The literature

This page exists so that nobody has to take the rest of the article on trust.  Every source it rests on is listed here with what it actually measured, who wrote it, and which way it cuts.

A catalogue of the literature bearing on erosion rates and the age of the earth — continental denudation, cosmogenic nuclide measurements, coastal retreat, isostasy, exhumation, sediment budgets, stratigraphic hiatuses, and the young-earth arguments built on all of them.

It is deliberately not a reading list of things that agree with this site.  Sources are tagged by the direction they cut, and a reader who wants only the strongest material against the argument can filter for undercuts and read those.

How to read an entry

Measurement or model.  The most important distinction here.  A cosmogenic nuclide concentration in a quartz grain is an observation.  A "global denudation rate" is the output of a model that extrapolates from a biased sample of gauged rivers, and it inherits every assumption in that extrapolation.  Both are useful.  They are not the same kind of thing, and conflating them is how most bad arguments on this subject get made — on both sides.

Timescale.  Erosion rates depend strongly on the interval over which they are measured.  A rate from a decade of river gauging, a rate from ten thousand years of cosmogenic nuclide accumulation, and a rate from a hundred million years of thermochronology are three different quantities, and they systematically disagree in a predictable direction.  Every entry states the integration time.

Direction.  Whether the source supports, undercuts, or complicates the young-earth erosion argument.  Assigned by what the source shows, not by what its authors believe — several mainstream papers are tagged supports and at least one creationist review is tagged undercuts.

Camp.  Mainstream, creationist, old-earth Christian, counter-apologetics, or unaffiliated.  This is disclosure, not a verdict.  A creationist paper reporting original field measurement is worth more than a mainstream paper restating a consensus, and the tag exists so a reader can weigh it rather than being told what to think.

Field definitions

The founding young-earth documents

Evolution: The Ocean Says No!

Nevins, Stuart E. [pseudonym of Steven A. Austin].  Impact 8, Institute for Creation Research  (November 1973).  Reprinted in Morris, Gish & Hillestad (eds.), Creation: Acts, Facts, Impacts, pp. 164-172 (1974).

https://www.icr.org/article/56

topic: denudation-global camp: creationist type: model access: OA timescale: annual gauging direction: supports

Key figures: continents above sea level 30.4 million cubic miles = 383 × 10¹⁵ t; ocean sediment 77 million cubic miles = 820 × 10¹⁵ t; total sediment input to the ocean 27.5 Gt/yr (20.2 suspended river load, 4.6 dissolved, 2.2 ice, 0.47 groundwater, 0.28 marine erosion, 0.06 wind, less 0.29 salt spray); continents levelled in 14 Myr; all ocean sediment accumulated in 30 Myr; sea-floor spreading destroys only 2.75 Gt/yr, a tenth of the input.

The founding document of the modern erosion argument, and better than its reputation.  Every calculation reproduces from its own inputs.  Its ocean-sediment volume, 3.21 × 10⁸ km³, is within about five per cent of the 2019 GlobSed measurement of 3.37 × 10⁸ km³ — a 1973 estimate that survived forty-six years of seismic surveying.  The failure is structural, not numerical: it divides a stock by a flux and calls the quotient an age, with no source term for uplift, no isostatic term, and no recycling term.  Its ocean-sediment half additionally compares against the age of the ocean rather than the age of the ocean floor, which is 64.2 Myr.

Some Questions about Geochronology

Roth, Ariel A.  Origins (Geoscience Research Institute), 13(2):64-85  (1986).

topic: denudation-global camp: creationist type: review access: OA timescale: various direction: supports

Key figures: Judson & Ritter's 6.1 cm/1000 yr for the United States; mean continental elevation 623 m; continents levelled in 10.2 Myr; Ahnert's basin range from 1.6 cm/1000 yr (Texas, England) to 100 (Himalaya) and 1,900 (a New Guinea volcano); Judson 1968 correcting for human activity gives 34 Myr; Gilluly et al. 1968 and Judson 1968 put the agricultural amplification at less than 2× and about 2.5×.

The most careful document in the tradition, and the source nearly every later creationist writer copies from.  Roth quotes the mainstream literature accurately, states the standard counter-explanations — episodic uplift, human amplification, drier past climates, sediment recycling, the incompleteness of the record — and answers each in turn rather than ignoring them.  Its central objection, that a full cycle of uplift and erosion would have eradicated the geologic column preserved in mountains, is the strongest form of the argument and is answered in the main article by metamorphic barometry.

Those Gaps in the Sedimentary Layers

Roth, Ariel A.  Origins (Geoscience Research Institute), 15(2):75-85  (1988).

topic: hiatus camp: creationist type: review access: OA timescale: n/a direction: supports

The paraconformity compilation that the flat-contact argument rests on, and the better half of Roth's case.  Everything later — Woodmorappe's Studies in Flood Geology, Fleming's Grand Canyon examples, the Creation magazine interview "'Millions of years' are missing" — traces here.

A strange admixture of erosion geology and tendentious theology

Woodmorappe, John [pen name].  Review of Fleming, Stories About Earth's History.  Journal of Creation, 38(3):33-35  (December 2024); posted online 20 April 2026.

https://creation.com/en/articles/erosion-challenges

topic: denudation-global camp: creationist type: review access: OA timescale: n/a direction: mixed

Key figures, all attributed to Fleming: Pacific coast average erosion 20 cm, Atlantic coast 78 cm (the review omits the time unit; reconstruction from the stated conclusion makes them cm/yr of shoreline retreat); North America gone in 4.2 Myr, with the two erosional fronts meeting near Salt Lake City; rivers alone level all continents in 22 Myr; Mount Everest at most 4.6 Ma against a conventional ~40 Ma; Himalayan long-term erosion rates on uniformitarian assumptions only 2% of present rates; the Pisco Formation's 640 m assigned 13 Myr.

Tagged mixed for one sentence.  Woodmorappe endorses the geology warmly and criticises only a theological digression — but he inserts the objection that sinks the argument: "However, I am not sure that erosion rates are independent of the topography.  Could it be, instead, that erosion rates are high at times of high relief (as today) but become much smaller as the continents approach peneplanation?  This should be considered."  It has not been.

Geologic Challenges to a Young Earth

Morton, Glenn R.  Proceedings of the First International Conference on Creationism, vol. 1, article 50, pp. 137-146  (1986).

topic: denudation-global camp: creationist type: review access: OA timescale: various direction: mixed

Key figures: continents lowered to sea level in 14 to 35 Myr; continental average erosion about 1 inch per thousand years (25.4 m/Myr); Zion high plateaus 1.5-3 ft/kyr (457-914 m/Myr); California Transverse Ranges 7.5 ft/kyr (2,286 m/Myr); buried southern Appalachian folds 10,000-20,000 ft high with up to 10,000 ft planed off, then covered by 2,500 ft of undisturbed Cretaceous, taking 1.3 Myr to strip at the fastest listed rate.

Morton was a young-earth creationist and an oil-industry geophysicist when he wrote this, and he calls continental denudation "the best geological argument" for a young earth.  In the same paper he supplies the double edge: "However, the young earth creationists, like the old earthers, have generally ignored the problem presented by these same erosion rates when regional unconformities are discussed."  He later left young-earth creationism.

Discussion of Morton, "Geologic Challenges to a Young Earth"

DeVilbiss, John W.  Same proceedings, pp. 144-145  (1986).

topic: uplift camp: creationist type: review access: OA timescale: n/a direction: undercuts

The single most important item in the young-earth corpus, and almost unknown.  A creationist, at a creationist conference, in print, gives the coupled-system rebuttal: "A mountain chain does not start out at maximum height.  It 'grows' in the presence of erosion, and as it 'grows' higher, the erosion becomes exponentially greater… Thus, in the presence of erosion proportional to altitude, a mountain would never get unreasonably high even if very, very old."  And then: "The balance is between growth, altitude and erosion rate, not merely between maximum altitude and erosion rate."  Morton's printed reply — that the Appalachians stopped growing 200 Myr ago and so are the hard case — is a fair counter and is the live crux.

Moon Dust and the Age of the Solar System

Snelling, Andrew A. and Rush, David E.  Creation Ex-Nihilo Technical Journal, 7(1):2-42  (1993).

https://creation.com/en/articles/moon-dust-and-the-age-of-the-solar-system

topic: history camp: creationist type: review access: OA timescale: n/a direction: undercuts

Key figures: the retired argument's input, 14,300,000 t/yr of meteoritic dust (Pettersson 1960), against modern satellite and radar estimates of 10,000-20,000 t/yr — wrong by roughly three orders of magnitude.

Not about erosion, and included because it is the procedural template.  Two young-earth creationists spent seven years killing a popular young-earth argument in their own peer-reviewed journal, and said so plainly: "Should creationists then continue to use the moon dust as apparent evidence for a young moon, earth and solar system?  Clearly, the answer is no."  The test they applied — is the input rate measured, is it representative, and has the removal term been accounted for — is exactly the test the erosion argument fails.

The global sediment flux to the ocean

Erosion of the land, or what's happening to our continents?

Judson, Sheldon.  American Scientist, 56(4):356-374  (1968).

topic: denudation-global, human-impact camp: mainstream type: model access: paywalled timescale: mid-20th-century river gauging direction: undercuts

Key figures: present global denudation 24 × 10⁹ t/yr; pre-human global denudation 9 × 10⁹ t/yr; amplification 2.67×.  At 2.7 t/m³ over 148.48 × 10⁶ km² of land, 24 Gt/yr is 59.9 m/Myr and 9 Gt/yr is 22.4 m/Myr.  Levelling times against 3.19 × 10¹⁷ t of rock above sea level: 13.3 Myr at the modern flux, 35.4 Myr at the pre-human flux.

The most damaging single item in this group, and it comes from the man who supplied the creationists' headline number.  Judson published the 6.1 cm/1000 yr United States rate with Ritter in 1964; four years later he said in print that nearly two-thirds of the modern global flux is anthropogenic.  Any use of the 1964 rate as a natural rate is quoting half a source.  Ariel Roth is the one creationist writer who reports the correction, and reports it accurately: "Judson (1968), while correcting for human activity, suggests 34 Ma for complete erosion of the continents."

Provenance, stated plainly.  The paper itself was not obtained — American Scientist v.56 sits behind JSTOR, and the issue index was the furthest this research reached.  The 9 and 24 Gt/yr figures come from Gregor, B. (1970), "Denudation of the Continents," Nature 228:273, whose fetched abstract says "man's intervention has increased the overall denudation rate from 9 to 24 billion tons a year."  Roth's "34 Ma" does not reproduce from height over rate (623 m ÷ 22.4 m/Myr = 27.8 Myr) but does reproduce from mass over flux (35.4 Myr), so Roth's figure is substantially verified even though Judson may never have written the number 34.  Whether Judson's totals include dissolved load is probable but not certain; Gregor's wording, "overall denudation rate," suggests yes.

Impact of Humans on the Flux of Terrestrial Sediment to the Global Coastal Ocean

Syvitski, James P.M.; Vörösmarty, Charles J.; Kettner, Albert J.; and Green, Pamela.  Science, 308(5720):376-380  (2005).  doi 10.1126/science.1109454

https://geodynamicsprogram.whoi.edu/wp-content/uploads/sites/14/2018/10/of_Humans_on_the_Flux_of_Terrestrial_Sediment_75325.pdf

topic: sediment-budget, human-impact camp: mainstream type: model access: paywalled timescale: 1960-1995 gauging against a modelled pre-human state direction: mixed

Key figures: pre-human suspended flux 14 Gt/yr, or 15.5 Gt/yr with bed load; modern flux 12.6 Gt/yr, "or 10% less than the pre-Anthropocene load"; soil erosion adds +2.3 ± 0.6 Gt/yr, reservoirs subtract −1.4 ± 0.3 Gt/yr; gross modern flux absent reservoirs 16.2 Gt/yr (17.8 with bed load); large reservoirs trap 20% of the global flux and small reservoirs another 6%; over 100 Gt of sediment now sequestered behind dams.  Observed loads for rivers draining 70% of the land surface, 217 of them with before-and-after-dam records.  Converted over all land at 2.7 t/m³: 14 Gt/yr = 34.9 m/Myr, 16.2 = 40.4, 12.6 = 31.4.  Levelling times 22.8, 19.7 and 25.3 Myr bare; 137, 118 and 152 Myr with the ×6.0 Airy isostatic correction.

The most careful modern sediment budget, and it cuts both ways at once, which is why it is tagged mixed rather than either extreme.  Read as a global mass balance it undercuts the young-earth argument: the natural flux of 14 Gt/yr levels the continents in 23 Myr bare and 137 Myr with isostasy, against a 4.5 Gyr earth.  Read as a test of the human-amplification rebuttal it does the opposite: on Syvitski's own numbers the modern ocean flux is lower than the pre-human flux, 12.6 against 14, because reservoirs trap more than agriculture adds.  If this paper is right, no downward correction of the modern flux to a "natural" value is warranted at all.  The two readings are not in conflict; they answer different questions, and both belong in the article.

The honest caveat is that the 12.6 Gt/yr modern figure is far below the Milliman-family estimates of 19-20 Gt/yr suspended, and the two literatures have never been reconciled.  That unresolved gap of nearly 2× is larger than the human-amplification signal being argued over.

Erosion and sediment yield: a global overview

Walling, Desmond E. and Webb, Bruce W.  In Erosion and Sediment Yield: Global and Regional Perspectives, IAHS Publication 236:3-19  (1996).

topic: sediment-budget camp: mainstream type: compilation access: OA timescale: years to decades of gauging, estimates from 1950 to 1996 direction: background

Key figures: their Table 1 collects eighteen published estimates of total suspended sediment transport to the oceans, spanning 8.3 to 51.1 Gt/yr — a factor of six.  In m/Myr over all land at 2.7 t/m³ that is 20.7 to 127.5 m/Myr, and in bare levelling time 6.2 to 38.4 Myr (37 to 231 Myr with the ×6.0 isostatic correction).  Named endpoints: Fournier 1960 at 51.1 Gt/yr, Kuenen 1950 at 32.5, Gilluly 1955 at 31.7, Holeman 1968 at 18.3, Milliman & Meade 1983 at 13.5, Mackenzie & Garrels 1966 at 8.3.

The table that makes the whole dispute legible, and the reason no single number should be quoted without saying which one and why.  Walling & Webb's own verdict on the high end is blunt: Fournier's 51.1 Gt/yr came from a database over-weighted with high-yield Chinese rivers and "was undoubtedly an overestimate."  Their verdict on the centre is that "values closer to 15-20 X 10⁹ t year-1 have been most frequently cited."  They also note Milliman & Meade's own upward revision of 13.5 to about 14 Gt/yr to represent the natural flux — an independent arrival at Syvitski's 14.

Four axes explain nearly the whole spread, and a reader who holds them in mind will not be misled by any figure in the literature.  Which load is counted: suspended only is the default, dissolved load adds roughly 15-35%, bed load a further 10%.  How ungauged small rivers are handled: Milliman & Syvitski's correction moved the total from 13.5 to 20 Gt/yr by itself.  Where the measurement is taken: "to tide water," "to the coast," and "generated on land" are three different quantities.  And whether human amplification is corrected for, which is the axis the creationist literature ignores.

A density trap recorded here so nobody repeats it.  Roth 1986 attributes to Gilluly 1955 both "13.6 km³ of solid material" per year and "about 31,000 million tons/year."  Those are consistent only at 2.33 t/m³.  Spread over 148.48 × 10⁶ km², 13.6 km³/yr is 91.6 m/Myr, while 31.7 Gt/yr at 2.7 t/m³ is 79.1 m/Myr.  The article must pick one density and say which.  Gilluly 1955 itself could not be fetched — GeoScienceWorld returns HTTP 403 to automated requests — so the 31.7 Gt/yr figure here rests on Walling & Webb's compilation, and the volumetric form rests on Roth alone.

Geomorphic/Tectonic Control of Sediment Discharge to the Ocean: The Importance of Small Mountainous Rivers

Milliman, John D. and Syvitski, James P.M.  Journal of Geology, 100(5):525-544  (1992).  doi 10.1086/629606

topic: sediment-budget camp: mainstream type: compilation access: paywalled timescale: mid-20th-century gauging, pre-dam direction: mixed

Key figures: 280 rivers analysed; pre-dam mid-twentieth-century flux to the ocean about 20 Gt/yr, the highest well-supported modern estimate; pre-agricultural flux "probably was less than half the present level," so under 10 Gt/yr; small mountainous river fluxes had been underestimated in earlier global budgets "perhaps by as much as a factor of three."  At 2.7 t/m³ over all land, 20 Gt/yr is 49.9 m/Myr, levelling the continents in 15.9 Myr bare and 96 Myr with isostasy.

The paper that raised the accepted global flux by half, and it is the strongest number a young-earth writer can reach for.  Sediment load turns out to be a log-linear function of basin area and maximum basin elevation, with climate and runoff secondary — which is why thousands of short, steep rivers draining Oceania and southern Asia had been missed.  Nearly half the 20 Gt/yr comes from Oceania alone and another third from southern Asia.

Two things must be said alongside the 20 Gt/yr.  The same authors put the natural pre-agricultural flux at under half of it, and they warn that in some basins "sediment loads may have increased by a factor of 2-10 since humans began farming."  And the 20 Gt/yr is suspended load only; the authors say so explicitly, noting "larger contributions from bedload material, which seldom is included in the sediment load values reported in the literature."  Adding bed load pushes the number up; correcting for agriculture pushes it further down.

The changing sediment loads of the world's rivers

Walling, Desmond E.  In Sediment Dynamics in Changing Environments, IAHS Publication 325:323-338  (2008).

https://iahs.info/uploads/dms/14531.49-323-338-42-325-walling.pdf

topic: sediment-budget, human-impact camp: mainstream type: review access: OA timescale: decades of gauging direction: mixed

Key figures: the twenty-five-year consensus on global land-ocean flux is "values around 15 Gt year-1."  Against Syvitski's 3.6 Gt/yr of reservoir trapping, Walling argues for 24 Gt/yr, giving a gross contemporary flux absent dams of 36.6 Gt/yr — the largest number in the literature, 91.3 m/Myr, a bare levelling time of 8.7 Myr and 52 Myr with isostasy.  He holds the pre-human flux at 14 Gt/yr, so his own human amplification is 160%.  His route: the ICOLD register of about 33,000 dams, 0.8%/yr storage loss on about 6,000 km³ of capacity, giving ~48 km³/yr, which at 1.2 t/m³ dry bulk density is ~60 Gt/yr sequestered, then a tentative 60% floodplain conveyance loss.

Double-edged in the sharpest possible way.  A creationist could quote the 36.6 Gt/yr and halve the standard levelling time; Walling's entire point in producing it is that the excess is anthropogenic.  Take his numerator and his denominator must come too.  He also calls his own new estimates "speculative" in the text, which any use of them should repeat.

The paper is also the only accessible source for Walling & Fang (2003), Global and Planetary Change 39(1-2):111-126, doi 10.1016/S0921-8181(03)00020-1, which could not be fetched.  Walling restates it verbatim: of 145 rivers with long records, "approx. 48% were essentially stationary, whilst the remaining 52% provided evidence of statistically significant trends.  Of these, about one tenth were increasing and the remainder were decreasing."  Roughly nine in ten changing rivers are carrying less sediment, not more, which undercuts any claim that erosion is accelerating.  Walling supplies his own sampling caveat: the 145 rivers were "drawn exclusively from the northern hemisphere and included no rivers in Africa or South America."  The Yellow River is the extreme case — a load "frequently cited as 1.6 Gt year-1," delivering 1.08 Gt/yr to the ocean in the 1980s, about 0.4 Gt/yr in the 1990s and possibly 0.15 Gt/yr by 2007, a decline equal on its own to 5-7% of the global land-ocean flux.

An error in the source, recorded so it is not propagated.  Walling's Table 1 lists the Syvitski column as "increase due to human activity 22%" and "reduction from reservoir trapping 16%."  The two are transposed: 14 to 16.2 Gt/yr is +16%, and 16.2 to 12.6 is −22%.  His running text has it right.  Cite the text, not the table.

Eroding ages: If our continents were old, they would no longer be here

Walker, Tas.  Creation, 22(2):18-21  (March 2000); web version updated 10 February 2024.

https://creation.com/en/articles/eroding-ages

topic: denudation-global camp: creationist type: popular access: OA timescale: annual river gauging direction: supports

Key figures: "The average height reduction for all the continents of the world is about 60 mm (2.4 inches) per 1,000 years, which equates to some 24 billion tonnes of sediment per year" — that is 60 m/Myr and 24 Gt/yr.  North America levelled in 10 Myr.  Fastest quoted regional rate, "19,000 mm (750 inches) per 1,000 years from a volcano in Papua New Guinea," or 19,000 m/Myr.

The version of the argument most readers have met, and its arithmetic is sound.  24 Gt/yr ÷ 2.7 t/m³ ÷ 148.48 × 10⁶ km² is 59.9 m/Myr, so the two headline numbers are internally consistent and use total land area at rock density.  The mean continental height of 623 m matches Roth's.  Creation Ministries International also flags, creditably, two uncertainties that push the true figure up: unmeasured bed load — "It is difficult to be exact because some sediment is rolled or pushed along the bottom of the river.  'Bed load', as it is called, is not easily observed" — and rare catastrophic events.

The flaw is one the article must not repeat.  24 Gt/yr is Judson 1968's human-amplified figure, and Judson's own natural figure in the same paper is 9 Gt/yr, or 22.4 m/Myr.  Presenting 60 mm/1000 yr as "the average height reduction for all the continents" without noting that the source attributed roughly five-eighths of it to agriculture and deforestation states less than the source does.  The uncertainty that pushes the number down is the one left out.

Denudation

Dole, Roger B. and Stabler, Herman.  In Papers on the Conservation of Water Resources, U.S. Geological Survey Water-Supply Paper 234, pp. 78-93  (1909).

https://pubs.usgs.gov/wsp/0234/report.pdf

topic: denudation-global camp: mainstream type: measurement access: OA timescale: one year of daily sampling, about 150 rivers direction: mixed

Key figures: United States total, Great Basin excluded, 270 × 10⁶ short tons dissolved plus 513 × 10⁶ short tons suspended = 783 × 10⁶ short tons/yr = 710.7 × 10⁶ t/yr over 3,088,500 sq mi (7.999 × 10⁶ km²), a specific yield of 88.8 t/km²/yr.  Stated as "thirteen ten-thousandths of an inch per year, or 1 inch in 760 years" — 33.4 m/Myr.  Including the Great Basin, 1 inch in 730 years, or 34.8 m/Myr.  Per-basin years required for 1 inch: Colorado 440, Mississippi 500, South Atlantic 710, North Atlantic 1,100, Hudson Bay 3,900.  Dissolved load is 34% of the total by mass.  About 5,000 complete mineral analyses underlie it.

The oldest and least human-disturbed United States measurement, and it is little more than half the rate the creationist literature uses.  Judson & Ritter's 61 m/Myr is 1.83× this number.  That ratio confirms Roth's own statement that Dole & Stabler "gave figures indicating that it would take about twice as long," and it shows that the spread in this quantity has been public since 1909.

Their arithmetic is internally consistent: 88.8 t/km²/yr at their own stated density of 165 lb/ft³ (2.643 t/m³) gives 33.6 m/Myr, within 0.6% of their stated figure.  Their caveats are their own and should travel with the number: run-off and dissolved solids within about 10%, but "it is possible that denudation estimates, based on average suspended matter for one year, may be in error as much as 50 per cent," and calculated denudation "generally within 20 per cent of the true values."  They also warn that suspended figures "represent more properly material carried to tide water" — that is, to the point where the current slows and deposition begins, not to the deep ocean.  Bed load is excluded outright.

Fluvial-Sediment Discharge to the Oceans from the Conterminous United States

Curtis, Willard F.; Culbertson, James K.; and Chase, Edith B.  U.S. Geological Survey Circular 670, 17 pp.  (1973).

topic: sediment-budget camp: mainstream type: measurement access: OA timescale: 20 years of gauging, 1950-1969 direction: undercuts

Key figures: 27 drainage areas over 1950-1969; total suspended discharge to the oceans 491,449,600 short tons/yr = 445.8 × 10⁶ t/yr from 2,658,776 sq mi (6.886 × 10⁶ km²), a mean yield of 64.7 t/km²/yr.  At 2.7 t/m³ that is 24.0 m/Myr, or 26.4 m/Myr with the paper's own 10% bed-load allowance.  Regional yields, short tons per sq mi per year: Atlantic 49.5, Pacific 156.6, Gulf of Mexico 217.4.  Colorado River at Yuma fell from 966 short tons/sq mi/yr in 1911-16 to 0.63 in 1965-67.

The best open United States cross-check on the number the whole argument runs on, and it comes in at 40% of Judson & Ritter's 61 m/Myr.  Part of the gap is load type — this is suspended only, and Dole & Stabler put dissolved load at about a third of the total mass.  Part is reservoir trapping in the five years between the two studies.  Either way, the highest-quality mid-century United States measurement is well below the figure the creationist literature reuses.

The authors state both limits themselves.  On bed load: "the data are based on suspended-sediment samples obtained with standard United States depth-integrating samplers and, therefore, do not include that part of the total sediment discharge moving as bedload.  The quantity of sediment transported as bedload may be estimated at about 10 percent."  On dams: "yields computed on the basis of total drainage area can be deceiving because good land-use practices and multiple reservoirs in the drainage area can dramatically reduce the amount of sediment delivered to the sea."  The circular also documents that Holeman 1968 over-predicts the United States total by 33% because Holeman used the pre-reservoir Colorado record of 1925-57.

Rates of regional denudation in the United States — what could not be verified

Judson, Sheldon and Ritter, Dale F.  Journal of Geophysical Research, 69(16):3395-3401  (1964).  doi 10.1029/JZ069i016p03395

topic: denudation-global, method camp: mainstream type: measurement access: paywalled timescale: years of river gauging direction: background

Key figures: the reported rate is 6.1 cm/1000 yr = 61 m/Myr, which against a 623 m mean continental elevation gives Roth's 10.2 Myr.  Per-river suspended yields survive in USGS Circular 670's Table 2, in short tons per sq mi per year: Mississippi 244, Delaware 147, Columbia 125, Alabama 97, Sacramento 94, Colorado 1,190, Eel 5,846, and the whole Pacific drainage 623.

This entry exists to record a gap, not a finding.  The 1964 paper could not be obtained.  Wiley returns HTTP 403 on both the landing page and the PDF; Unpaywall reports is_oa: false with no OA locations; OpenAlex reports oa_status: closed and any_repository_has_fulltext: false across 198 citations; Semantic Scholar reports CLOSED with the abstract elided; and the Wayback Machine holds no snapshot of any AGU or Wiley URL form.  Everything the creationist literature says about this paper is second-hand.

The attribution chain is short and runs through one man.  Roth 1986 writes: "Judson and Ritter (1964) have estimated that for the United States the rate of erosion averages 6.1 cm/1000 years.  At this rate of denudation the continents, which average 623 m above sea level, would be eroded to sea level in a mere 10.2 Ma."  Every later creationist treatment copies Roth.  Independent corroboration that the paper says something close to this does exist: SCOPE 32 chapter 12 states that at 0.75 t/ha/yr "the land mass of the United States lying above sea level could be moved to the ocean in 11-12 million years (Judson and Ritter, 1964)," and Bierman & Nichols 2004 record that "in 1964, Judson & Ritter used contemporary suspended sediment loads to infer the erosion rate of North America."

Four things remain unknown and all of them matter: the full regional table, the sampling period, the authors' own caveats, and — most consequentially — whether 6.1 cm/1000 yr is suspended load only.  Bierman & Nichols say suspended; but 61 m/Myr is high enough for a suspended-only United States figure that dissolved load is probably included, and the two readings are not compatible.  So far as this research could establish, nobody in the creationist literature has ever checked the number against the original.  The article should say so.

River Discharge to the Coastal Ocean: A Global Synthesis

Milliman, John D. and Farnsworth, Katherine L.  Cambridge University Press  (2011).  doi 10.1017/CBO9780511781247

https://www.cambridge.org/core/books/abs/river-discharge-to-the-coastal-ocean/introduction/1A5DB5678A49D084B100143A6D344C9C

topic: sediment-budget camp: mainstream type: book access: book timescale: modern gauging direction: background

Key figure: rivers "annually discharge about 36 thousand km3 of freshwater and more than 20 billion tons (Bt) of solid and dissolved material to the world ocean."

Included as a warning label.  This is the standard modern reference and almost certainly the source of the 24.0 Gt/yr modern comparator that Wittmann et al. 2020 measure their cosmogenic background against — which makes it load-bearing for the human-amplification concession.  But its ">20 Bt" combines solid and dissolved load, and every other figure in the global tables is suspended only.  Comparing it directly with Syvitski's 12.6 Gt/yr or Milliman & Syvitski's 20 Gt/yr is comparing different quantities.

The book itself was not obtained.  Only the publisher's chapter-one abstract was fetched, and the quoted sentence comes from it.  The frequently cited "1,534 rivers" database size is publisher marketing copy and is not verified.

Human amplification: what got multiplied, and what did not

The impact of humans on continental erosion and sedimentation

Wilkinson, Bruce H. and McElroy, Brandon J.  Geological Society of America Bulletin, 119(1/2):140-156  (2007).  doi 10.1130/B25899.1

http://www.rpgroup.caltech.edu/aph150_human_impacts/assets/pdfs/Wilkinson_2007.pdf

topic: human-impact, denudation-global camp: mainstream type: compilation access: paywalled timescale: annual soil plots against 10^8^ yr rock volumes direction: mixed

Key figures: Phanerozoic mean continental denudation from Ronov's preserved sedimentary rock volumes, corrected for erosional and subductional destruction, 16 m/Myr = about 5 Gt/yr, with the twenty-five epoch intervals running from about 4 m/Myr in the Middle Triassic to about 53 m/Myr in the Pliocene and a standard deviation only 80% of the mean.  Modern large-river loads require about 62 m/Myr = about 21 Gt/yr over 118 × 10⁶ km² of ice-free land.  Global farmland soil loss taken as 600 m/Myr over 50 × 10⁶ km² of cropland and pasture, giving 75 Gt/yr.  USDA-ARS soil plots, 1,406 plots and over 700 annual observations, mean 885 m/Myr; the USDA USLE map of United States croplands, mean 250 m/Myr.  Post-settlement alluvium at 15 United States sites accumulates at 12,600 m/Myr.  Their fitted relation is denudation (m/Myr) = 7.21 e^0.0015E, where E is elevation in metres, r = 0.65.  Dams retain about 4 Gt/yr.  Contiguous-United States mean natural denudation about 21 m/Myr.

The paper the whole amplification argument runs on, and the one that quietly answers it.  Two sentences from it settle where the world's sediment comes from: "~83% of the global river sediment flux is derived from the highest 10% of Earth's surface," and "~83% of cropland erosion occurs over the lower 65% of Earth's surface."  The natural sediment machine is mountains; agriculture is on the plains.  The two barely overlap, which is why a 30-fold multiplier on farm fields does not become a 30-fold multiplier on the planet.  Their post-settlement alluvium data then shows where the farm soil goes: it needs an aggrading area only about 5% of the degrading area to absorb it, so it is stored, not exported.  It was never in the 21 Gt/yr river figure to begin with.

Their deep-time series is also a direct rebuttal to any claim that pre-human erosion ran orders of magnitude faster.  Across twenty-five Phanerozoic epochs the range is 4 to 53 m/Myr, the maximum falls in the Pliocene and the minimum in the mid-Mesozoic, and the standard deviation is smaller than the mean.  Whatever varied over 540 Myr, it did not vary by orders of magnitude.

Three internal inconsistencies, recorded so the article does not inherit them.  The Pliocene is 16 Gt/yr in the abstract and 19 Gt/yr in the text.  The modern river requirement is "~62 m/m.y." in one place and "~71 m/m.y. over the globe's entire exposed land" in another.  The agricultural flux is 75 Gt/yr in the text and 63 Gt/yr in the figures.  And their stated ratios do not follow from their own numbers: they say 600 m/Myr is "~30 times" the deep-time rate of 16 and "~7 times" the modern river rate of 62, but 600/16 is 37.5 and 600/62 is 9.7.  Do not repeat "30×" as though it followed from the paper's figures.  To their credit, the authors flag the largest weakness themselves, citing Trimble's finding of an 80% reduction in sediment flux at Coon Creek, Wisconsin between 1853 and 1975: "the more recent cropland flux estimated at 75 Gt/yr might in fact be too high… at this point in time, we just do not know if this is the case."

Soil erosion and agricultural sustainability

Montgomery, David R.  Proceedings of the National Academy of Sciences, 104(33):13268-13272  (2007).  doi 10.1073/pnas.0611508104

https://pmc.ncbi.nlm.nih.gov/articles/PMC1948917/

topic: human-impact camp: mainstream type: compilation access: OA timescale: plot-years to 10^6^ yr direction: mixed

Key figures, medians converted to m/Myr: conventional agriculture 1,537 (n = 448); conservation agriculture 82 (n = 47); native vegetation 13 (n = 65); soil production 17 (n = 188); geological 29 (n = 925).  Paired comparisons: no-till against conventional, median reduction 20× (n = 39, range 2.5× to 7,620×); native vegetation against conventional cultivation, median 18× (n = 46).  USDA tolerable-soil-loss "T values" are 400-1,000 m/Myr, set 20-60× above his own median soil-production rate.  By setting: ancient cratons below 0.1 to 10 m/Myr; soil-mantled moderate hillslopes 1-1,000 m/Myr; steep tectonically active alpine terrain 100 to over 10,000 m/Myr.

The origin of the hundredfold rebuttal, and the table that shows what the hundredfold is a fact about.  Montgomery's own summary is exact: "erosion rates from conventionally plowed agricultural fields average 1-2 orders of magnitude greater than rates of soil production, erosion under native vegetation, and long-term geological erosion."  That is a per-unit-area statement about cultivated ground, and it is well replicated across 1,673 measurements.  Creationists should not dispute it.

What belongs in a global mass balance is not the 1,537 but the 29 and the 13 — his geological and native-vegetation medians — and those bracket Wilkinson & McElroy's 16 m/Myr from an entirely different method.  Montgomery also volunteers a bias that runs in the young-earth argument's favour: geologists over-sample scenic alpine terrain, "which would imply that the median value reported here is, if anything, biased high."  His independent check on the low end is worth noting too — Montgomery & Brandon's relation E = 0.2R applied to a global median relief of 86 m returns 17 m/Myr, "identical to the median soil production value" in his compilation.

The human impact on North American erosion, sediment transfer, and storage in a geologic context

Kemp, David B.; Sadler, Peter M.; and Vanacker, Veerle.  Nature Communications, 11:6012  (2020).  doi 10.1038/s41467-020-19744-3

https://www.nature.com/articles/s41467-020-19744-3

topic: timescale-bias, human-impact camp: mainstream type: compilation access: OA timescale: 41 kyr down to 1 yr, timespan-matched direction: mixed

Key figures: 4,754 alluvium accumulation rates from 400 sites, mined from 183 publications, ages 41 ka BP to 2007 CE, spanning eight orders of magnitude in rate.  From 41 ka to about 200 years ago the median accumulation rate is 0.7 mm/yr = 700 m/Myr, with little variation with age.  The raw median for rates younger than 200 years is 75 mm/yr — and that number must not be used.  Timespan-matched, at averaging windows of 40-400 years, post-settlement medians are 6-24 mm/yr against pre-settlement medians of 0.6-1 mm/yr, "consistently ~10× faster," resting on 238 pre-settlement and 281 post-settlement rates after de-duplication by site.  94% of sites (119 of 126) exceed the expected geologic rate; 39% exceed it by at least 10×.  Change-point analysis places the step between about 1720 and 1920 CE, median about 1820.  Alluvial stores cover under 7% of the agricultural landscape.

The methodological gift of this group, and the demonstration that settles how erosion rates may be compared at all.  The raw post-settlement median is 75 mm/yr and the honest timespan-matched multiplier is 10×, not 100×, and the difference is entirely the Sadler effect — shorter averaging windows return higher rates, because they miss the quiet intervals.  Kemp and colleagues prove this empirically with 4,754 measurements.  Any erosion-rate comparison in this article that is not timespan-matched, or explicitly flagged as unmatched, is worthless.  That cuts against sloppy creationist arithmetic and sloppy uniformitarian arithmetic in equal measure.

The headline line — "humans have moved as much material in North America during the last century as would be moved by natural processes in ~700-3000 y" — is a strong soundbite against a young earth and a weak argument, because it concerns material moved a short distance into local alluvial storage covering under 7% of the farmed landscape.  Their post-settlement figure of about 10 mm/yr independently matches Wilkinson & McElroy's 12.6 mm/yr from 15 United States sites.

Humans as geologic agents: A deep-time perspective

Wilkinson, Bruce H.  Geology, 33(3):161-164  (2005).  doi 10.1130/G21108.1

topic: human-impact camp: mainstream type: model access: paywalled timescale: 5 × 10^8^ yr against the present direction: undercuts

Key figures, from the abstract only: mean denudation over the past half-billion years has lowered continental surfaces by "a few tens of meters per million years"; construction plus agriculture now move enough material to lower all ice-free continental surfaces by "a few hundred meters per million years"; humans became the prime agents of erosion "sometime during the latter part of the first millennium A.D."

The source of the sentence most often quoted at creationists: "Humans are now an order of magnitude more important at moving sediment than the sum of all other natural processes operating on the surface of the planet."  That sentence is true and it is about total material displaced, including construction and mining, not about sediment delivered to the ocean.  Wilkinson's own 2007 follow-up shows the displaced material stays on the continent.  Quoting the 2005 line against a global mass-balance argument, without the 2007 partition, is a category error.

Only the abstract was obtained.  GeoScienceWorld returns HTTP 403, Unpaywall reports is_oa: false, and Semantic Scholar reports CLOSED.  No number from the body of this paper is verified here, and any specific Gt/yr figure attributed to Wilkinson 2005 in the secondary literature should be treated as unchecked.

On the history of humans as geomorphic agents

Hooke, Roger LeB.  Geology, 28(9):843-846  (2000).  doi 10.1130/0091-7613(2000)28<843:OTHOHA>2.0.CO;2

https://www.uvm.edu/~geomorph/geol151/2012/Hooke_2000_2605.pdf

topic: human-impact camp: mainstream type: model access: paywalled timescale: 5,000 yr direction: mixed

Key figures: intentional earth moving for construction, roads and mineral production runs at about 30 t/yr per capita in the United States and about 6 t/yr per capita worldwide.  Model inputs for unintentional soil loss: developing-world cropland about 35 t/ha/yr, pasture about 10 t/ha/yr, which at 2.7 t/m³ are 1,300 m/Myr and 370 m/Myr.  Headline: "The total earth moved in the past 5000 yr would be sufficient to build a 4000-m-high mountain range, 40 km wide and 100 km long."  Treated as a rectangular prism at 2.0 t/m³ that is 16,000 km³, or about 6.4 Gt/yr averaged over five millennia; the shape is ambiguous, so the conversion is order-of-magnitude only.

Included for one sentence that geomorphologists write and rebuttals never quote.  Of the products of human-accelerated erosion Hooke writes: "They accumulate as colluvium on slopes and as alluvium on floodplains downstream from the fields."  Moved "only a short distance" is his phrase.  That is the concession that human-accelerated soil erosion does not translate into human-accelerated sediment delivery, stated by a mainstream author with no stake in the creation debate.

A misattribution to stop repeating.  The widely circulated figures "1,500 Gt plowed annually of which 75 Gt/yr is eroded" and "humans displace 129 billion tonnes/yr against 111 for all natural processes" appear nowhere in the four pages of this paper.  They probably come from Hooke 1994, GSA Today 4:217, or from Haff 2003.  Attribute them correctly or drop them.

Natural streams and the legacy of water-powered mills

Walter, Robert C. and Merritts, Dorothy J.  Science, 319(5861):299-304  (2008).  doi 10.1126/science.1151716

topic: human-impact, sediment-budget camp: mainstream type: measurement access: paywalled timescale: 300-400 yr direction: supports

Key figures, from the abstract only: 1 to 5 metres of slackwater sediment accumulated behind "tens of thousands of 17th- to 19th-century milldams" across the mid-Atlantic United States.

The physical mechanism behind the 6% figure, and it favours the young-earth reading.  An entire region's worth of accelerated soil erosion was intercepted a few kilometres downslope by tens of thousands of small dams and never reached the ocean at all.  Most mid-Atlantic floodplains turn out to be fill terraces, not natural meander floodplains.

The paper also bears on what the natural landscape was doing before any of that: "before European settlement, the streams were small anabranching channels within extensive vegetated wetlands that accumulated little sediment but stored substantial organic carbon."  Slow-moving, sediment-poor systems — not a landscape shedding material fast.

Only the abstract was obtained.  Three PDF mirrors failed (Columbia LDEO 404, Carleton SERC 403, Franklin & Marshall 404).  Any burial rate, radiocarbon date or valley-fill volume beyond "1 to 5 meters" is unverified.

Anthropogenic sediment retention: major global impact from registered river impoundments

Vörösmarty, Charles J.; Meybeck, Michel; Fekete, Balázs; Sharma, Keshav; Green, Pamela; and Syvitski, James P.M.  Global and Planetary Change, 39(1-2):169-190  (2003).  doi 10.1016/S0921-8181(03)00023-7

topic: human-impact, sediment-budget camp: mainstream type: model access: paywalled timescale: decades direction: supports

The text of this paper was not obtained and nothing from it is quoted here.  ScienceDirect returns 403, the ADS abstract page returned empty, OpenAlex stores no abstract, Semantic Scholar's is publisher-elided, and Unpaywall reports is_oa: false.  The bibliographic metadata above is verified through Crossref; the content is not.

Listed because it is the standard citation for reservoir trapping and the article will need it.  What can be said safely comes from sources that were fetched.  Wilkinson & McElroy 2007 report that "Vörösmarty et al. (2003) reported a tripling (from 5% to 15%) in the storage of sediment behind dams in large reservoirs" and "a 6-fold increase in the amount of riverine particulate sediment retained in large reservoirs between 1950 and 1985."  Syvitski et al. 2005 give the magnitudes directly: large reservoirs trap 20% of the global sediment flux and small reservoirs another 6%.  Widely circulated figures of 633 large reservoirs, over 40% of global river discharge intercepted, and 30% plus 23% trapping came from a search summary rather than a fetched source, and should not be printed until the paper is in hand.

Which way it cuts is clear even at second hand: dams intercept a quarter to a half of the moving sediment, so the modern flux measured at the coast understates the natural flux rather than overstating it.

An assessment of the global impact of 21st century land use change on soil erosion

Borrelli, Pasquale and others.  Nature Communications, 8:2013  (2017).  doi 10.1038/s41467-017-02142-7

topic: human-impact camp: mainstream type: model access: OA timescale: one year, modelled direction: supports

Key figures: 35.9 Pg/yr (= 35.9 Gt/yr) of soil eroded globally in 2012, modelled with RUSLE at 250 m resolution.  The 2020 follow-up — Borrelli et al., PNAS 117(36):21994-22001, doi 10.1073/pnas.2001403117 — gives a 2015 baseline of 43 (−7 / +9.2) Pg/yr, with current conservation-agriculture practice cutting it by about 5%; gullying and river-bank erosion are excluded.

Useful because it deflates the largest number the other side uses.  Wilkinson & McElroy's headline 75 Gt/yr of agricultural erosion is roughly double Borrelli's modelled global total, and Borrelli says so directly: "We challenge the previous annual soil erosion reference values as our estimate, of 35.9 Pg yr⁻¹ of soil eroded in 2012, is at least two times lower."  For a defensible modern global hillslope figure the article should use 36-43 Gt/yr and flag 75 Gt/yr as a high-end outlier that its own authors doubted.

Note the arithmetic that follows.  Even 43 Gt/yr of hillslope erosion yields only about 24 Gt/yr of delivery to the ocean, and that delivery figure includes dissolved load.  Most of what comes off the hillslope stops before it reaches the sea.

Only the abstracts of both papers were obtained, via OpenAlex and Semantic Scholar respectively.  Open-access PDFs of both exist and were not fetched.

People have shaped most of terrestrial nature for at least 12,000 years

Ellis, Erle C. and others.  Proceedings of the National Academy of Sciences, 118(17):e2023483118  (2021).  doi 10.1073/pnas.2023483118

https://pmc.ncbi.nlm.nih.gov/articles/PMC8092386/

topic: human-impact camp: mainstream type: model access: OA timescale: 12,000 yr direction: mixed

Key figures: "Even 12,000 y ago, nearly three quarters of Earth's land was inhabited and therefore shaped by human societies, including more than 95% of temperate and 90% of tropical woodlands."  Uninhabited "Wildlands" were therefore about 27% of land at 10,000 BCE.  Of land under protection today, only 36.4% was uninhabited Wildland at 10,000 BCE, falling to 26.4% by 2017 CE.  Built on the HYDE 3.2 land-use reconstruction, 73 time steps from 10,000 BCE to 2017 CE.

Included because it complicates the word "pre-human" in every baseline in this group.  Holocene-integrated background rates — Wittmann's 11 kyr cosmogenic average, Reusser's ¹⁰Be rates — already contain some human signal, which biases them slightly high and implies an even lower true natural rate.  That is bad for the young-earth argument, and the article should say so rather than wait for a critic to.

It needs careful handling in the other direction too.  "Inhabited" is not "cleared and plowed."  Ellis's Intensive category requires at least 20% crop, pasture or urban cover, and that threshold was not crossed at scale until the last few centuries.  For an erosion calculation the relevant number is cropland area, not habitation: Wilkinson & McElroy put cropland at about 11% of global land today, cropland plus pasture at about 50 × 10⁶ km², and note that global cropland area has grown only about 11% since 1961 while population roughly doubled.

Did a bare or differently vegetated earth erode faster?

Evolution of alluvial mudrock forced by early land plants

McMahon, William J. and Davies, Neil S.  Science, 359(6379):1022-1024  (2018).  doi 10.1126/science.aan4660

topic: denudation-global, ancient-surfaces camp: mainstream type: compilation access: paywalled timescale: 3.5 Ga to 0.3 Ga direction: mixed

Key figures, from the abstract only: all 704 known alluvial formations from the Archean (3.5 Ga) to the Carboniferous (0.3 Ga).  Alluvium deposited after land plants evolved contains a proportion of mudrock "1.4 orders of magnitude greater," about 25×, than alluvium from the preceding 90% of earth history.  The onset is constrained to the Ordovician-Silurian.

The best evidence for the mechanism behind the "bare continents eroded faster" claim, and it stops short of the conclusion the claim needs.  Pre-vegetation continents retained almost no fine sediment — braided, sheet-flood systems that flushed material straight through.  That much is genuinely consistent with the young-earth intuition that a plantless earth moved sediment more efficiently.

But the measured quantity is retention and river style, not total denudation rate.  Plants both increased mud production, through soil formation and chemical weathering, and increased mud retention.  The authors attribute the shift to how vegetation revolutionised mud production and sediment flux from continental interiors — production and flux together.  No total-denudation multiplier can be read out of this paper in either direction.  Only the abstract was obtained; an open-access copy exists in the Hull repository and was not fetched.

Latitudinal effect of vegetation on erosion rates identified along western South America

Starke, Jessica; Ehlers, Todd A.; and Schaller, Mirjam.  Science, 367(6484):1358-1361  (2020).  doi 10.1126/science.aaz0840

topic: cosmogenic, denudation-global camp: mainstream type: measurement access: paywalled timescale: millennial, 10^3^-10^4^ yr direction: undercuts

Key figures, from the abstract only: 86 cosmogenic-radionuclide erosion rates along 3,500 km of the Andean Western Cordillera, from 6°S to 36°S, spanning an extreme climate and vegetation gradient, analysed with multivariate statistics.

The dataset that kills the simple model.  The result is "a bidirectional response to vegetation's influence on erosion whereby correlations between vegetation cover and erosion range from negative (dry, sparsely vegetated settings) to positive (wetter, more vegetated settings)."  In wet settings more vegetation goes with more erosion, because vegetation accelerates weathering and supplies erodible regolith.  A plantless earth would have less soil to erode, and bare bedrock resists detachment.  The net sign of removing plants is not obvious, and the best modern dataset says it flips with climate.

The classic statement of the same point is Langbein, W.B. and Schumm, S.A. (1958), "Yield of sediment in relation to mean annual precipitation," Eos, Transactions AGU, 39(6):1076-1084, doi 10.1029/TR039i006p01076, in which sediment yield peaks at intermediate semi-arid effective precipitation and falls at higher precipitation as vegetation thickens.  That paper was not obtained — Wiley returned HTTP 402 — so its peak-precipitation value must be verified before it is printed.  Only the abstract of Starke et al. was obtained.

Cenozoic expansion of grasslands and climatic cooling

Retallack, Gregory J.  The Journal of Geology, 109(4):407-426  (2001).  doi 10.1086/320791

topic: denudation-global camp: mainstream type: model access: paywalled timescale: 40 Myr direction: undercuts

Key claims, from the abstract only: grassland soils hold more organic matter than woodland and desert soils of comparable climate, and "when eroded, their crumb clods form sediment unusually rich in organic matter."  Grasslands "promote export of bicarbonate and nutrient cations to lakes and to the oceans," because they "preferentially exploit fertile young soils in the first flush of weathering and their soils have a crumb structure with much higher internal surface area for weathering."

Runs the opposite way to the argument that invokes it.  The young-earth move is that a pre-Miocene world without grasses eroded faster; Retallack's grasses raise chemical weathering rates, so their Neogene spread should have increased, not decreased, total denudation.

Two honest limits.  This is a hypothesis paper about climate, not a denudation-rate measurement, and Retallack's grassland chronology is contested by other workers.  Only the abstract was obtained.  Two standard citations for the parallel claim about rooted plants and chemical weathering — Berner, R.A. (1997), Science 276(5312):544-546, doi 10.1126/science.276.5312.544, and Gibling, M.R. and Davies, N.S. (2012), Nature Geoscience 5(2):99-105, doi 10.1038/ngeo1376 — could not be reached at all and are listed here unread.

Exhumation depths: how much rock demonstrably came off

Defining the Timing, Extent, and Conditions of Paleozoic Metamorphism in the Southern Appalachian Blue Ridge Terranes of Tennessee, North Carolina, and Northern Georgia

Thigpen, J.R. and others.  Tectonics  (2022).  doi 10.1029/2022TC007406

topic: uplift camp: mainstream type: measurement access: paywalled timescale: peak burial ~455 Ma to now direction: undercuts

Key figures: western Blue Ridge staurolite- and kyanite-grade units record 600-650 °C and 5.8-8.9 kbar; the central Blue Ridge Taconic metamorphic core records 775 °C and ~11.5 kbar; monazite U-Pb weighted means 450-441 Ma (west), 459-457 Ma (centre), 458-453 Ma (east).  At 2,750 kg/m³ and g = 9.81, one kbar is 3.7 km of rock, so those pressures are burial depths of 21.5 km, 33.0 km and 42.6 km.  Spread over the ~450 Myr since peak burial, the removal averages 47.8, 73.3 and 94.7 m/Myr (m/Myr = mm/kyr).

This is the single most important entry on this page for the deep-time half of the argument, because it does not involve an erosion rate at all.  Rocks now sitting in roadcuts in North Carolina, Tennessee and northern Georgia contain mineral assemblages that cannot form except under tens of kilometres of overburden.  Thermodynamic modelling of those assemblages gives the pressure; the pressure gives the depth; the overburden is gone.  The removal implied is not fast.  Portenga and Bierman's mean for basins in seismically inactive regions is 182 ± 30 m/Myr, so the long-term Appalachian average of 48-95 m/Myr is roughly half what the same class of terrain is doing today.

Two honest qualifications.  First, only the abstract was obtained — Wiley returns 403 and Unpaywall reports no open copy — so the pseudosection uncertainty bars on the 5.8-8.9 and 11.5 kbar figures are not in hand.  Second, the Barrovian metamorphism is Taconic (Ordovician), not Alleghanian, and the authors attribute the unroofing to "post-Taconic unroofing during Neoacadian and Alleghanian thrusting."  Part of that 20-43 km came off tectonically, as thrust sheets removed by later faulting, not purely by rainfall.  The correct statement is that 21 to 43 km of overburden was removed by a combination of tectonic unroofing and erosion — not that 43 km was eroded.

Ultrahigh-pressure metamorphism: tracing continental crust into the mantle

Chopin, C.  Earth and Planetary Science Letters, 212:1-14  (2003).

https://www.geology.cwu.edu/facstaff/tim/g501/pdf/Chopin2003EPSL.pdf

topic: crust-recycling camp: mainstream type: review access: OA timescale: 10^7^-10^8^ yr direction: undercuts

Key figures: coesite and metamorphic diamond indicate minimum pressures "on the order of 3 and 4 GPa, respectively" — burial to ~100 km and beyond 120 km; the Dora-Maira ultrahigh-pressure terrane is small, 15 × 5 × 1 km; a Hercynian metagranite from Brossasco sat at 700-750 °C at 100 km depth in Alpine times; some terranes imply present burial estimates of 150-200 km, and pyroxene exsolutions in garnet suggest pressures above 6 GPa, meaning more than 200 km; Indian crust beneath the Himalaya reached coesite-forming depths of 90-100 km with average exhumation near 10 mm/yr.

The clearest lay-facing evidence that the crust does not stay where it starts.  Silica has two structures — quartz below about 2.7-3 GPa, coesite above — and carbon has graphite below about 4 GPa and diamond above.  These are one-way switches, not calibrated models, so they carry none of the activity-model error that dogs multivariant barometers.  They survive the return trip because they are armoured: the host garnet or zircon acts as a pressure vessel and physically blocks the volume increase that back-transformation would need.  A piece of ordinary-looking granite that can be picked up at Brossasco was once 100 km down.

Read the limit carefully, because it is a trap for anyone arguing this side.  Chopin says the exhumation rates for ultrahigh-pressure terranes "are typical plate velocities (1-2 cm/yr)" and bear no relation to erosion rates.  This is buoyancy-driven return flow up a subduction channel, not denudation.  It proves deep burial and return; it does not prove that 100 km of rock was eroded away.  The type-locality discovery — Chopin, C. (1984), Contributions to Mineralogy and Petrology, 86(2):107-118, doi 10.1007/BF00381838 — was bibliographically verified through OpenAlex but the text was not obtained, so nothing is quoted from it here.

Ascent of the ultrahigh-pressure Western Gneiss Region, Norway

Hacker, B.R.  Geological Society of America Special Paper 419  (2007).

https://hacker.faculty.geol.ucsb.edu/viz/Hacker07_Ascent_Western_Gneiss_region.pdf

topic: crust-recycling camp: mainstream type: measurement access: OA timescale: 5 Myr ascent at 415-400 Ma direction: undercuts

Key figures: eclogite-facies conditions from ~2.0 GPa and 400 °C to ~3.5 GPa and 800 °C, which at a crust-plus-mantle column density of 3,000 kg/m³ is 68 to 119 km of burial; three discrete ultrahigh-pressure domains ranging from ~2,500 km² down to under 100 km²; a final amphibolite-facies overprint at 0.5 GPa and 700 °C, about 17 km; isothermal decompression to under 20 km depth in ~5 Myr, from 130 km.  That ascent is 22 km/Myr = 22,000 m/Myr = 22 mm/yr.

The extreme case, and it must be used with discipline.  A body of continental crust more than 20-30 km thick went to mantle depths and came back, and the mineral record of the trip is preserved.  That by itself destroys the intuition that rocks now at the surface have always been near the surface.  But 22 mm/yr is roughly four hundred times the fastest measured erosion rates anywhere on earth, which is the tell that this is a tectonic process rather than an erosional one.

The erosional part of the argument has to rest on the last twenty kilometres or so, and on Barrovian belts like the Appalachians, not on this.  A skeptic who knows the literature will catch an overclaim here immediately, and the credibility is worth more than the rhetorical gain.

Error propagation for barometers: 2. Application to rocks

Kohn, M.J. and Spear, F.S.  American Mineralogist, 76:138-147  (1991).

http://www.minsocam.org/ammin/AM76/AM76_138.pdf

topic: method camp: mainstream type: model access: OA timescale: n/a direction: mixed

Key figures: barometer accuracy as applied to real rocks "may typically range from ±600 to ±3250 bars (1σ or quarterwidth)"; converted to depth that is ±2.1 to ±11.4 km, narrowing to ±1.6 to ±6.8 km under a square-wave distributional assumption and ±1.5 to ±6.6 km under a Gaussian one; the authors summarise the overburden uncertainty as ±1.5 to ±11.4 km, or 5-30% of the thickness of normal continental crust; the kyanite-sillimanite transition limits pressure to about 3.75 kbar at 500 °C and 10 kbar at 800 °C; their worked example gives 6.4 kbar and 620 °C.

Included because it is the strictest published error analysis available, and because it is the number a hostile reader should be handed first rather than made to find.  A single barometer on a single sample is worth ±2 to ±11 km.  That is a large error bar and there is no honest way to shrink it.

Three things keep the conclusion standing anyway.  The pessimistic bar does not reach it: 42.6 ± 11 km is still 31 km of missing rock at the low end.  Independent phase equilibria clamp the answer, as the authors themselves point out — the presence of an aluminosilicate polymorph pins the result to part of the error field.  And polymorph indicators like coesite and diamond are not subject to activity-model error at all.  Note that a modern, more optimistic uncertainty figure was not obtained; Powell and Holland (2008), Journal of Metamorphic Geology, 26:155-179, is not open access and was not read, so the article's uncertainty statement rests on this older and more conservative source.

Metamorphic pressure variation in a coherent Alpine nappe challenges lithostatic pressure paradigm

Luisier, Baumgartner, Schmalholz, Siron and Vennemann.  Nature Communications, 10:4734  (2019).  doi 10.1038/s41467-019-12727-z

https://www.nature.com/articles/s41467-019-12727-z

topic: method camp: mainstream type: measurement access: OA timescale: n/a direction: mixed

Key figures: in the Monte Rosa nappe a whiteschist lens records ~2.2 GPa while its host metagranite records 1.05-1.40 GPa, with pressure rising toward the lens across about 30 m of rock; "A pressure difference of 0.8 ± 0.3 GPa indicates a significant deviation from the lithostatic pressure."

The strongest technical objection to the whole barometry argument, and it belongs on this page rather than in a footnote.  Converting pressure to depth assumes P = ρgz.  If rocks can carry non-lithostatic pressure, then recorded pressures overstate burial depth, and 0.8 GPa of tectonic overpressure is about 27 km of spurious depth at deep-crustal densities.  That is a real concession and there is no point pretending the debate does not exist.

Then notice what survives it.  The host metagranite in the same nappe still records 1.05-1.40 GPa, which is 40 to 50 km of real burial, and the authors' argument is that the whiteschist's 80-plus kilometres is the inflated figure.  The overpressure debate compresses the deepest ultrahigh-pressure numbers.  It does not turn 40 km into 4 km.

Ductile shearing to brittle thrusting along the Nepal Himalaya

Searle, M.; Avouac, J.-P.; Elliott, J.; and Dyck, B.  Tectonophysics, 714-715:117-124  (2017).

https://web.gps.caltech.edu/~avouac/publications/Searle_2017tecto_Ductle.pdf

topic: uplift camp: mainstream type: measurement access: OA timescale: 24-11 Ma direction: undercuts

Key figures: Greater Himalayan Sequence rocks "were formed by partial melting at 15-18 km depth >50-100 km north of the Himalaya and have been extruded southward"; Main Central Thrust and South Tibetan Detachment ductile shear zones active together in the Early Miocene at ~21-16 Ma (Langtang), 23.6 to ~13 Ma (Sikkim), and 13-11 Ma (Everest-Rongbuk).

An orogen doing now what the Appalachians did in the Ordovician, with the difference that the transport can still be watched.  The rocks at the top of the Greater Himalayan Sequence were partially molten mid-crustal material that has been squeezed southward and upward between two simultaneously active shear zones.

Weaker as fetched than the Appalachian case, and the entry says so.  The verified text number is only 15-18 km of burial for the melting.  The paper's Figure 5 plots two pressure-temperature paths, one in the kyanite field after Kohn (2008) and one in the sillimanite field after Inger and Harris (1993), but those are figure readings and no kilobar value is stated in the text, so none is quoted here.  A Greater Himalayan Sequence source with explicit peak pressures is still wanted.

Thermochronology, cratons, and why old orogens survive

Temporally and spatially uniform rates of erosion in the southern Appalachian Great Smoky Mountains

Matmon, A.; Bierman, P.R.; Larsen, J.; Southworth, S.; Pavich, M.; and Caffee, M.  Geology, 31(2):155-158  (2003).

topic: cosmogenic camp: mainstream type: measurement access: paywalled timescale: 10^4^-10^5^ yr direction: undercuts

Key figures: ¹⁰Be in 27 fluvial sediment samples from eight Great Smoky Mountain drainages of 1-330 km²; sediment generation of 73 ± 11 t km⁻² yr⁻¹, which the authors convert to 27 ± 4 m/Myr of bedrock erosion (the conversion reproduces exactly at a rock density of 2.7 t/m³); fission-track, long-term sediment-budget and sediment-yield data all give ~30 m/Myr for the Mesozoic and Cenozoic; Paleozoic orogenic unroofing rates were ≥100 m/Myr.

The single most important number for the "why are the Appalachians still there" question, and it is a measurement with four independent methods agreeing.  At 27 m/Myr the roughly 2,000 m of the Great Smokies takes 74 Myr to remove with no isostatic compensation at all, and something like 440 Myr with a sixfold Airy multiplier.  Nothing about a 300 Myr-old mountain range still standing needs a special explanation.

The comparison that matters is with an active orogen.  Taiwan erodes at 3-6 mm/yr, which is 3,000-6,000 m/Myr, so the Appalachians are going 110 to 220 times slower.  Measured ¹⁰Be says erosion rate spans two orders of magnitude between an active belt and a dead one, and the authors say why: rates fell "after termination of tectonically driven uplift," leaving a deep crustal root that isostatically maintains the range.  That is precisely the relief-dependence that Woodmorappe wondered aloud about in 2024 and that Schumm fitted a curve to in 1963.  Only the abstract was obtained, through OpenAlex; GeoScienceWorld blocks the full text.

Colorado Plateau uplift and erosion evaluated using GIS

Pederson, J.L.; Mackley, R.D.; and Eddleman, J.L.  GSA Today, 12(8):4-10  (2002).

https://www.geosociety.org/gsatoday/archive/12/8/pdf/i1052-5173-12-8-4.pdf

topic: uplift, isostasy camp: mainstream type: measurement access: OA timescale: 70 Myr and 30 Myr direction: mixed

Key figures: mean Cenozoic rock uplift since the Late Cretaceous 2,117 m, from 90 structural-contour data points on uppermost Cretaceous strata; mean net erosional exhumation over the same interval 406 m, which is 5.8 m/Myr; mean post-Laramide erosion since about 30 Ma 843 m, which is 28.1 m/Myr; maximum post-Laramide erosion under 2,000 m, in the Canyonlands region and along the axis of Grand Canyon; isostatic rebound from that 843 m of unloading, 639 m, using crust 2,500 and mantle 3,300 kg/m³, leaving 204 m of net surface lowering; about 1,800 m of the rock uplift needs a source other than erosional unloading.

This is the measured answer to the creationist Grand Canyon talking point.  Writers in that tradition assert that several kilometres of Mesozoic section were stripped from above the rim, and treat the implied rate as evidence for catastrophic removal.  The measured maximum along the canyon axis is under 2,000 m and the plateau-wide mean is 843 m since 30 Ma.  The scale of removal is real and roughly what the Grand Staircase implies.  The rate, 28 m/Myr, is above the global outcrop mean of 12 m/Myr and well below the global drainage-basin median of 54 m/Myr.  Nothing about it is anomalous.

The plateau also gained section in the Paleocene and Eocene, which is why net Cenozoic exhumation (406 m) is smaller than post-30 Ma erosion (843 m).  Any argument that treats erosion as monotonic gets this wrong.  The authors state their own weakness plainly and it should be quoted rather than buried: the post-Laramide reconstruction is "a work in progress and somewhat subjective," they take a conservative approach, and the result is a minimum estimate.  One caution for anyone reusing this paper: it reports mean thicknesses only and gives no eroded volume.  Converting to a volume requires a plateau area, which the paper does not state, so a volume figure quoted from this source is a derivation and should be labelled as one.

Unroofing, incision, and uplift history of the southwestern Colorado Plateau from apatite (U-Th)/He thermochronometry

Flowers, R.M.; Wernicke, B.P.; and Farley, K.A.  GSA Bulletin, 120(5/6):571-587  (2008).  doi 10.1130/B26231.1

http://tecto.gps.caltech.edu/publications/pdf/Flowers_etal2008GSAB.pdf

topic: thermochronology camp: mainstream type: measurement access: OA timescale: 10^8^ yr direction: mixed

Key figures: 230 (U-Th)/He analyses on 36 samples, with an assumed geothermal gradient of 25 °C/km; pre-mid-Cretaceous stripping of ~1,500 m of Lower Triassic through Upper Jurassic section in ~56 Myr, stated as 0.03 mm/yr and recomputing to 26.8 m/Myr; aggradation of at least 1,500 m of Upper Cretaceous sediment along the plateau margin; the Kaibab surface cooled below 45 °C before 60 Ma in western Grand Canyon and 60-35 Ma along the Mogollon Rim; plateau-interior unroofing 28-16 Ma; mid-Tertiary relief of 850-1,400 m at Salt River Canyon and ~600 m at the Mogollon Rim; net structural relief over 5,500 m developed; the plateau reached its present ~1.9 km elevation "with little internal upper crustal strain (<1%)."

The independent cross-check on geobarometry, and it works on different physics.  Barometry reads the deepest point of the trip from mineral equilibria and volumes.  Low-temperature thermochronology reads the last few kilometres from helium that leaks out of apatite above roughly 70 °C and is retained below it.  The two methods share no assumptions.

The entry is tagged mixed because the paper cuts both ways honestly.  It confirms kilometre-scale rather than tens-of-kilometres-scale Mesozoic and Cenozoic removal on the plateau, at rates of tens of m/Myr.  And it supplies the single most useful cratonic datum in the whole file: like most of the North American craton, the Colorado Plateau "remained near sea level for 500 m.y." during slow subsidence and deposition.  That is a stratigraphy-and-thermochronology statement from a mainstream journal with no stake in this debate, and it is why flat-lying Paleozoic rock is still there.  The authors also caution against the obvious misuse of their own result, warning readers not to presume that unroofing is a direct proxy for surface uplift.

Miocene rejuvenation of topographic relief in the southern Appalachians

Gallen, S.F.; Wegmann, K.W.; and Bohnenstiehl, D.R.  GSA Today, 23(2):4-10  (2013).  doi 10.1130/GSATG163A.1

https://www.geosociety.org/gsatoday/archive/23/2/pdf/i1052-5173-23-2-4.pdf

topic: uplift camp: mainstream type: measurement access: OA timescale: 10^7^ yr direction: mixed

Key figures: knickpoint analysis on a 6 m LiDAR digital elevation model of the Cullasaja River basin, about 300 km² in western North Carolina; relief has increased by more than 150% since the Miocene.

Tagged mixed because the rejuvenation literature concedes the creationist observation before answering it.  These papers exist because the southern Appalachians look too rugged for a range that has had no tectonic forcing for over 200 Myr — locally high relief, steep slopes, incised gorges, frequent mass wasting.  That is the honest starting point and the article should not hide it.

The mainstream resolution is that the relief is young even though the orogen is old, driven by post-orogenic epeirogenic uplift in the Miocene.  That is the surface-uplift versus rock-uplift distinction again.  Note that the field has not settled it: the competing view is Hack's dynamic equilibrium, in which relief is controlled by variable rock erodibility and has long persisted, and the authors state that thermochronology and cosmogenic radionuclide results "have not led to a consensus."  One citation correction worth recording: the third author is D.R. Bohnenstiehl, not McKeon, and the paper is sometimes cited wrongly.  Hack (1960, 1982) and Pazzaglia and Brandon (1996) are cited inside this paper but were not fetched, so no full citations for them are given here.

Why are the Appalachians high? New insights from detrital apatite laser ablation (U-Th-Sm)/He dating

Jess, S.; Enkelmann, E.; and Matthews, W.  Earth and Planetary Science Letters, 597:117794  (2022).  doi 10.1016/j.epsl.2022.117794

topic: thermochronology camp: mainstream type: measurement access: paywalled timescale: 10^7^-10^8^ yr direction: mixed

Key figure, unverified: apatite fission-track and (U-Th-Sm)/He dates across the Appalachians are consistently Mesozoic, implying that late Cenozoic exhumation has likely failed to exceed about 1.2 km.

Listed with an explicit warning.  The figure above comes from a web-search snippet, not from the paper.  Unpaywall reports the article as open access, but the only location it gives is the Elsevier landing page, which returns 403 to every fetcher tried.  Treat the 1.2 km as provisional until someone reads the paper.

It is included anyway because it cuts against this article's own side of the ledger in one respect and for its own side in another, and both should be visible.  It limits recent Appalachian erosion to roughly a kilometre in the Cenozoic, which shows the field is not simply assuming fast steady erosion everywhere.  It also coexists with the 21-43 km of Paleozoic overburden without contradiction, because that removal was front-loaded into the interval when there was still a mountain belt and a thrust pile to take off.  What was not found, and should be flagged as a real gap, is any single paper that closes the loop quantitatively from peak metamorphic depth through cooling ages to the offshore sediment volume.  Pazzaglia and Brandon (1996), Basin Research, 8:255-278, did the offshore mass-balance half and was not obtained.

Uplift rates, Everest, and orogenic mass balance

Recent uplift of Chomolungma enhanced by river drainage piracy

Han, X.; Dai, J.-G.; Smith, A.G.G.; Xu, S.; Liu, B.-R.; Wang, C.; and Fox, M.  Nature Geoscience, 17:1031-1037  (2024).  doi 10.1038/s41561-024-01535-w

https://www.nature.com/articles/s41561-024-01535-w

topic: uplift, isostasy camp: mainstream type: model access: OA timescale: 89 kyr direction: mixed

Key figures: Everest's short-term GPS rock uplift is ~2 mm/yr = 2,000 m/Myr, its long-term thermochronological rock uplift ~1 mm/yr = 1,000 m/Myr; isostatic response to a river capture adds 0.16-0.53 mm/yr at Everest and 0.18-0.77 mm/yr at Makalu, which is 10-50% of the GPS rate; best-fit capture age 89 ka, with well-fitting models spanning 50-100 ka and independent optically stimulated luminescence ages on upstream pluvial sediment of 30-90 ka; anomalous elevation attributed to the mechanism 15-50 m of Everest's 8,849 m, comprising 13-41 m at Everest and 14-60 m at Makalu; almost 800 m of non-steady incision through the highest Himalaya since capture; effective elastic thickness explored at 10, 20 and 30 km; background rock uplift imposed in the flexural model at 4 mm/yr in the mid-crustal duplex, 0.5 mm/yr to the south, 1 mm/yr to the north.

The most interesting source in this group, because a 2024 Nature Geoscience paper states the creationist observation in plain language before resolving it.  The authors write of "a disconnect between the short-term rock uplift rate of Chomolungma measured from GPS data (~2 mm yr⁻¹) and its long-term rock uplift rate recorded by thermochronology (~1 mm yr⁻¹)."  The measured modern rate really is about twice the long-term rate.  Anyone answering the erosion argument should concede that first and explain it second.

The explanation is a dated, modelled, locally sourced transient: the Arun river captured the Kosi drainage roughly 89,000 years ago, the resulting incision unloaded the crust, and the flexural rebound is still going.  It is not a claim that the rate has always been 2 mm/yr.  The authors state their isostatic figures are minimum estimates because hillslope erosion is not modelled.  Weakness to record: the ~2 mm/yr Everest GPS value is cited at one remove from Wang and Barbot (2023), and the station-level value was not located in that paper.

The Miocene elevation of Mount Everest

Gebelin, A.; Mulch, A.; Teyssier, C.; Jessup, M.J.; Law, R.D.; and Brunel, M.  Geology, 41(7):799-802  (2013).  doi 10.1130/G34331.1

https://pearl.plymouth.ac.uk/gees-research/1320/

topic: uplift camp: mainstream type: measurement access: paywalled timescale: ~15 Myr direction: undercuts

Key figures: stable-isotope paleoaltimetry on hydrous minerals deformed in the South Tibetan detachment shear zone gives δD of the exchanging meteoric water of −156 ± 5 ‰, against age-equivalent near-sea-level foreland paleosol records at δ¹⁸O of −5.8 ± 1.0 ‰; the difference is consistent with mean elevations of at least 5,000 m for the Mount Everest area in the late Early Miocene, described as similar to modern.

The cleanest empirical refutation of the whole extrapolation argument, and it needs no rate model at all.  The argument runs: measured uplift rate multiplied by deep time gives an impossible mountain.  Test it.  If Everest had been rising at even 1 mm/yr for 15 Myr it would have gained 15 km of elevation.  It stands at 8,849 m, and this paper says it was already at 5,000 m or more when the South Tibetan detachment was active.  Surface uplift over that interval is close to nil.

That is the England and Molnar distinction observed in the field rather than argued in the abstract.  Rock uplift is real and continuous; surface uplift has been near zero for millions of years because erosion has kept pace with it.  Two limitations, stated because a reader will find them.  Only the repository abstract was obtained, and it is truncated; GeoScienceWorld returns 403.  And the age is given as "late Early Miocene" with no figure in millions of years, so no Ma value is asserted here.

Surface uplift, uplift of rocks, and exhumation of rocks

England, P. and Molnar, P.  Geology, 18(12):1173-1177  (1990).

https://www.osti.gov/biblio/5798508

topic: uplift, method camp: mainstream type: review access: paywalled timescale: n/a direction: undercuts

The paper that names the confusion the young-earth uplift argument runs on, published in 1990, aimed at professional geologists, and long predating the creationist use of the numbers.  Three different quantities get called "uplift": how fast the land surface rises, how fast a given rock rises relative to the geoid, and how much rock is removed from above a sample.  They are not the same, and a mountain in steady state can have large rock uplift and zero surface uplift.

The abstract is blunt about how often the field itself got this wrong: many reports of surface uplift in mountain ranges mistake exhumation or rock uplift for surface uplift, and "provide no information whatsoever on the rates of surface uplift."  This is the best kind of hostile-witness citation, because the target is mainstream geologists and not creationists.

Only the OSTI abstract was obtained.  Unpaywall reports the article closed with no repository copy and GeoScienceWorld returns 403.  The relation surface uplift = rock uplift − exhumation is universally attributed to this paper, but it was not read in the source, so it should be stated in the article's own words and this paper cited for the distinction, which the abstract does establish verbatim.

Interseismic coupling on the main Himalayan thrust

Stevens, V.L. and Avouac, J.-P.  Geophysical Research Letters, 42(14):5828-5837  (2015).  doi 10.1002/2015GL064845

https://authors.library.caltech.edu/records/7qgcw-dkb40/files/grl53219.pdf

topic: uplift camp: mainstream type: measurement access: OA timescale: interseismic, decades direction: undercuts

Key figures: arc-normal convergence increases eastward from 13.3 ± 1.7 to 21.2 ± 2.0 mm/yr; the Main Himalayan Thrust is fully locked over about 100 ± 20 km width along its whole length; moment deficit 15.1 ± 1 × 10¹⁹ N m/yr for the arc; modelled interseismic uplift correlates with topography at R = 0.78; the fraction of anelastic, permanent deformation is up to 10% at the 70% confidence level, and the probability that anelastic convergence exceeds 2 mm/yr is only 30%; the Himalaya accommodates about half of the ~40 mm/yr India-Eurasia convergence.

The measurement that dismantles the geodetic step in the argument.  A GPS or levelling uplift rate is not a rock uplift rate.  Most of the interseismic signal is elastic strain stored in the hanging wall of a locked thrust, and earthquakes hand it back.  The authors do not assume this; they test it by comparing geodetic rates against independent geological and geomorphic slip rates and finding them equal, which is only possible if the geodetic signal is dominated by recoverable strain.

They are honest about the limit, and the entry repeats it: because interseismic uplift rates correlate with topography, the anelastic deformation cannot be null.  Some of it is permanent.  It is roughly a tenth, not the whole.  This is also the direct answer to the highest vertical numbers in circulation, the ~6-7 mm/yr interseismic levelling signal for the High Himalaya usually traced to Jackson and Bilham (1994), Journal of Geophysical Research, 99(B7):13,897-13,912 — a paper whose reference was verified inside Ader et al. (2012) but which was not itself fetched, so no number is quoted from it here.

Convergence rate across the Nepal Himalaya and interseismic coupling on the Main Himalayan Thrust

Ader, T.; Avouac, J.-P.; Liu-Zeng, J.; and others.  Journal of Geophysical Research, 117:B04403  (2012).  doi 10.1029/2011JB009071

https://authors.library.caltech.edu/records/tavqm-6yn80/files/Ader2012p18083J_Geophys_Res-Sol_Ea.pdf

topic: uplift camp: mainstream type: measurement access: OA timescale: decades vs 10^4^ yr direction: undercuts

Key figures: 30 continuous GPS stations in Nepal and southern Tibet plus campaign GPS plus the 1977-1990 Survey of Nepal spirit-levelling line along the Birganj-Kathmandu-Kodari road; geodetic shortening 17.8 ± 0.5 to 20.5 ± 1 mm/yr; independent Holocene geologic shortening from deformed terraces 21.5 ± 1.5 mm/yr in central Nepal (Lavé and Avouac 2000) and 19 ± 6 mm/yr in western Nepal (Mugnier et al. 2003).

The cross-check that makes the elastic-strain result testable rather than merely plausible.  A rate measured over two decades by satellite geodesy and a rate measured over ten thousand years from deformed river terraces agree inside their uncertainties.  That agreement is exactly what should not happen if the modern instrumental signal were permanent shortening running far faster than the long-run average.

Useful methodological note for anyone reusing the vertical numbers: the authors state that the most reliable vertical velocities available are the levelling data, and that the GPS vertical velocities carry large uncertainties.  The levelling line here is the same dataset behind the higher published Himalayan uplift figures.

Links between erosion, runoff variability and seismicity in the Taiwan orogen

Dadson, S.J.; Hovius, N.; Chen, H.; Dade, W.B.; Hsieh, M.-L.; Willett, S.D.; Hu, J.-C.; Horng, M.-J.; Chen, M.-C.; Stark, C.P.; Lague, D.; and Lin, J.-C.  Nature, 426(6967):648-651  (2003).  doi 10.1038/nature02150

topic: uplift, timescale-bias camp: mainstream type: measurement access: paywalled timescale: 10^1^, 10^4^ and 10^6^ yr direction: undercuts

Key figures: erosion rates in the Taiwan mountains are 3-6 mm/yr = 3,000-6,000 m/Myr on all three timescales — modern river sediment loads over decades, Holocene river incision over thousands of years, and thermochronometry over millions — although the spatial pattern of erosion has shifted with the migration of localised deformation.

The direct empirical rebuttal to "if erosion were this fast the mountains would already be gone."  Taiwan erodes several times faster than the average modern orogenic uplift rate Schumm calculated in 1963, has done so for millions of years, and is still a mountain range.  Fast erosion in an active orogen is not a paradox.  It is a steady state, maintained because uplift feeds rock into the erosion zone as fast as the rivers and landslides take it out.

It is also a second, independent instance of the Kirchner result: three chronometers with nothing in common, spread over five orders of magnitude in integration time, returning the same number.  Only the PubMed abstract was obtained; the body of the paper was not read.

The Himalaya-Bengal Fan denudation-accumulation system during the past 20 Ma

Einsele, G.; Ratschbacher, L.; and Wetzel, A.  The Journal of Geology, 104(2):163-184  (1996).  doi 10.1086/629812

topic: sediment-budget camp: mainstream type: measurement access: paywalled timescale: 20 Myr direction: undercuts

Key figures: Himalayan denudation of 7.1 × 10⁶ km³ against sediment accumulation in the Subhimalayan basins and Bengal deep-sea fan of 7.4 × 10⁶ km³ over the past 20 Myr, both at ±20% and a rock density of 2.75 g/cm³ — the two halves of the budget agree within their stated error; vertical unroofing along the High Himalayan crystalline chain averaged ~1,000 m/Myr; northward lateral retreat of the monsoon-facing southern slope up to 3.5 km/Myr; the Tethyan sedimentary zone to the north averaged only 150 m/Myr; modern river solute loads give a mean chemical denudation rate of 17 m/Myr.

The mass balance that tests the Himalayan version of the argument directly, and it closes.  Rock removed from the range and sediment sitting in the basins downstream are the same quantity to within 4%, over 20 million years.

It is also the check on a specific claim.  Woodmorappe's 2024 review reports Fleming citing "a uniformitarian estimate" that long-term Himalayan erosion was only 2% of present rates.  Running the numbers: 7.4 × 10⁶ km³ over 20 Myr at 2.75 t/m³ is 1.02 Gt/yr, against a modern Ganges plus Brahmaputra suspended load of about 1.06 Gt/yr, a ratio of 0.96.  The discrepancy is a factor of about one, not a factor of fifty.  Three cautions on that arithmetic, all of which should be printed with it: only the OpenAlex abstract of Einsele et al. was obtained, not the body; the modern load figure comes from Milliman and Syvitski (1992) via a search summary and was not fetched; and the two totals do not cover identical source areas, since the accumulation figure includes peninsular Indian input and excludes material lost to the Nicobar fan.  No reasonable handling of those caveats moves a ratio near unity to 0.02.

On steady states in mountain belts

Willett, S.D. and Brandon, M.T.  Geology, 30(2):175-178  (2002).

topic: uplift camp: mainstream type: model access: paywalled timescale: 10^6^-10^7^ yr direction: undercuts

Key figures: four distinct kinds of steady state are identified — flux, topographic, thermal and exhumational; exhumational steady state shows up as reset cooling-age zones nested by closure temperature, illustrated from New Zealand, the Cascadia accretionary margin and Taiwan.

The conceptual frame behind the Taiwan measurement, and the reason "uplift rate × time = impossible mountain" fails as a syllogism.  Tectonics and erosion are coupled with feedbacks, so an orogen tends toward a balance in which rock enters and leaves at similar rates and the mean topography stops changing.  Rock uplift can continue indefinitely without the summit getting higher.

Do not overclaim it, and the authors say why: models suggest a steady mean form at the scale of a whole orogenic belt, but "perfect topographic steady state is unlikely to be achieved at shorter length scales."  It is a mean-form, orogen-scale statement, not a claim that any particular peak is fixed.  Abstract only, reconstructed from the OpenAlex record; GeoScienceWorld returns 403 and there is no open copy.

The Ordovician rocks of Mount Everest

International Union of Geological Sciences.  IUGS Geological Heritage Sites  (undated; fetched July 2026).

https://iugs-geoheritage.org/geoheritage_sites/the-ordovician-rocks-of-the-mount-everest/

topic: uplift camp: mainstream type: popular access: OA timescale: 450 Myr direction: background

Key figures: the summit succession is Middle Ordovician, Darriwilian, about 450 Ma; the fauna includes brachiopods, conodonts and crinoids, with shelly assemblages dominated by suspension feeders including orthide and strophomenide brachiopods.

The summit of the highest mountain on earth is marine limestone, "deposited in a warm, shallow-water sea some 450 million years ago."  Both sides accept the observation and both sides like it.  Creationists read it as flood deposition followed by rapid uplift; mainstream geology reads it as a 450 Myr-old shallow-marine platform carried 8.8 km up by continental collision.  Tagged background for that reason, tilting toward undercutting, because the fossils themselves are the part a skeptic will press.  Conodonts and orthide and strophomenide brachiopods are stratigraphically diagnostic of the Ordovician and do not occur higher in the column, so the limestone is not neutral evidence and the article should not present it as such.

One naming caution.  This page calls the summit succession the Chiatsun (or Jiacun) Group; Sakai et al. (2005) and Myrow et al. use Qomolangma Formation.  They are the Chinese-literature and international-literature names for the same rocks, and the article should give both rather than silently pick one.  Sakai, H. and others (2005), Island Arc, 14(4):297-310, doi 10.1111/j.1440-1738.2005.00499.x, was reachable only as a title and author list, so nothing is quoted from it.

Topographic controls on erosion rates in tectonically active mountain ranges

Montgomery, D.R. and Brandon, M.T.  Earth and Planetary Science Letters, 201(3-4):481-489  (2002).  doi 10.1016/S0012-821X(02)00725-2

topic: uplift camp: mainstream type: measurement access: paywalled timescale: 10^3^-10^6^ yr direction: undercuts

Listed with the text not obtained, and no figure from it is printed here.  OpenAlex and Semantic Scholar both return a null abstract, there is no open copy, and Elsevier was not reached.

It is included because it is the formal modern answer to the objection Woodmorappe raised against his own side in 2024 — whether erosion rates depend on topography — and because its proposition is attested at one remove by a paper that was read in full.  Gallen et al. (2013) open with the statement that topographic relief exerts an essential control on the rates and processes of landscape denudation, citing Ahnert (1970) and this paper.  That is enough to cite it for the proposition and not for a number.  It is the successor to Schumm's Figure 2, and the proposition is what dissolves the continent-levelling extrapolation: the rate is not constant as relief falls, so dividing a mean elevation by a present-day rate does not give a time.

Origin of Sedimentary Rocks, 2nd edition

Blatt, H.; Middleton, G.V.; and Murray, R.C.  Prentice-Hall, Englewood Cliffs, New Jersey, 634 pp.  (1980).  ISBN 0136427022

topic: uplift camp: mainstream type: book access: book timescale: n/a direction: mixed

Claim in circulation, NOT VERIFIED: that page 30 states a mountain sustaining 10 mm/yr of uplift would reach about 45 km in height.

Recorded here as an open failure rather than dropped, because the figure is in circulation and someone will eventually ask where it comes from.  The book exists and the second edition is correctly cited above.  The Internet Archive holds a scan at originofsediment0000blat_a0l5_2, but the item is lending-restricted, the legacy search-inside endpoint is retired, the bulk full-text API returned empty, and the BookReader search endpoint 404s.  Nobody has read page 30.

Three reasons for caution beyond the non-verification.  Schumm's independently computed crossover, where maximum denudation rates catch up with his average uplift rate of 25 ft per 1,000 years, is at relief above 20,000 feet — about 6.1 km, not 45 km, a discrepancy of 7.5× between two textbook-level statements.  A sustained 10 mm/yr uplift premise is itself the error at issue, since roughly 90% of a measured geodetic rate is elastic.  And if Blatt, Middleton and Murray wrote the figure as a reductio showing that 10 mm/yr cannot be a long-term rate, then quoting it as a problem for uniformitarianism inverts the authors' meaning — the same failure mode as the standard misuse of Schumm.  Do not print the 45 km figure until the page is read.

The ocean sediment budget

Mass/age distribution and composition of sediments on the ocean floor and the global rate of sediment subduction

Hay, W.W.; Sloan, J.L. II; and Wold, C.N.  Journal of Geophysical Research: Solid Earth, 93(B12):14933-14940  (1988).  doi 10.1029/JB093iB12p14933

topic: sediment-budget camp: mainstream type: compilation access: paywalled timescale: 0-180 Myr of sea floor direction: background

Key figures: total ocean-floor sediment mass 262 × 10²¹ g = 2.62 × 10¹⁷ t; pelagic 140 × 10²¹ g of which about 74% is calcium carbonate; detritus 122 × 10²¹ g, of which about 6 × 10²¹ g is volcanic; mass/age distribution fitted as (11.02 × 10²¹ g)·e^−0.0355t per Ma, which is a half-mass age of 19.5 Myr and an e-folding time of 28.2 Myr; subduction removal "about 1 × 10²¹ g per million years" = 1 Gt/yr.  Reservoir divided by that flux gives a residence time of 262 Myr.

The single paper both sides build on, and the reason to read it first is that the young-earth argument and its answer both come out of the same three numbers.  Snelling's "not even 1,300 feet (400 m)" is Hay's deep-sea thickness.  The 1 Gt/yr removal term that every young-earth version divides by is Hay's, and Hay states it as a minimum — Stern's 2002 review gives the range as 1 × 10¹⁵ to 3-4 × 10¹⁵ g/yr.  Two further cautions belong here before anyone quotes the paper.  Hay's 262 × 10²¹ g is ocean-floor sediment and deliberately excludes the continental shelves, so it is not comparable to a whole-ocean figure.  And his half-mass age of 19.5 Myr, derived from mass, sits 1.5× below the ~30 Myr volumetric half-life Husson and Peters later report; that gap is a real uncertainty and should not be hidden by quoting whichever suits.

Only the abstract was obtained.  Wiley paywalls the full text and Unpaywall reports no open copy, so every figure above is from the abstract as stored by OpenAlex.  One cross-check was run: 2.62 × 10¹⁷ t at a dry bulk density near 2.0 t/m³ implies 1.31 × 10⁸ km³, against GlobSed's independently mapped deep-sea volume of 1.13 × 10⁸ km³ — two methods thirty years apart agreeing within 16%.

GlobSed: updated total sediment thickness in the world's oceans

Straume, E.O.; Gaina, C.; Medvedev, S.; and others.  Geochemistry, Geophysics, Geosystems, 20(4):1756-1772  (2019).  doi 10.1029/2018GC008115

topic: sediment-budget camp: mainstream type: compilation access: paywalled timescale: 0-180 Myr of sea floor direction: undercuts

Key figures: a 5-arc-minute global grid.  All ocean and marginal seas: 3.63 × 10⁸ km² carrying 3.37 × 10⁸ km³ of sediment, mean thickness 927 m.  Continental margins: 0.469 × 10⁸ km², 1.43 × 10⁸ km³, mean 3,044 m.  Deep sea: 2.79 × 10⁸ km², 1.13 × 10⁸ km³, mean 404 m.  The grid is "an increase of 29.7%, compared to previously published global maps."  Deep-sea thickness against the 64.2 Myr mean age of the sea floor is 6.3 m/Myr (= 6.3 mm/kyr); the whole-ocean 927 m over the same interval gives 14.4 m/Myr, which overstates, because margin sediment sits on stretched continental crust far older than the mean sea-floor age and is never subducted.

This is the measurement that shows what the "only 400 metres of sediment" claim is actually a measurement of.  The 400 m every young-earth source quotes is the deep-sea subset — 2.3 times thinner than the whole-ocean mean of 927 m.  A reader who has met the argument in its usual form has been shown the thinnest of the three numbers and invited to compare it against the age of the whole ocean.

Weakness of this entry, stated plainly: the primary paper was not obtained.  Wiley returns 402 and 403 to every route tried; the Oslo handle resolves to a metadata-only record; the figshare and GEUS records carry no PDF.  The table above is quoted second-hand from the Creation Research Society article catalogued below, which quotes Straume directly.  It was checked for internal consistency — 0.927 km × 3.63 × 10⁸ km² = 3.365 × 10⁸ km³, matching the stated total to 0.2%, and the margin and deep-sea rows sum correctly — but it should be labelled second-hand until the PDF is in hand.

Observations at convergent margins concerning sediment subduction, subduction erosion, and the growth of continental crust

von Huene, R. and Scholl, D.W.  Reviews of Geophysics, 29(3):279-316  (1991).  doi 10.1029/91RG00969

topic: crust-recycling camp: mainstream type: review access: paywalled timescale: contemporary and 30 Myr averages direction: undercuts

Key figures: "the contemporary rate of solid-volume sediment subduction at convergent ocean margins (∼43,500 km) is calculated to be 1.5 km³/yr" = 4.05 Gt/yr at 2.7 t/m³; the long-term rate, corrected for high terrigenous sedimentation over the past 30 Myr, is 1.0 km³/yr = 2.7 Gt/yr; total subcrustally subducted material 1.3-1.8 km³/yr = 3.5-4.9 Gt/yr; material recycled to the mantle "at a rate near 1.6 km³/yr" = 4.3 Gt/yr, which the authors note is "effectively equivalent to the commonly estimated rate at which the mantle adds juvenile igneous material."  Subduction erosion of the forearc runs at 25-50 km³/Myr per kilometre of margin, conservatively 30, giving about 1.1 km³/yr globally.  Interstitial water expelled: 0.9 km³/yr.

The removal term the argument needs and rarely gets right.  It is worth being fair to Stuart Nevins here: his 1973 figure of 2.75 Gt/yr destroyed by sea-floor spreading, which critics routinely call a lowball, matches von Huene and Scholl's long-term rate of 2.7 Gt/yr to within 2%.  What Nevins missed is the contemporary rate, half again as large, and subduction erosion, which he had no term for at all.  His "a tenth of the input" claim only survives because his input of 27.5 Gt/yr counted dissolved load, wind dust and salt spray, none of which is sea-floor sediment.  Against the modern particulate flux of 17.8 Gt/yr the contemporary subduction flux is 23% of input, not 10%.

Abstract only; Wiley paywalls the text and no open copy exists.  Every figure above is verbatim from the abstract.

Controls on tectonic accretion versus erosion in subduction zones

Clift, P. and Vannucchi, P.  Reviews of Geophysics, 42:RG2001  (2004).  doi 10.1029/2003RG000127

https://app.ingemmet.gob.pe/biblioteca/pdf/RG000127.pdf

topic: crust-recycling camp: mainstream type: review access: OA timescale: 10^6^-10^8^ yr direction: undercuts

Key figures: "57% of the global active margins seem to be of the erosive type"; "Only 7 - 37% of the sediment reaching the trenches appears to be added to the accretionary complexes, with the bulk subducted to depth," median accreted fraction 17%; "Even at the accretionary margins, typically 83% of the incoming sedimentary pile is subducted beneath the forearc wedge"; eroded forearc basement is "approximately 55% of the crustal material subducted to depth"; tectonic erosion is favoured where convergence exceeds 6 ± 0.1 cm/yr and sediment cover is under 1 km; maintaining the present crustal volume requires magmatic productivity of 90 km³/Myr per kilometre of active margin; continental crust volume taken as 8.0 × 10⁹ km³.  Table 1 gives per-margin sediment delivery of 43-179 km³/Myr and accretion of 4-29 km³/Myr.

The direct answer to "sediment cannot really be destroyed, it just piles up somewhere."  Destruction is measured margin by margin, from seismic sections and drilling, and the dominant fate of trench sediment is subduction rather than accretion.  Half the world's active margins are not merely failing to accrete; they are losing their own forearc basement into the trench.  This is the only entry in the group that quantifies the removal term from the geology of individual margins rather than from a global average, and full text was read.

Estimates of sediment flux to global oceans

Cohen and others.  Water Resources Research, 58:e2021WR031583  (2022).  doi 10.1029/2021WR031583

topic: sediment-budget camp: mainstream type: model access: paywalled timescale: modern annual direction: background

Key figures: "Total global particulate load of 17.8 Gt/y is delivered to global oceans, 14.8 Gt/y as washload, 1.1 Gt/y as bedload, and 2.6 Gt/y as suspended bed material."  "The largest 25 rivers are predicted to transport more than half of the total sediment flux to global oceans."

The best modern number for the particulate flux, and the one to use instead of the 1992 figure when a present-day rate is wanted.  It matters most for the concentration result: half the world's sediment delivery comes down twenty-five rivers, so any calculation that spreads a global flux evenly across a global land area is describing a landscape that does not exist.  Abstract only.

The oceans point to a young earth

Tomkins, J.P. and Clarey, T.  Acts & Facts, Institute for Creation Research  (2021).

https://www.icr.org/article/oceans-young-earth

topic: sediment-budget camp: creationist type: popular access: OA timescale: n/a direction: supports

Key figures: "about 22 billion tons of sediment disappear into the oceans every year"; after removal, the "remaining 21 billion tons would have accumulated on the seafloor in less than 12 million years"; framed against oceans "over 200 million years old," for a shortfall of about 17×.

The strongest modern young-earth statement of the ocean-sediment argument, and it is stronger than the others precisely because it uses 200 Myr rather than the 1 to 3 billion years its predecessors used.  Credit where it is due: that is the right order of magnitude for the oldest sea floor.  It is still the wrong denominator.  The reservoir is being drained continuously, so the quantity to compare against is the area-weighted mean age of the floor, 64.2 Myr, not its maximum.  Three corrections close the gap on their own terms: use GlobSed's whole-ocean 927 m rather than the deep-sea 404 m, use a measured subduction flux of 4.05 Gt/yr rather than Hay's stated minimum of 1 Gt/yr, and use the mean age rather than the maximum.

Very little sediment on the seafloor

Snelling, A.A.  Answers in Genesis  (2012).

https://answersingenesis.org/geology/sedimentation/1-very-little-sediment-on-the-seafloor/

topic: sediment-budget camp: creationist type: popular access: OA timescale: n/a direction: supports

Key figures: "Every year water and wind erode about 20 billion tons of dirt and rock debris from the continents and deposit them on the seafloor"; "the average thickness of all these sediments globally over the whole seafloor is not even 1,300 feet (400 m)"; "An estimated 1 billion tons of sediments are removed this way each year"; "At this rate, 1,300 feet of sediment would accumulate in less than 12 million years."  The internal rate implied is 400 m in 12 Myr = 33 m/Myr, five times the 6.3 m/Myr that GlobSed and Seton together actually require.

The most-quoted version of the argument, and it has three defects that a reader can check without leaving the sources it cites.  The 400 m is the deep-sea subset, not the whole sea floor as the sentence claims.  The 1 Gt/yr removal figure is Hay's explicit minimum, presented here as the value.  And 12 Myr is set against roughly 3 Gyr of ocean rather than 64 Myr of ocean floor.  Correcting only the last two removes the discrepancy entirely, without touching the 20 Gt/yr input figure at all.

Where the Sediments Are

Creation Research Society.  No author or date given on the page.

https://www.creationresearch.org/where-the-sediments-are

topic: sediment-budget camp: creationist type: popular access: OA timescale: n/a direction: mixed

Key figures: quotes GlobSed verbatim — "3.37 x 10⁸ km³ of sediment over an oceanic area of about 3.63 x 10⁸ km²" at "an average thickness of 927 m"; margins "0.469 x 10⁸ km²", "1.43 x 10⁸ km³", "3,044 m"; deep sea "2.79 x 10⁸ km²", "1.13 x 10⁸ km³", "404 m."

Tagged mixed because it is the most accurate creationist handling of the GlobSed numbers found anywhere — it is the source of the full three-row table used throughout this section, including the whole-ocean 927 m that other young-earth treatments omit — while its own argument is the one the mainstream literature has already partly conceded and partly answered.  Its conclusion: "there are not enough rocks for uniformitarian processes acting over deep time.  Issues like this force numerous ad hoc excuses, such as erosion or nondeposition."  The honest reply is split.  For deep-sea sediment the exponential-decay model holds and the destruction term is measured, so the objection fails.  For shallow-marine sediment on continental crust the objection is partly right, and Peters and Husson say so in Geology — see that entry.

Ocean sediments and Deep Time

Mitchell, S.  jesusinhistoryandscience.com, 27 January 2025.

https://jesusinhistoryandscience.com/wp-content/uploads/2025/01/23-Ocean-Floor-Mud.pdf

topic: sediment-budget camp: old-earth-christian type: blog access: blog timescale: n/a direction: undercuts

Key figures: quoting Dedkov and Gusarov (2006), "The total global suspended sediment yield into the World Ocean equals 15.5 × 10⁹ t year," and "Recent human activity has increased suspended sediment yield into the World Ocean by 2.6 times."

A ten-page rebuttal by an old-earth Christian and former petroleum geologist, and it arrives independently at four of the same corrections made in this section: Snelling's input flux is high and uncorrected for human amplification, the 400 m omits the shelf, the removal term uses Hay's minimum, and the comparison should be against 150-200 Myr rather than 3 Gyr.  Listed here partly because it is useful and partly because of who wrote it — a reader who assumes the answer to this argument comes only from hostile secular geologists should know that a Christian with field experience got there first and published it.  Full text read.

Recycling: where the eroded rock goes

Atmospheric oxygenation driven by unsteady growth of the continental sedimentary reservoir

Husson, J.M. and Peters, S.E.  Earth and Planetary Science Letters, 460:68-75  (2017).  doi 10.1016/j.epsl.2016.12.012

https://strata.geology.wisc.edu/reprints/Husson_Peters_2017.pdf

topic: crust-recycling camp: mainstream type: model access: OA timescale: 4 Gyr of earth history direction: undercuts

Key figures: global sediment flux through the surface environment about 6 × 10⁶ km³/Myr = 6 km³/yr ≈ 15 Gt/yr at 2.5 t/m³, which spread over 1.4848 × 10⁸ km² of land is 40 m/Myr (= 40 mm/kyr) of mean denudation; Ronov's preserved Phanerozoic sedimentary volume "is 639 million km³" ≈ 1.60 × 10¹⁸ t; of the gross delivery, only "∼20% on average in the Phanerozoic, ∼2% in the Proterozoic, and ∼0.5% in the Archean" is destined for preservation on continental crust.

This is the paper that dissolves Ariel Roth's strongest sediment argument, and it does so by agreeing with his arithmetic.  Roth, following Gregor, computed that 3 cm per thousand years over 3,500 Myr should have produced 23 times more sediment than survives.  Checked here, his sum is right: 30 m/Myr over 3,500 Myr is 105 km of vertical denudation, which over the land area is 1.56 × 10¹⁰ km³, or 24.4 times Ronov's preserved volume.  But rebuilding the same ratio from Husson and Peters' measured preservation fractions — 6 × 10⁶ km³/Myr × (539 × 0.20 + 1,959 × 0.02 + 1,000 × 0.005) — gives 9.1 × 10⁸ km³ preserved out of 2.1 × 10¹⁰ delivered, a ratio of 23.0×.  Roth's number and the geologists' number are the same number.  The disagreement is not about the mass balance; it is about whether recycling is real.

The mass balance closes internally too: 1.60 × 10¹⁸ t at 15 Gt/yr takes 107 Myr to deliver, against a 539 Myr Phanerozoic, for a preserved fraction of 19.8% — reproducing the paper's independently stated ~20% to within one per cent.  The weakness to state: the preservation fractions are themselves derived partly by comparing the preserved record against a modelled flux, so quoting them back to reproduce 23× is not a fully independent check.  The independent evidence that recycling happens is in Blatt and Jones and in the detrital zircon work below.  Full text read.

Sediment cycling on continental and oceanic crust

Peters, S.E. and Husson, J.M.  Geology, 45(4):323-326  (2017).  doi 10.1130/G38861.1

topic: crust-recycling camp: mainstream type: compilation access: paywalled timescale: Phanerozoic direction: mixed

Key figures: "the predictions of a model in which the destruction of sedimentary rock is the predominant process signal are borne out only among sediments deposited on oceanic crust and among sediments deposited above sea level in non-marine environments"; "Most of the surviving volume of sedimentary rock (∼75%) was deposited in and adjacent to shallow seas on continental crust and does not exhibit any steady decrease in quantity with increasing age."

The most important concession in this whole group, and it belongs on the page because it is the one place where the standard geological picture genuinely fails rather than merely being misapplied.  The simple story that old sediment thins out because it has all been recycled away is false for three-quarters of the surviving sedimentary record.  Shallow-marine sediment on continental crust does not decay with age; its abundance fluctuates.  The Creation Research Society objection catalogued above is, on this point, partly correct, and it is a mainstream journal that says so.  What survives is the distinction the paper actually draws: continental crust is not recycled and ocean crust is, so the decay model applies where the crust is being destroyed and not where it is not.  Abstract only; GeoScienceWorld blocks the full text.

Nature of the sedimentary rock record and its implications for Earth system evolution

Husson, J.M. and Peters, S.E.  Emerging Topics in Life Sciences, 2:125-136  (2018).  doi 10.1042/ETLS20170152

https://strata.geology.wisc.edu/reprints/Husson_Peters_2018.pdf

topic: crust-recycling camp: mainstream type: review access: OA timescale: Phanerozoic direction: mixed

Key figures: "Deep sea sediment quantity indeed declines approximately exponentially, with an empirical half-life of ∼30 millions years"; "ocean floor must be destroyed to balance the crustal production rate of ∼3.0 km² year⁻¹ … at mid-ocean ridges"; on their own North American data, "an exponential fit … provides a poor description of the data (Shapiro-Wilk test for normality of residuals, P = 3 × 10⁻⁴⁴)."

The companion to the Geology paper, and it states both halves of the answer in one place.  Deep-sea sediment does decay exponentially, at a rate that follows directly from the measured rate at which sea floor is manufactured and destroyed — "it is impossible for crust, and any sediments deposited on it, to survive intact for long periods of time."  Continental sediment does not, and the authors quote Ronov and colleagues from 1980 saying so before them: the diagram "does not show any regular decrease of the relative mass of the rocks with increasing age and demonstrates only its periodic fluctuations."  The ~3.0 km²/yr crustal production rate also gives a check that needs no sediment data at all — a conveyor producing 3.0 × 10⁶ km²/Myr and carrying a mean age of 64.2 Myr implies a standing ocean-crust area of order 3.9 × 10⁸ km², which is the size of the ocean basins.  Note against the previous entry that this paper's ~30 Myr volumetric half-life and Hay's 19.5 Myr mass half-life differ by 1.5×.  Full text read.

Proportions of exposed igneous, metamorphic, and sedimentary rocks

Blatt, H. and Jones, R.L.  GSA Bulletin, 86(8):1085-1088  (1975).  doi 10.1130/0016-7606(1975)86<1085:POEIMA>2.0.CO;2

topic: crust-recycling camp: mainstream type: compilation access: paywalled timescale: 0-4 Gyr of outcrop direction: undercuts

Key figures: "Of the rocks exposed on the Earth's surface, 66 percent (±3.5 percent) are sedimentary and 34 percent are crystalline, at the 95 percent confidence level."  The age-versus-outcrop-area relation "is well described by a decay curve with a half-life of 130 × 10⁶ yr.  That is, one-half of all existing sedimentary rocks are younger than Jurassic.  Such a short half-life indicates that the rate of sedimentary recycling must be very rapid."  And: "Less than 5 percent of all Precambrian rocks are mapped as sedimentary."

The answer to Roth's objection that recycling is assumed rather than evidenced, and it is a map measurement rather than a model.  Two-thirds of exposed bedrock is already sedimentary rock.  A river eroding a random square kilometre of continent is, two times in three, eroding sediment that has been through the cycle at least once.  Whatever else is uncertain about the sedimentary mass balance, the claim that the eroded material is fresh crystalline crust on its first pass is contradicted by the geological map of the world.

Cross-check: a 130 Myr half-life implies a mean residence of 188 Myr, against the 107 Myr turnover computed from Ronov's mass divided by 15 Gt/yr.  Same order, a factor of 1.8 apart, which is inside the combined uncertainty in mass, density and flux.  Abstract only; GeoScienceWorld serves PDF alone.

Gregor's denudation of the continents

Garrels, R.M. and Mackenzie, F.T.  Nature, 231(5302):382-383  (1971).  doi 10.1038/231382a0

https://www.nature.com/articles/231382a0

topic: denudation-global camp: mainstream type: review access: paywalled timescale: Carboniferous to Recent direction: undercuts

Key figures, verbatim from the first paragraph: "IN his article, 'Denudation of the Continents', Gregor estimated the pre-human global denudation rate at about 90-100 × 10¹⁴ g/yr, and obtained an average rate of about 110 × 10¹⁴ g/yr for the time interval from Carboniferous to Recent."  In modern units: 9-10 Gt/yr pre-human, 11 Gt/yr for the last ~300 Myr; spread over the land area those are about 25 and 27 m/Myr (= mm/kyr).

Catalogued because of what it does to the way Gregor is used.  Roth cites Gregor (1968) for the claim that 14-23 times more sediment should exist than does, and treats it as a missing-mass crisis.  Gregor's own published conclusion runs the other way: the denudation rate reconstructed from the sedimentary mass-age record for the last 300 Myr is equal to or slightly greater than the modern pre-human rate.  That is a finding that present rates are representative of the past — a result the young-earth argument needs and a rebuttal of the "past rates were much lower" reply at the same time.  It is not a deficit.

Handle with care.  Gregor's two papers were not obtained.  The 1968 Koninklijke Nederlandse Akademie van Wetenschappen paper has no digital copy anywhere that was found; Nature 228(5268):273-275 (1970), doi 10.1038/228273a0, has no abstract in PubMed, Crossref or OpenAlex and sits behind a sign-in gate.  Everything attributed to Gregor here comes from the one paragraph of Garrels and Mackenzie that renders without authentication.  Do not quote Gregor directly from this entry.

U-Pb ages of detrital zircons in Jurassic eolian and associated sandstones of the Colorado Plateau

Dickinson, W.R. and Gehrels, G.E.  GSA Bulletin, 121(3-4):408-433  (2009).  doi 10.1130/B26406.1

topic: crust-recycling camp: mainstream type: measurement access: paywalled timescale: Jurassic deposition, Precambrian sources direction: undercuts

Key figures: "U-Pb ages for 1655 individual detrital zircon grains in 18 samples" of Glen Canyon and San Rafael Group sandstones; age peaks at 420, 615, 1055 and 1160 Ma, matching Appalachian, Neoproterozoic and Grenvillian source terranes.

The cleanest physical evidence that sediment is second-hand, because each grain carries its own isotopic address.  "Most detrital zircon grains in Jurassic eolianites were derived ultimately from basement provinces older than 285 Ma in eastern and central Laurentia, rather than from rock assemblages of the nearby Cordilleran margin."  And within the section itself: "Zircons in marine Curtis sandstone were largely reworked from underlying Entrada eolianite."  The authors describe a transcontinental river and wind system that spread mature quartz detritus across the southwest, after which "durable zircon grains were recycled by multiple intraregional depositional systems."

The relevance to a one-year flood is direct.  A single depositional event does not give a zircon population time to be eroded from a distant orogen, transported across a continent, deposited, lithified, re-exhumed and re-deposited locally — nor does it explain why the grains sort into four age peaks that match four distinct and widely separated basement provinces.  Cawood, Hawkesworth and Dhuime (Geology 40(10):875-878, 2012, doi 10.1130/G32945.1) generalise the point: the gap between grain crystallisation age and sediment depositional age is systematic and varies with tectonic setting, which a single global event would not produce.  Both were read as abstracts only; GeoScienceWorld blocks both full texts.

Isostasy

Mantle-driven dynamic uplift of the Rocky Mountains and Colorado Plateau and its surface response

Karlstrom, K.E. and others.  Lithosphere, 4(1):3-22  (2012).  doi 10.1130/L150.1

http://geomorphology.sese.asu.edu/Papers/karlstrom_etal_mantle_uplift_colorado_river_2012.pdf

topic: isostasy camp: mainstream type: model access: paywalled timescale: 10 Myr direction: mixed

Key figures: for broad, locally compensated loads, "a given amount of denudation (D) leads to surface lowering of ~0.2D and 0.8D of rock uplift."  Effective elastic thickness across the region varies from 40 km at the plateau centre to about 10 km at each end; the model used a mean of 25 km.  Modelled rebound reaches up to 1 km in 10 Myr at the plateau centre, which is "~55% of total post-10 Ma incision (compared to 80% for Airy isostasy)"; in Grand Canyon and the southern Rockies it is only 25% of post-10 Ma exhumation.  Denudation along the Colorado and Green River corridor is about 1,500 m, and up to about 1,800 m near the Green River confluence.  The rebound filter has a radius near 150 km.

The one peer-reviewed source that states the wavelength rule plainly, which is why it sits first in this section.  "When the scale of topographic loads is less than several hundred kilometers … the surface load is partially supported by elastic strength of the lithosphere, and the magnitude of rebound is diminished.  Consequently, denudation on a lateral scale of hundreds of kilometers will produce a strong component of rock uplift due to isostatic rebound, whereas smaller-scale incision of narrow canyons results in little rock uplift due to isostatic response."  Continent-wide denudation gets close to the full Airy response; a ten-kilometre-wide canyon gets almost none.  Any argument that says "the Grand Canyon rebounds" is wrong, and a hostile reader will know it.

Tagged mixed rather than undercuts for two reasons.  The rebound is real, modelled from independent data — basalt palaeosurfaces plus thermochronology — and it reaches a kilometre in ten million years, which no young-earth treatment accounts for.  But it comes in at 55-80% of full Airy, not 100%, so the isostatic multiplier used to lengthen levelling times should not be pushed to the top of its range.  The authors add their own caution: "Rebound models need to be interpreted with caution.  Our model, as well as previous models … does not have firm constraints on absolute paleoelevation except for Cretaceous and present times."  Full text read.  The companion paper, Lazear, Karlstrom, Aslan and Kelley (Geosphere 9(4):792-814, 2013, doi 10.1130/GES00836.1), is listed as open access but returns 403 by every route tried and has not been read.

Local relief and the height of Mount Olympus

Montgomery, D.R. and Greenberg, H.M.  Earth Surface Processes and Landforms, 25(4):385-396  (2000).

https://gis.ess.washington.edu/grg/publications/pdfs/61_ftp.pdf

topic: isostasy camp: mainstream type: model access: paywalled timescale: Quaternary valley incision direction: mixed

Key figures: for uniform periodic ridge-and-valley topography with narrow valley floors, "at most (i.e. for C = 1) 42 per cent of local relief translates into higher peaks," from their equation 3, in which the peak gain equals a compensation fraction times crustal density over mantle density times half the local relief, with densities of 2,800 and 3,300 kg/m³.  Rebound at Mount Olympus from valley development is 500 to 750 m, or 21 to 32 per cent of its height, for effective elastic thickness of 5-10 km, and 310-750 m over a flexural rigidity range of 10²¹ to 10²³ N m, that is elastic thickness 5-24 km.  Banks and colleagues put United States flexural rigidity at 10²¹-10²² N m, so elastic thickness 5-11 km; Bechtel and colleagues report elastic thickness under 10 km for western North America.

The free source that prints the flexural machinery with a worked example and real elastic thicknesses, and the honest ceiling on how much of a mountain erosion can build.  Even in a weak, hot, low-rigidity orogen, carving the valleys buys about a quarter to a third of peak height — the authors note this is "comparable to the maximum estimates of 25 to 30 per cent in other" ranges.  Montgomery's earlier paper (JGR Solid Earth 99(B7):13913-13921, 1994, doi 10.1029/94JB00122, abstract only) gives the same result across three ranges: isostatically compensated valley incision accounts for "at most 5-10% of the present elevation of mountain peaks" in the central Sierra Nevada and Tibet, and "as much as 20-30%" in the Himalaya.  Full text of the 2000 paper read.

Observation of glacial isostatic adjustment in "stable" North America with GPS

Sella, G.F. and others.  Geophysical Research Letters, 34:L02306  (2007).  doi 10.1029/2006GL027081

https://www.ngs.noaa.gov/CORS/Articles/2006GL027081.pdf

topic: isostasy camp: mainstream type: measurement access: paywalled timescale: decade of GPS direction: background

Key figures: 360 GPS sites across Canada and the United States; uplift of 10 mm/yr near Hudson Bay "the site of thickest ice at the last glacial maximum," falling with distance and changing "to subsidence (1-2 mm/yr) south of the Great Lakes"; the hinge line separating uplift from subsidence agrees with water-level gauge data; vertical velocity uncertainties "less than 1 mm/yr."

The proof of concept for the whole isostasy section, and it is worth stating in the same units as everything else: 10 mm/yr is 10,000 m/Myr, two hundred to two thousand times any measured continental denudation rate.  The mantle really does flow back under an unloaded plate, it does so fast, and it can be watched happening with satellite geodesy over a decade.  Nobody arguing about erosion and isostasy needs to treat the mechanism as theoretical.

This entry also carries a warning about a use of isostasy that does not work.  Sella's map puts the southeastern United States south of the hinge line, in the collapsing forebulge, subsiding or near zero.  There is no large measured GPS uplift on the American east coast to attribute to erosional unloading, and any argument that asserts one has the sign wrong.  Full text read via the NOAA mirror; AGU and the author's repository both blocked.

Glacial isostatic uplift of the European Alps

Mey, J. and others.  Nature Communications, 7:13382  (2016).  doi 10.1038/ncomms13382

https://www.nature.com/articles/ncomms13382.pdf

topic: isostasy camp: mainstream type: model access: OA timescale: since the Last Glacial Maximum direction: mixed

Key figures: with an average effective elastic thickness of 50 km and an upper-mantle viscosity of 2.2 ± 0.5 × 10²⁰ Pa s, the model reproduces "virtually all of the geodetically measured uplift with ~90% of this uplift caused by deglaciation and only ~10% by erosional unloading."  Elastic thickness was tested from 10 to 70 km; bedrock density 2,700 kg/m³.  Their viscosity "is lower than the viscosity estimated for an old craton (3-10 × 10²⁰ Pa s, Fennoscandia), but higher than that for a region with recent crustal thinning (0.18 × 10²⁰ Pa s, Basin and Range province)."

Catalogued next to Champagnac because the two disagree, and the disagreement should be reported rather than resolved by picking the convenient one.  Champagnac and colleagues attribute half of measured Alpine uplift to erosional rebound; Nocquet and colleagues (Scientific Reports 6:28404, 2016, doi 10.1038/srep28404) measure uplift reaching about 2.5 mm/yr in the northwestern Alps, model erosional unloading at 0.3-0.5 mm/yr, and conclude that erosion and glacial adjustment together "only explain at most 60% of the highest uplift rates."  Mey and colleagues get 90% from deglaciation alone.  All three are recent and peer-reviewed.  What they agree on is that erosional rebound in the Alps is real and is of order a few tenths of a millimetre per year; what they disagree on is its share.  Full text of Mey read; the Nocquet figures were read through a fetch summariser rather than raw and should be re-checked before printing.

Quaternary erosion-induced isostatic rebound in the western Alps

Champagnac, J.-D. and others.  Geology, 35(3):195-198  (2007).  doi 10.1130/G23053A.1

topic: isostasy camp: mainstream type: model access: paywalled timescale: 1 Myr direction: undercuts

Key figures: modelled isostatic response to erosional unloading "reaches a maximum of [~]500 m since 1 Ma in the inner Swiss Alps," which is 500 m/Myr = 0.5 mm/yr; assuming steady erosion since 1 Ma, "this rebound accounts for half of the measured vertical motion of 1.1 mm/yr in the southern Valais"; vertical movement "extends across the entire belt, including peri-Alpine basins."

The clearest single measurement-plus-model of erosional rebound in a mountain belt, and it also removes the obvious alternative explanation.  The Alps are currently extending, and as the authors point out, "isostatic compensation of thinning crust should lead not to increased but to decreased mean elevations."  So the observed uplift cannot be crustal thickening; it has to be unloading of one kind or another.  Abstract only — GeoScienceWorld blocks the paper and the HAL record carries no deposited PDF, so the figures above are the abstract as served by the HAL API.

Isostatic rebound due to glacial erosion within the Transantarctic Mountains

Stern, T.A.; Baxter, A.K.; and Barrett, P.J.  Geology, 33(3):221-224  (2005).  doi 10.1130/G21068.1

topic: isostasy camp: mainstream type: model access: paywalled timescale: Cenozoic glacial incision direction: mixed

Key figures: "In temperate climates, ~25% of peak elevations in mountain ranges can be created by isostatic rebound as a response to erosional incision."  In polar conditions the figure is larger: flexural modelling shows rebound from glacial incision "can account for as much as 2000 m or 50% of peak elevation in the central Transantarctic Mountains."  Relief differences of "at least 5500 m over lateral distances of just 40 km."

The upper end of what erosional rebound can do to a mountain range, and a mechanism for longevity rather than for youth: "Because isostatic rebound results in permanent peak uplift, this mechanism provides an explanation of why the Transantarctic Mountains are one of the higher and more long-lived continental rift margins on Earth."  A range that carves itself deeper and rises as it does so is not a range on its way to being planed flat.  Abstract only.

Citation note: this is the Transantarctic Mountains paper, not a Stern paper on the Southern Alps of New Zealand.  No such New Zealand isostasy paper by Stern was located, and the two should not be conflated.

Elastic strength of the Slave craton at 1.9 Gyr and implications for the thermal evolution of the continents

Grotzinger, J.P. and Royden, L.  Nature, 347(6288):64-66  (1990).  doi 10.1038/347064a0

https://www.nature.com/articles/347064a0

topic: isostasy camp: mainstream type: measurement access: paywalled timescale: 1.9 Gyr direction: undercuts

Key figures: flexural analysis of the Proterozoic Kilohigok basin "shows that 1.9 Gyr ago the craton had an effective elastic thickness of only 12 ± 4 km.  Because the present effective elastic thickness of the region is ~100 ± 25 km, these results imply a two- to fourfold decrease in the thermal gradient beneath the Slave craton since 1.9 Gyr."

The answer to the strongest form of the "you would burn through the whole crust" objection, and it is a measurement rather than a model.  Effective elastic thickness is read at two epochs 1.9 billion years apart from the flexural signature preserved in a sedimentary basin, and the direction of change is toward more strength, not less.  Cratons do not soften and thin with age; they cool, stiffen and stabilise.  The modern figure of about 100 km is also the number that sets the flexural wavelength on a shield at 800-1,100 km, which is why a canyon on a craton produces no rebound at all while continent-wide stripping produces nearly the full Airy response.  Abstract only — the Nature abstract renders without authentication, the full text does not.

Continental erosion and large-scale relief

Pinet, P. and Souriau, M.  Tectonics, 7(3):563-582  (1988).  doi 10.1029/TC007i003p00563

topic: denudation-global camp: mainstream type: compilation access: paywalled timescale: 10^6^-10^8^ yr direction: mixed

Key figures: two regression laws for mechanical denudation Ds against mean basin elevation H.  For basins in orogens younger than 250 Ma, "Ds (m/10³ yr) = 419 × 10⁻⁶ H (m) - 0.245," explaining 95.1% of variance.  For older orogens, "Ds (m/10³ yr) = 61 × 10⁻⁶ H (m)," explaining 86.5%.  At H = 1,000 m those give 174 m/Myr and 61 m/Myr respectively; at the global mean land elevation of 795 m, 88 and 48 m/Myr.  The bare relief e-folding time is 2.4 Myr for young orogens and 16.4 Myr for old ones — about 98 Myr for old terrain once a ×6 isostatic multiplier is applied — against the paper's own reported range of "15 to 360 m.y."

The measured version of the objection John Woodmorappe raised and nobody followed up: erosion rate is not independent of topography.  It is proportional to elevation, with a different constant for young and old terrain, and old terrain erodes 6.9 times more slowly per metre of elevation than young orogenic terrain.  That single ratio is why cratons survive and why a global average rate applied uniformly to a continent gives the wrong answer.

The intercepts carry a second result that bears directly on any continental mass balance.  For young orogens "the negative intercept is interpreted as a continental sedimentation rate of 245 m/m.y." — a large share of the eroded rock is stored on the continent rather than exported to the ocean.  For old orogens, "the null intercept suggests the lack of continental storage."  Abstract only.

CRUST1.0: a 1-degree global crustal model

Laske, G.; Masters, G.; Ma, Z.; and Pasyanos, M.  (2013).

https://igppweb.ucsd.edu/~gabi/crust1/crust1.0.tar.gz

topic: isostasy, method camp: mainstream type: compilation access: OA timescale: present-day snapshot direction: undercuts

Key figures, all computed here from the 64,800-cell grid using exact WGS84 cell areas: area-weighted mean continental crustal thickness 39.07 km over 149.83 × 10⁶ km² of land cells; median 39.0 km; 5th percentile 30.0 km, 95th percentile 49.0 km; 3.4% of land has crust under 30 km.  Cells under 0-200 m of water: 20.36 × 10⁶ km², mean crustal thickness 31.48 km.  Land minus shelf is 7.59 km of crust, 19.4% of the total, separating a continent at +795 m from one at −70 m.

Listed as a source because the article's answer to the strongest young-earth counter rests on it, and a reader should be able to redo the sum.  Monte Fleming's objection is that isostatic rebound cannot rescue an old earth, because stripping the surface and letting it pop back up repeatedly means "you're going to get through the whole thickness of the crust pretty pretty quickly."  The physics is right and he deserves credit for granting it.  The arithmetic is not.  Levelling the continents from a mean elevation of 794.5 m with an isostatic multiplier of 6.0 requires removing 4.77 km of rock, which is 12.2% of a 39.07 km crust — 8.8% to 14.9% across the density bracket.  Not the whole thickness, and off by roughly a factor of eight.  The process also self-terminates: subaerial erosion stops at sea level.

The complication, stated before a geophysicist raises it.  Regressing surface elevation against crustal thickness over all cells with 25-60 km of crust (n = 27,403, area-weighted) gives elevation (km) = 0.0707 × thickness (km) − 2.122, with r = 0.466.  The observed slope is less than half the Airy prediction of 0.14-0.21, and the correlation is weak.  Continental elevation is not set by crustal thickness alone — cratons ride high on buoyant depleted mantle roots, orogens are dragged down by cold dense lithosphere, and mantle flow adds hundreds of metres of dynamic topography.  This does not rescue the counter-argument, because the static elevation-thickness relation is a different question from the response to unloading, where the relevant contrast is between the rock removed and the asthenosphere flowing in beneath.  But the weak static correlation is real and should be said out loud.

Total-Earth area from the same computation came to 510.066 × 10⁶ km² against the accepted 510.07 × 10⁶ km², which is the check that the area weighting is right.  The standard journal citation for CRUST1.0 was not recorded in the source notes; the model itself was downloaded and used directly.

Cliff retreat: the millennial measurements

Recent acceleration in coastal cliff retreat rates on the south coast of Great Britain

Hurst, Martin D.; Rood, Dylan H.; Ellis, Michael A.; Anderson, Robert S.; and Dornbusch, Uwe.  PNAS, 113(47):13336-13341  (2016).  doi 10.1073/pnas.1613044113

https://europepmc.org/articles/pmc5127313?pdf=render

topic: coastal, cosmogenic camp: mainstream type: measurement access: OA timescale: 7 kyr against ~140 yr direction: undercuts

Key figures: two transects in Upper Cretaceous chalk, East Sussex — Hope Gap, cliff height 15 m, and Beachy Head, cliff height 50 m.  ¹⁰Be measured in in-situ flint nodules on the shore platform, collected at spring tides on 24-25 July 2013.  Millennial retreat 5.7 (+0.3/−0.3) cm/yr at Hope Gap and 2.6 (+0.2/−0.2) cm/yr at Beachy Head.  Convert horizontal retreat as 1 cm/yr = 10,000 m/Myr = 10 km per million years, so those are 26,000 to 57,000 m/Myr.  Historic rates from mapped clifftop positions since 1870: about 32 cm/yr at Hope Gap and about 22 cm/yr at Beachy Head, with a 0.05-0.8 m/yr range along this coast.  Historic exceeds millennial by roughly 10×.  Modelled step at Beachy Head: up to 30.4 (+8.3/−106) cm/yr, 293 (+170/−80) years ago.

The single most damaging measurement against the coastal version of the young-earth argument, and it must be conceded first.  The abstract states the result without hedging: "The ¹⁰Be concentrations are consistent with retreat rates of chalk cliffs that were relatively slow (2-6 cm·y−1) until a few hundred years ago.  Historical observations reveal that retreat rates have subsequently accelerated by an order of magnitude (22-32 cm·y−1)."  The geometric argument behind it needs no model at all: at 32 cm/yr the 250-350 m of platform the authors sampled would have formed in the last 1,500 years and would carry almost no ¹⁰Be, which is not what the samples contain.  Their explanation is beach thinning, not sea level — the fringing flint-and-sand beaches were larger in the nineteenth century, and a beach both supplies abrasive tools and shields the cliff toe.  The bias runs one way only: more beach cover in the mid-late Holocene would force the long-term rate lower still, widening the gap.

Two weaknesses to record.  The Hope Gap best fit steps down, to 1.3 (+1.1/−0.3) cm/yr about 308 years ago, while the mapped clifftop record at that same site gives about 32 cm/yr since 1870; the modelled step and the historical observation point opposite ways there, and the headline acceleration rests on Beachy Head plus the platform-width geometry.  And the Beachy Head step carries an uncertainty of +8.3/−106 cm/yr, larger than the value itself.  The paper also does not solve the deep-time problem it is often used to solve: even the slow Holocene chalk rate of 2-6 cm/yr is 20-60 km/Myr, which erases any continental shelf on earth in ten million years.  It moves the discrepancy from a factor of thirty to a factor of three.

GlobR2C2: a global relational database of coastal rocky cliff erosion rates

Prémaillon, Mélody; Regard, Vincent; Dewez, Thomas J.B.; and Auda, Yves.  Earth Surface Dynamics, 6:651-668  (2018).  doi 10.5194/esurf-6-651-2018

https://esurf.copernicus.org/articles/6/651/2018/esurf-6-651-2018.pdf

topic: coastal, timescale-bias camp: mainstream type: compilation access: OA timescale: 1 month to 7,100 yr; half the records 56-63 yr direction: mixed

Key figures: 58 sources — 47 peer-reviewed articles plus 11 national databases — covering 1,530 cliff sites and 1,680 erosion-rate records.  Medians by rock strength: hard rock 2.9 cm/yr (n = 341, median absolute deviation 3.4), medium 10 cm/yr (n = 63, MAD 7.8), weak 23 cm/yr (n = 403, MAD 25, with rates above 10 m/yr).  In m/Myr those are 29,000, 100,000 and 230,000 m/Myr, or 29, 100 and 230 km per million years.  Chalk is classed as weak rock.  Geography is lopsided: Europe 42 studies and 1,579 records, Oceania 3 and 94, North America 4 and 50, Asia 2 and 4, South America 1 and 1, and "No literature was found for the entire African continent."  Aerial-photo and historic-map comparison supplies 88% of all records; lidar and structure-from-motion supply 5%.

This is the best global number available for hard-rock cliff retreat, and the entry a reader should check any coastal claim against.  It also confirms the argument's core factual premise, which is worth stating plainly: cliffs only retreat.  Unlike a sandy shoreline, a rock cliff has no mechanism to advance seaward, so every metre it loses is permanent.

The finding that matters most, though, is what the median leaves out.  The French CEREMA national inventory alone contributes 465 records, 53% of the whole database, surveying 1,800 km of French rocky cliff every 200 m — and it cannot see slow coasts: "CEREMA acknowledges that the quality of photographs limits the detectable cliff recession to rates higher than 10 cm yr−1.  Below this value, they deem recession rates as undetermined."  The authors then write that "These null values represent 67 % of the studies of rocky coasts, which means that slowly eroding rocky coasts are common and ignoring this information can affect conclusions."  The published 2.9 cm/yr median is computed after those nulls are deleted; Table 2 shows that setting them to 1 mm/yr instead drops the hard-rock median to 0.001 m/yr, a factor of thirty.  Note the loose wording — the sentence as printed says 67% of the studies, where the sense required is 67% of the surveyed sections — so quote it with care.  The authors also quantify short-survey bias directly: at a Normandy chalk site, one 70,000 m³ rockfall during 2.5 years of terrestrial lidar caused more than 19 m of local clifftop retreat and multiplied the estimated annual rate sevenfold "just by including this random and definitely unrepresentative event."  At the same site the long and short records agree well, 15 cm/yr from 29 years of aerial photos against 11-13 cm/yr from ¹⁰Be over 3,000 years (Regard et al. 2012, known here only through this paper).  Minor internal inconsistency, recorded so nobody trips on it: the abstract says "more than 1680" records, Section 3.1 says 1,530 sites and 1,680 records, and the Conclusions say "ca. 2000."

Evidence of slow millennial cliff retreat rates using cosmogenic nuclides in coastal colluvium

Bossis, Rémi; Regard, Vincent; Carretier, Sébastien; and Choy, Sandrine.  Earth Surface Dynamics, 13:71-79  (2025).  doi 10.5194/esurf-13-71-2025

https://esurf.copernicus.org/articles/13/71/2025/esurf-13-71-2025.pdf

topic: coastal, cosmogenic camp: mainstream type: measurement access: OA timescale: millennial direction: undercuts

Key figures: ¹⁰Be measured in the colluvium shed off the cliff face, at 8 sites on the Mediterranean coast of the eastern Pyrenees and 3 on the desert coast of southern Peru.  Retreat 0.3-0.6 mm/yr for five Mediterranean samples = 300-600 m/Myr, i.e. 0.3-0.6 km per million years.  Two vegetated-colluvium samples give about 0.1 mm/yr = 100 m/Myr.  The wave-attacked Peruvian site gives 0.5 mm/yr; two Peruvian sites out of reach of waves for tens of thousands of years erode about 20 times slower.  Comparison catchment denudation in the same regions: Andes 0.015-0.02 mm/yr, Pyrenees 0.06-0.14 mm/yr.

The most important number in the whole coastal file, because it is the only measured rate slow enough for a 100-200 Myr passive margin to absorb.  At 0.3-0.6 mm/yr a coast retreats 30-60 km in 100 Myr, which is the order of an actual shelf width.  That is 50 to 100 times slower than the GlobR2C2 hard-rock median, and the method is the reason: platform ¹⁰Be needs a platform, so slow coasts have simply never been measured over millennia.  The authors say so directly — "there are no available methods that can be used to quantify slow coastal erosion (< 1 cm yr−1) averaged over millennia" — and their colluvium technique is the first that reaches them.

State the limits honestly, because a careful skeptic will not.  This is n = 11 on two low-energy coasts, one of them microtidal Mediterranean and one of them a desert where two of three sites are not being attacked by waves at all.  It is not a global result and it does not yet establish that ordinary open-ocean rock coasts run this slow.  The authors also report the opposite comparison, which cuts the other way: "These data show that rocky coasts are eroding 1 to 20 times faster than catchments in the same regions on average over the last few thousand years."  Coasts are still the fast edge of the landscape.

Cosmogenic exposure dating reveals limited long-term variability in erosion of a rocky coastline

Swirad, Zuzanna M.; Rosser, Nick J.; Brain, Matthew J.; Rood, Dylan H.; Hurst, Martin D.; Wilcken, Klaus M.; and Barlow, John.  Nature Communications, 11:3804  (2020).  doi 10.1038/s41467-020-17611-9

https://www.nature.com/articles/s41467-020-17611-9.pdf

topic: coastal, timescale-bias camp: mainstream type: measurement access: OA timescale: 7 kyr against 7 yr of monitoring direction: mixed

Key figures: Hartle Loup, North Yorkshire — a 300 m wide shore platform of Jurassic shale and fine sandstone, mean surface slope 0.4 degrees, below a cliff about 60 m high.  Cliff retreat steady at 4.5 ± 0.63 cm/yr over the last 7 kyr = 45,000 m/Myr, or 45 km per million years.  Platform down-wearing 0.314 ± 0.044 mm/yr = 314 m/Myr.  Contemporary monitored cliff retreat at the same site: 2.7 ± 2.9 cm/yr over 7 years, statistically indistinguishable from the 7-kyr rate.  Relative sea level rose about 5.8 m over the modelled 7 kyr.

The honest counterweight to over-reliance on Hurst 2016, and the reason the acceleration story cannot be treated as a general law.  Here the modern rate and the millennial rate match.  The paper also undercuts the simplest mainstream explanation for cliff retreat: "Our results imply a lack of a direct relationship between relative sea level over centennial to millennial timescales and the erosion response of the coast."  If sea-level rise does not drive retreat on this coast, then the appeal to a special Holocene sea-level regime does not automatically rescue the deep-time budget either.

The authors flag their own limitation, which should be quoted alongside the result: their sampling could not have detected a Sussex-style post-industrial acceleration, because only two of their sample sites lie on the strip of foreshore that would record it.  A steady 4.5 cm/yr is also not a young-earth-friendly number in deep time — it removes 4,500 km of coast in 100 Myr.

Reconciling short- and long-term measurements of coastal cliff erosion rates

Acharya-Chowdhury, L.; Dickson, Mark E.; Norton, Kevin P.; Rowland, Julie V.; Hall, B.; and Stephenson, Wayne J.  Engineering Geology, 341:107703  (2024).  doi 10.1016/j.enggeo.2024.107703

https://api.openalex.org/works/doi:10.1016%2Fj.enggeo.2024.107703

topic: coastal, timescale-bias camp: mainstream type: measurement access: paywalled timescale: 6 kyr platform width against 1-10 yr laser scanning direction: mixed

Key figures: Rothesay Bay, Hauraki Gulf, New Zealand — low wave energy, meso-tidal, cliffs 10-30 m high in horizontally bedded soft flysch, fronted by a 140 m wide near-horizontal shore platform with an abrupt seaward edge.  17 terrestrial laser scans from 2014 to 2023, plus monthly scans from July 2021 to July 2022, along 120 m of scan line.  Long-term rate from platform width, assuming 6,000 years of stable Holocene sea level and a fixed platform edge: 1.4 ± 0.1 to 14.3 ± 0.1 mm/yr = 1,400 to 14,300 m/Myr.  Decadal laser-scan rate 41 ± 2 mm/yr = 41,000 m/Myr.  Monthly rates range from 30 to 288 mm/yr.

Reproduces the Hurst discrepancy at a second site in different rock, and states the ambiguity that decides the whole question: "The cliff recession rate detected with TLS is 3 to 30 times higher than erosion rates derived based on the shore platform width.  If erosion had been constant at this rate over approximately 6000 years, a total cliff retreat of >245 m would be expected, whereas the contemporary shore platform is only 140 m wide."  Two readings are possible and the authors offer both: modern retreat really is faster than past retreat, or the platform's seaward edge is not the Holocene starting line.  New bathymetry seaward of the edge shows "multiple rocky features" with steep slope breaks and planar surfaces, which would mean the platform-width method gives a lower bound on Holocene retreat rather than a measurement of it.

Access caveat, stated plainly: only the reconstructed abstract was obtained through OpenAlex.  The full text is closed and Unpaywall lists no open copy, so the site description, scan counts and rate ranges above come from the abstract and not from the body of the paper.

Spatial and temporal trends in California coastal cliff retreat

Swirad, Zuzanna M. and Young, Adam P.  Geomorphology, 412:108318  (2022).  doi 10.1016/j.geomorph.2022.108318

https://api.openalex.org/works/doi:10.1016%2Fj.geomorph.2022.108318

topic: coastal camp: mainstream type: measurement access: paywalled timescale: 5-7 yr (2009-2011 to 2016) direction: mixed

Key figures: airborne lidar along 866 km of California coastal cliffs, comparing 2009-2011 with 2016.  Erosion above the level of detection at 55% of cliffs, which leaves 45% showing none.  State-averaged cliff face and cliff top retreat rates both 0.06 m/yr = 60,000 m/Myr, or 60 km per million years.  Highest face retreat rates, above 1 m/yr, occur at weak rocks.  No significant correlation with rock hardness, rainfall, groundwater, waves or relative sea-level change.

The modern statewide number, and the one that settles what Monte Fleming's California figure is worth.  His stated "erosion rate of about six centimeters a year" is exactly this statewide lidar mean, so his input is correct.  The local values he also reads off a cliff-retreat viewer — 0.85 m/yr and 2.14 m/yr near San Francisco — are 14 and 36 times that mean, and using them for a continental budget is the error, not the 6 cm/yr.  The paper's own comparison is equally awkward for anyone quoting a single number: mean clifftop retreat rates came out lower than in the two earlier survey intervals, while cliff face rates and their statistical distributions were nearly identical between 1998-2011 and 2011-2016 (r² = 0.99).

Two honesty notes.  Only the abstract was obtained; the ScienceDirect text is paywalled and the NOAA repository copy returns 403.  And no California statewide median cliff retreat rate is reported anywhere in the accessible material, so no median should be quoted for the state.

National Assessment of Shoreline Change Part 4: Historical Coastal Cliff Retreat along the California Coast

Hapke, Cheryl J. and Reid, David.  USGS Open-File Report 2007-1133  (2007).

https://pubs.usgs.gov/of/2007/1133/of2007-1133.pdf

topic: coastal camp: mainstream type: measurement access: OA timescale: ~70 yr (1930s to 1998) direction: mixed

Key figures: "The average rate of coastal cliff retreat for the State of California was -0.3±0.2 m/yr, based on rates averaged from 17,653 individual transects measured throughout all areas of California's rocky coastline.  The average amount of cliff retreat was 17.7 m over the 70-year time period of our analysis."  That is 300,000 m/Myr, or 300 km per million years.  Regional means: northern California -0.5 m/yr, central -0.3, southern -0.2.  Coverage is 353 km of measured cliff out of a 1,603 km region length, about 22%.  Highest single transect -2.3 m/yr, at a small headland 3 km north of the Klamath River mouth, 168 m of retreat in about 70 years; -1.8 m/yr in unlithified Quaternary dunes in southern Monterey Bay where dune sand had been mined; -1.3 m/yr at Point Sal.

Historical topographic sheets against 1998 lidar, and the statewide anchor for any claim about the Californian coast before the airborne-lidar era.  Read it beside Swirad and Young 2022, which measured the same coast with lidar over 2009-2016 and got 0.06 m/yr — one fifth of this figure.  That fivefold difference between two careful surveys of the same coastline is itself the strongest argument that a single measured cliff retreat rate is not a stable quantity, and it is a caution that applies to whichever number a writer prefers.

Southern California is the slowest region, and the report attributes that "potentially because of the abundance of protective structures" — a reminder that even the historical baseline is not a natural rate.

Sea-level rise will likely accelerate rock coast cliff retreat rates

Shadrick, Jennifer R.; Rood, Dylan H.; Hurst, Martin D.; Piggott, Matthew D.; Hebditch, Brendan G.; Seal, Alexander J.; and Wilcken, Klaus M.  Nature Communications, 13  (2022).  doi 10.1038/s41467-022-34386-3

https://www.nature.com/articles/s41467-022-34386-3.pdf

topic: coastal, cosmogenic camp: mainstream type: model access: OA timescale: 8 kyr modelled, ~130 yr historic direction: mixed

Key figures: historic retreat over about 130 years of 5.8 ± 4.0 cm/yr at Bideford and 5.9 ± 4.3 cm/yr at Scalby, with local variability spanning 2 to 25 cm/yr; 58,000-59,000 m/Myr at the site means.  The model reconstructs 8,000 years of retreat and forecasts increases of "13 times at Bideford and 6 times at Scalby" by 2100 under RCP8.  "Rock coasts make up >50% of global coastlines."

The paper that ties the rate of Holocene sea-level rise to the rate of cliff retreat, which is the mechanism that would let one explanation cover both the measured Holocene slowdown and the deep-time budget.  If retreat scales with the rate of relative sea-level change, then the slow, near-static sea level of the last few thousand years is a period of unusually slow cliff retreat, and rates measured in it are not representative of anything longer.  That is the strongest available form of the mainstream reply, and it is a model, not a measurement.

Recorded weakness: only the abstract and scattered results lines of this paper were read, so the model's structure and its fit statistics have not been checked here, and its conclusion is directly contested by Dickson et al. 2023 below.  Swirad et al. 2020, at Hartle Loup, found no relationship between relative sea level and cliff erosion over exactly the timescales this model invokes.

Sea-level rise may not uniformly accelerate cliff erosion rates

Dickson, Mark E.; Matsumoto, Hironori; Stephenson, Wayne J.; Swirad, Zuzanna M.; Thompson, Colin F.; and Young, Adam P.  Nature Communications, 14:8485  (2023).  doi 10.1038/s41467-023-44149-3

topic: coastal camp: mainstream type: review access: OA timescale: decades to ~130 yr direction: mixed

Key figures: Shadrick's model matched the observed rate at Scalby, 0.059 against 0.054 m/yr, but not at Bideford, where 0.058 m/yr was observed and 0.017 m/yr modelled — a factor of 3.4 miss at one of the two sites the model was built on.

A formal Matters Arising against the entry above, and it belongs in the register precisely because it shows the mainstream account of cliff retreat is unsettled rather than closed.  Its second point cuts directly at how the young-earth argument selects its numbers: "historical cliff erosion observations in California show that locations with higher/lower historical cliff erosion rates generally had lower/higher recent erosion rates."  That is regression to the mean in the Californian record.  Picking the fastest 5 km of Californian coast and treating it as the continental rate is therefore not merely unrepresentative — the fastest sections are the ones most likely to be measured slow next time.

Recorded limits: three pages, read once, and it makes no measurement of its own.

Global distribution of coastal cliffs

Young, Adam P. and Carilli, Jessica E.  Earth Surface Processes and Landforms, 44(6):1309-1316  (2019).  doi 10.1002/esp.4574

https://escholarship.org/content/qt579326kr/qt579326kr.pdf

topic: coastal, method camp: mainstream type: compilation access: OA timescale: n/a direction: background

Key figures: "The results suggest coastal cliffs likely exist on about 52% of the global shoreline."  Method: SRTM 3-arcsecond elevation data plus the Arctic Coastal Dynamics Database, backshore elevations sampled at 1 km alongshore intervals at about 1,340,000 locations covering 81% of the world vector shoreline, trained against mapped European cliffs and extended to 89% coverage using literature values for Greenland and Antarctica.  Previous estimates had run near 80%.  Cliff retreat rates "have been quantified in at least one location within 33% of independent coastal states and 15% of non-independent regions."

The scope number, and it validates one of Fleming's complaints.  Global shoreline-change assessments based on satellite imagery cover sandy beaches, which are 31% of the ice-free shoreline; cliffs are 52% and no global rate assessment covers them.  The half of the world's coast that can only lose ground is the half nobody has systematically measured.  Shadrick et al. 2022 give a consistent figure independently, stating that rock coasts make up more than 50% of global coastlines.

Access note: the abstract was read through OpenAlex and the open eScholarship PDF downloaded intact, but the body of the paper was not read past the abstract, so nothing beyond the figures above should be attributed to it.

Global patterns of shore platform evolution and cliff retreat: insights from cosmogenic radionuclides

Jeong, A.; Seong, J.B.; Choi, K.H.; and Swirad, Zuzanna M.  Earth-Science Reviews, 272:105323  (2026).  doi 10.1016/j.earscirev.2025.105323

topic: coastal, cosmogenic camp: mainstream type: review access: paywalled timescale: millennial direction: background

Text not obtained.  Listed because it is, on its title and scope, the single most relevant paper in existence to the coastal argument — a global synthesis of the millennial cliff-retreat rates that the whole dispute turns on.  It is closed access with no open copy indexed on OpenAlex, and no part of it has been read.  Nothing in this article rests on it.  It is flagged here as an open obligation: any serious treatment of coastal retreat rates should obtain it, by interlibrary loan or author request, before claiming the millennial global picture is unmeasured.

The self-limiting feedbacks

Beach and sea-cliff dynamics as a driver of long-term rocky coastline evolution and stability

Limber, Patrick W. and Murray, A. Brad.  Geology, 39(12):1147-1150  (2011).  doi 10.1130/G32315.1

topic: coastal camp: mainstream type: model access: paywalled timescale: 10^3^-10^4^ yr model runs direction: undercuts

The strongest single statement of the negative-feedback answer to the coastal argument.  Sea-cliff retreat is not a constant nibble; it is "a nonlinear function of beach width, where cliff retreat is maximized by sediment abrasion and minimized by either a lack of beach sediment or too much sediment (which prevents waves from reaching the sea cliff)."  Because a retreating cliff manufactures the sediment that becomes its own protective beach, the system regulates itself.  The modelling conclusion is the part that matters: "Numerical experiments indicate that through such interactions, rocky coastlines can reach an equilibrium configuration where headlands and embayments remain stable through time, even in the absence of alongshore variations in sea-cliff lithology."

A rocky coastline can therefore hold a stable plan shape over long intervals with no uplift, no sediment import, and no difference in rock strength to explain it.  That is a mechanism the young-earth coastal argument does not address anywhere in its current literature, and it was published fifteen years before the argument's present revival — it is not a rescuing device invented in response.

Two limits stated plainly.  This is a model, not a measurement.  And the equilibrium it finds is a plan-view equilibrium of headlands and bays, not a claim that the shoreline stops moving landward; nobody in this literature claims the feedback drives retreat to zero over 100 Myr.  Only the reconstructed abstract was obtained, so the model's parameters and run lengths have not been checked here.

Feedback structure of cliff and shore platform morphodynamics

Payo, Andrés; Hall, Jim W.; Dickson, Mark E.; and Walkden, Mike J.A.  Journal of Coastal Conservation, 19:847-859  (2015).  doi 10.1007/s11852-014-0342-z

https://link.springer.com/content/pdf/10.1007/s11852-014-0342-z.pdf

topic: coastal camp: mainstream type: model access: paywalled timescale: decades to millennia direction: undercuts

Key figures: four feedback loops control horizontal cliff erosion — three balancing (cliff-toe wave energy depletion, groundwater pore-pressure diminution, cliff deposit protection) and one reinforcing (abrasion enhancement).  Loop strength for the cliff-toe energy depletion loop is given as order −10⁻¹⁰ to −10⁻⁴ for poorly lithified rock coasts.

A formal loop analysis of cliff and platform morphodynamics, and it puts numbers where Limber and Murray put a mechanism.  The authors' own summary: "The backwearing erosion rate (cliff horizontal erosion) has been found to be controlled by at least four feedback loops; three balancing (cliff toe wave energy depletion, ground-water pore pressure diminution and cliff deposit protection) and one positive loop (abrasion enhancement)."  Three of four identified controls are self-limiting.  A retreat rate measured on a young, narrow platform is therefore the fastest that stretch of coast will ever run, and extrapolating it backwards over millions of years extrapolates the system's maximum.

The full paper was fetched, but only the abstract was read closely and the body skimmed, so the loop-strength figure above should be re-checked against the text before it is used as a load-bearing number.

A predictive Mesoscale model of the erosion and profile development of soft rock shores

Walkden, Mike J.A. and Hall, Jim W.  Coastal Engineering, 52(6-7):535-563  (2005).  doi 10.1016/j.coastaleng.2005.02.005

topic: coastal, method camp: mainstream type: model access: paywalled timescale: decades to centuries direction: background

Text not obtained — nothing below is quoted from it.  This is the SCAPE model, the standard soft-rock shore profile model, and it is the reference behind the "equilibrium retreat" and "widening platform" alternatives that Hurst et al. 2016 test against their ¹⁰Be data.  Two follow-ups extend it to sea-level change and are listed here with verified citations only: Walkden, M. and Dickson, M., "Equilibrium erosion of soft rock shores with a shallow or absent beach under increased sea level rise," Marine Geology 251(1-2):75-84 (2008), doi 10.1016/j.margeo.2008.02.003; and Walkden, M. and Dickson, M., "Equilibrium responses of cliffed coasts to changes in the rate of sea level rise," Marine Geology 284(1-4):217-229 (2011), doi 10.1016/j.margeo.2011.01.007.

The same disclosure applies to the two rocky-coast standard works this field leans on, Trenhaile's papers and Sunamura's Geomorphology of Rocky Coasts (1992).  Both are cited repeatedly inside Prémaillon et al. 2018 and Hurst et al. 2016, which were fetched in full; neither original was obtained.  Sunamura is the source of the observation, quoted at second hand in Prémaillon's abstract, that measured rocky-coast erosion rates span four orders of magnitude — a spread worth remembering whenever a single global cliff retreat figure is quoted.  Anything attributed to these works in this article is attributed through the paper that quotes them, and is labelled as such.

Shoreline change assessments: the sandy coast

The State of the World's Beaches

Luijendijk, Arjen; Hagenaars, Gerben; Ranasinghe, Roshanka; Baart, Fedor; Donchyts, Gennadii; and Aarninkhof, Stefan.  Scientific Reports, 8:6641  (2018).  doi 10.1038/s41598-018-24630-6

https://ris.utwente.nl/ws/files/256151099/Luijendijk_2018_The_state_of_the_worlds_beaches.pdf

topic: coastal, human-impact camp: mainstream type: measurement access: OA timescale: 33 yr (1984-2016) direction: undercuts

Key figures: 2.2 million transects at 500 m spacing along the world's sandy shorelines, from Landsat and Sentinel imagery, 1984-2016.  "24% of the world's sandy beaches are eroding at rates exceeding 0.5 m/yr, while 28% are accreting and 48% are stable."  16% erode and 18% accrete faster than 1 m/yr; about 7% are severely eroding; 4% exceed 5 m/yr and 2% exceed 10 m/yr.  Global net: the beaches "have accreted on average 0.33 m/yr over the past three decades, i.e. a total gain of 3,663 km2 over this period."  Asia +1.27 m/yr, Australia −0.20, Africa −0.07.  Sandy shoreline is 31% of the world's ice-free coast (Africa 66%, Europe 22%).  Marine protected areas fare worse than average: 37% eroding against 32% accreting.  For contrast, Bird's earlier interview-based survey found "70% (10%) of the world's sandy beaches experienced net erosion (accretion) while 20% were stable."

The strongest single reply to "the coasts are disappearing," and the paper Fleming engages with directly in Let's Talk Creation Episode 87.  His scope objections are both correct.  It covers only sandy shoreline, 31% of the ice-free coast — his "about 31%" is exactly the paper's figure — and it says nothing about sea cliffs, which are the other half.  And the global accretion figure is contaminated by land-making, which the authors concede outright: "Two large-scale land reclamations appear in the top seven accretive beaches in the world," and Asia's +1.27 m/yr is "likely due to the artificial development of the Chinese coast and large land reclamations in, for example, Singapore, Hong Kong, Bahrain and UAE."  The paper reports no de-reclaimed global mean, so Fleming's demand for one is fair.

His third objection is wrong as stated and partly right in consequence, and both halves should be said.  He claims "the resolution of the measurements is only about 0.5 meters."  The paper says the shoreline position is accurate to about half a pixel, 15 m for Landsat, and that "The accuracy of the SDS data of ~0.5 pixel... and the study period of ~30 years allows for a classification of beach change rates with class boundaries of 0.5 m/yr."  0.5 m/yr is a rate class boundary derived from 15 m over 30 years, not a 0.5 m resolution; repeating his sentence misstates the paper by a factor of thirty.  What survives is the consequence: the "stable" bin runs from −0.5 to +0.5 m/yr, so a genuine 0.4 m/yr retreat — 400,000 m/Myr — is reported as stable, and the global mean of +0.33 m/yr sits inside that bin.  But each transect rate is a linear regression through up to 33 annual shorelines, not a difference of two images, and "more than 60% of the 2.2 million transects show an uncertainty bandwidth of less than 50% of the linear trend rate."  Validation against nine georeferenced aerial photographs along 63 km of Hatteras Island gave a mean offset of 2.0 m with an RMSE of 17 m.  Recorded discrepancy inside the paper itself: the abstract says 28% accreting, the results text says 27%, and which is correct could not be determined.

National Assessment of Shoreline Change: Historical Shoreline Change along the New England and Mid-Atlantic Coasts

Hapke, Cheryl J.; Himmelstoss, Emily A.; Kratzmann, Meredith G.; List, Jeffrey H.; and Thieler, E. Robert.  USGS Open-File Report 2010-1118  (2010).

https://pubs.usgs.gov/of/2010/1118/pdf/ofr2010-1118_report_508_rev042312.pdf

topic: coastal camp: mainstream type: measurement access: OA timescale: ~150 yr long-term, ~30 yr short-term direction: mixed

Key figures: "The average rate of long-term shoreline change for the New England and Mid-Atlantic coasts was -0.5 meters per year with an uncertainty in the long-term trend of ±0.09 meters per year.  The rate is based on shoreline change rates averaged from 21,184 individual transects, of which 65 percent were eroding."  Short term: "The net short-term rate as averaged along 17,045 transects was -0.3 meters per year," with 60% erosional.  Historic shorelines span the 1800s to the 2000s; the lidar shoreline is 1997 or 2000.  At 0.5 m/yr that is 500,000 m/Myr, or 500 km per million years.

The number that tests the load-bearing input of the coastal argument.  Fleming's East Coast figure of "about a meter a year," derived from lighthouse relocations, is twice the USGS long-term average and more than three times the short-term average.  Worse for the argument, the USGS figure is itself a measurement of the sandy shoreline — barrier islands, spits and inlets — which is the fastest-moving component of the coast and is not continental rock at all.  The report says why New England is slower than the Mid-Atlantic: "New England has a greater percentage of shore types that tend to erode more slowly (rocky coasts, pocket beaches, and mainland beaches), whereas the Mid-Atlantic is dominated by more vulnerable barrier islands and dynamic spit/inlet environments."

Tagged mixed rather than undercuts because half a metre a year is genuinely fast, and the deep-time arithmetic on it is not comfortable for anyone.  What it will not carry is the specific claim built on it, that continental bedrock retreats at a metre a year.

National Assessment of Shoreline Change Part 3: Historical Shoreline Change along Sandy Shorelines of the California Coast

Hapke, Cheryl J.; Reid, David; Richmond, Bruce M.; Ruggiero, Peter; and List, Jeffrey.  USGS Open-File Report 2006-1219  (2006).

https://pubs.usgs.gov/of/2006/1219/of2006-1219.pdf

topic: coastal camp: mainstream type: measurement access: OA timescale: decades to ~150 yr direction: undercuts

Key figures: "The average rate of long-term shoreline change for the State of California was 0.2±0.1 m/yr, an accretional trend.  This is based on shoreline change rates averaged from 14,562 individual transects, of which 40% were eroding."  Coverage is 45% of California's 1,100 km of coast.  Regional long-term and short-term means, m/yr: northern California +0.5 and +0.3, with 23% and 47% of transects eroding; central California 0.0 and −0.5, with 53% and 79% eroding; southern California +0.3 and −0.1, with 36% and 64% eroding.

The most awkward datum in the American record for the "coasts are vanishing" framing, and it is a federal survey rather than a rebuttal written for the occasion.  On the long-term record, California's sandy shoreline advanced.  The short-term record reverses the sign in central and southern California, which is a further caution against treating any measured shoreline rate as a fixed property of a coast.

The scope limit matters and should be granted: this is sandy shoreline only, on a state whose rocky cliffs are measured separately by the same programme at −0.3 m/yr.  A California-wide coastal budget needs both reports, not the convenient one.

National Assessment of Shoreline Change Part 1: Historical Shoreline Changes along the U.S. Gulf of Mexico

Morton, Robert A.; Miller, Tara L.; and Moore, Laura J.  USGS Open-File Report 2004-1043  (2004).

https://pubs.usgs.gov/of/2004/1043/ofr-2004-1043.pdf

topic: coastal camp: mainstream type: measurement access: OA timescale: 1800s to 1998-2002 long-term; 1970s to lidar short-term direction: mixed

Key figures: mean long-term shoreline change by state, m/yr, with the percentage of transects eroding — west Florida −0.1 ± 0.1 (58% eroding, 10,644 transects); Alabama −0.4 ± 0.8 (75%, 1,294); Mississippi −2.3 ± 1.9 (80%, 724); Louisiana −7.1 ± 4.5 (91%, 2,490); Texas −0.7 ± 1.7 (64%, 10,626).  Short-term means: Florida +0.2, Alabama +0.3, Mississippi −2.1, Louisiana −10.1, Texas −0.1.

The largest shoreline retreat rates in the United States are here, and they are the least relevant numbers in the country to a deep-time argument.  Louisiana's −7.1 m/yr is 7,100 km per million years on paper.  What it actually measures is the Mississippi delta compacting and subsiding while its barrier islands disintegrate — Holocene mud settling under its own weight, with the sediment supply that built it now leveed away from the marsh.  No rock is being eroded.  A budget for continental denudation cannot use it, and neither can a budget for coastal cliff retreat.

Included because it is the obvious place a fast headline number comes from, and because the honest handling of it is a good test of whether a writer on either side has read past the summary table.

National Assessment of Shoreline Change Part 2: Historical Shoreline Changes along the U.S. Southeast Atlantic Coast

Morton, Robert A. and Miller, Tara L.  USGS Open-File Report 2005-1401  (2005).

https://pubs.usgs.gov/of/2005/1401/ofr-2005-1401_print.pdf

topic: coastal camp: mainstream type: measurement access: OA timescale: 1800s to 1998-2002 long-term direction: mixed

Key figures: east Florida mean long-term change +0.2 ± 0.6 m/yr, with at least 39% of the Atlantic shoreline eroding long-term and the eroding segments averaging −0.5 m/yr.  Georgia: 65% of the shore stable or accreting, overall mean +1.0 ± 2.7 m/yr, eroding portion averaging −1.5 m/yr, maximum −9.4 ± 4.0 m/yr at Wolf Island.  North Carolina: overall −0.7 ± 1.3 m/yr, about 70% eroding.

Completes the American sandy-shore picture, and it does not run the way a reader of headlines expects.  Two of the three states with extractable means are net accretional over the long-term record, Georgia strongly so, and the state with the fastest reported retreat is North Carolina at −0.7 m/yr, which is still under Fleming's assumed metre.  The uncertainties are large relative to the means — ±2.7 m/yr on Georgia's +1.0 — and any use of these numbers should carry them.

Extraction failure, recorded rather than papered over: the South Carolina state mean could not be read out of the report's summary table, whose formatting did not survive text extraction, and no South Carolina figure should be quoted from this source until the table is read directly.

Sea level and marine terraces

Sea level and global ice volumes from the Last Glacial Maximum to the Holocene

Lambeck, Kurt; Rouby, Hélène; Purcell, Anthony; Sun, Yiying; and Sambridge, Malcolm.  PNAS, 111(43):15296-15303  (2014).  doi 10.1073/pnas.1411762111

topic: coastal, isostasy camp: mainstream type: model access: paywalled timescale: 35 kyr direction: undercuts

Key figures: an inversion of about 1,000 observations covering the past 35,000 years from far-field localities.  Sea level fell to −134 m from 29 to 21 ka BP, with grounded ice about 52 × 10⁶ km³ greater than today.  Deglaciation ran from about 16.5 to 8.2 ka BP at an average 12 m/kyr, with meltwater pulse 1A at 14.5-14.0 ka BP exceeding 40 mm/yr.  Then "a progressive decrease in the rate of rise from 8.2 ka to ∼2.5 ka BP, after which ocean volumes remained nearly constant until the renewed sea-level rise at 100-150 y ago, with no evidence of oscillations exceeding ∼15-20 cm in time intervals ≥200 y from 6 to 0.15 ka BP."

The reason every modern sea cliff is a young feature.  Sea level reached its present position only in the last few thousand years, so any cliff standing today has had at most about 7,000 years of wave attack where it now stands — which is exactly the window Hurst et al. 2016 and Swirad et al. 2020 model, and exactly the assumption behind the platform-width method in Acharya-Chowdhury et al. 2024.  A retreat rate measured at the modern shoreline is a rate for a line the sea has occupied for a few thousand years out of the last hundred thousand.

The horizontal consequence is the part the coastal argument has to answer.  On a shelf sloping at 0.1 degrees — Fleming's own assumed global figure — a 134 m lowstand puts the shoreline about 77 km seaward of today; at 0.25 degrees, 31 km; at 0.5 degrees, 15 km.  That offset is comparable to or larger than the shelf width the argument says is available, so Quaternary wave attack has been smeared across the whole shelf rather than concentrated at one line.  Fleming concedes the point himself on the podcast: most coastal erosion of the last 300,000 years "would have happened down on the shelf and not at the current coastline."  Two disclosures.  Only the Europe PMC abstract record was obtained; the article is not open access and the full text was not read.  And the shelf-offset table above is arithmetic done from the −134 m figure, not a result reported in the paper.

Interpreting Quaternary uplift rates at the Mendocino triple junction, northern California, from uplifted marine terraces

Merritts, Dorothy J. and Bull, William B.  Geology, 17(11):1020-1024  (1989).  doi 10.1130/0091-7613(1989)017<1020:IQURAT>2.3.CO;2

https://www.osti.gov/biblio/6213369

topic: uplift, coastal camp: mainstream type: measurement access: paywalled timescale: 3-81 ka terraces; uplift averaged over ≥330 kyr direction: mixed

Key figures: "Analysis of the altitudinal spacing of 14 flights of marine terraces indicates a spatial pattern of varying uplift rates that agrees with that determined from previously dated terraces for the past 3-81 ka, and temporal changes in uplift rates from <1 m/ka to 3-5 m/ka."  1 m/kyr = 1 mm/yr = 1,000 m/Myr, so this coast rises at 400 to 5,000 m/Myr.  The fastest uplift is 25-43 km south of the Mendocino triple junction.  At Point Delgada, 55 km south, uplift has averaged 1.2 m/kyr for at least 330 kyr.  On the Mendocino coast, more than a million years after the slab passed, uplift has been under 0.4 m/kyr for at least 330 kyr.

Named in the Let's Talk Creation Episode 87 show notes, and it is where the uplift half of the coastal budget is actually measured.  The pattern is the reason it is hard to dismiss: uplift is fastest exactly where the plate-tectonic model puts the slab window, 25-43 km south of the triple junction, and falls off with distance from it in both space and time.  That spatial coherence is not a free parameter fitted to the terraces.

Tagged mixed because of what it cannot do.  Uplift at 0.4-5 m/kyr can keep pace with cliff retreat on an active margin, which is precisely Fleming's Californian target, and cannot help at all on the US Atlantic passive margin, where the argument is strongest and where the mainstream must lean entirely on rate corrections, shifting shorelines and feedbacks.  Access note: only the abstract was obtained, twice and independently, through OpenAlex and OSTI; the GeoScienceWorld full text is blocked.

Coastal tectonics on the eastern margin of the Pacific Rim: late Quaternary sea-level history and uplift rates, Channel Islands National Park, California

Muhs, Daniel R.; Simmons, Kathleen R.; Schumann, R. Randall; Groves, Lindsey T.; DeVogel, Stephen B.; Minor, Scott A.; and Laurel, DeAnna.  Quaternary Science Reviews, 105:209-238  (2014).  doi 10.1016/j.quascirev.2014.09.017

topic: uplift, coastal camp: mainstream type: measurement access: paywalled timescale: 80-120 kyr direction: undercuts

Key figures: marine terraces on San Miguel and Santa Rosa Islands mapped and dated by uranium-series on fossil corals and amino acid geochronology on fossil mollusks.  The second terrace on both islands dates to about 120 ka and the first terrace on Santa Rosa Island to about 80 ka, matching the marine isotope stage 5.5 and 5.1 high stands.  Uplift rates 0.12-0.20 m/kyr = 120-200 m/Myr, similar to rates from neighbouring Santa Cruz Island.

The measurement that shows why the terrace story is expensive to reject.  Three of its legs are independent of each other.  The ages come from uranium-series decay in corals sitting in the terrace deposits, which is not tuned to the terraces.  The sea-level high stands they match, MIS 5.5 and 5.1, come from far-field oxygen isotope and coral records on other continents.  And the uplift rates that fall out are constant and geologically sensible — 0.12-0.20 m/kyr in a strike-slip regime here, against under 0.4 to 5 m/kyr in the Mendocino slab-window regime.  For the terraces to be Flood or immediately post-Flood features, all three legs would have to be wrong in a way that leaves them mutually consistent.

The authors also record a disagreement worth knowing about, since it shows the field is not uniform: their rates are "an order of magnitude lower than a recent study that generated uplift rates from an offshore terrace dating to the Last Glacial period," a difference they attribute to glacial isostatic adjustment effects the earlier work did not model.  Access note: only the USGS Publications Warehouse abstract was obtained; the article is closed and Unpaywall lists no open copy.

The Phanerozoic Record of Global Sea-Level Change

Miller, Kenneth G.; Kominz, Michelle A.; Browning, James V.; Wright, James D.; Mountain, Gregory S.; Katz, Miriam E.; Sugarman, Peter J.; Cramer, Benjamin S.; Christie-Blick, Nicholas; and Pekar, Stephen F.  Science, 310(5752):1293-1298  (2005).  doi 10.1126/science.1116412

topic: coastal camp: mainstream type: compilation access: paywalled timescale: 10^7^-10^8^ yr direction: undercuts

Key figures: "Long-term sea level peaked at 100 +/- 50 meters during the Cretaceous, implying that ocean-crust production rates were much lower than previously inferred."  Sea level tracks oxygen isotope variation, reflecting ice volume on 10⁴-10⁶ year scales, with the 10⁷ year link attributed to temperature.

Adds a second, slower shoreline migration on top of the Quaternary one.  About 100 m of long-term sea-level fall since the Cretaceous means the shoreline has been walking seaward across the shelf through the whole Cenozoic — on a 0.1 degree shelf, roughly 57 km of migration, which is again the same order as shelf width.  Fleming's own reading, that sea level has been falling for tens of millions of years, is consistent with this curve; what he does not draw from it is that a falling sea level moves the line of attack rather than holding it in place.

Only the reconstructed abstract was obtained, so the curve's structure, its error envelope and the backstripping method behind it have not been checked here.  The ±50 m uncertainty on the Cretaceous peak is half the value, and any argument that turns on the exact amplitude is resting on a soft number.

Monte Fleming: the book, the man, and its reception

Chapter 2 Update — pages 11 to 17 of Stories About Earth's History

Fleming, Monte.  Repository of Meaning (the author's own site), February 2022.

https://www.repositoryofmeaning.com/stories-about-earths-history-a-geologists-dissent-from-deep-time-chapter-2-update/

topic: denudation-global camp: creationist type: book access: OA timescale: n/a direction: supports

Key figures: land area 148,428,950 km² and mean land elevation 835 m, giving 123.9 million km³ above sea level; world rivers 15 Gt/yr = 5.6 km³/yr, so the continents level in 22.1 Myr; the Amazon alone at 0.33 km³/yr would take 372 Myr; reservoirs trapping ~50 km³/yr would fill the equivalent in 2.48 Myr; New Guinea's 786,000 km² at 667 m removed in 1.6 Myr; ocean-floor sediment 337 million km³, which is 2.7 times the continental volume above sea level, so 59 Myr of accumulation.

The only part of the book available anywhere.  It is eight photographs of the printed pages, carrying the running heads, the page numbers 11 to 17, and footnotes 5 to 13.  The Wayback Machine never captured it, and pages outside 11 to 17 return 404, so Chapter 1 — which holds the coastal argument and the 78 cm and 20 cm figures — has never been posted.  Every number in the chapter reproduces from its own stated inputs; the arithmetic is not where the problem is.

The chapter's citation apparatus is the finding.  Six of its nine footnotes point at popular web pages — Universe Today, Amazon Waters, WorldData, International Rivers, an Eos research spotlight — rather than at literature.  And the chapter closes with a list headed "Search terms:", nine questions to type into a search engine, plus "(Check Google Earth for the canyons.)"  That is the method, stated openly by the author, and it is more useful for understanding the book than any objection to its conclusions.  One paragraph on page 15 divides a volume of loose reservoir sediment by a volume of dense continental rock; that is the passage Fleming publicly retracts on Let's Talk Creation Episode 88, and it was still standing in this February 2022 update.

Evidence from Geology that Earth is Young

Fleming, Monte.  Repository of Meaning (YouTube), 21 July 2023.  17 min.

https://www.youtube.com/watch?v=zV8AW_0mCpU

topic: coastal camp: creationist type: podcast access: video timescale: n/a direction: supports

Key figures: Pacific coast "about 20 centimeters per year" (00:28); Atlantic "about a meter per year" (01:05); Africa 1.8 m/yr; Britain 0.4 m/yr; and the closing question, "if all of this erosion is happening for millions of years, why do the continents still fit together?"

This is where the book's headline number is sourced, in the author's own voice: "North America East Coast is about a meter per year — that's the last figure, the last figure I pulled off of the top Google search was 1.2, before that it was 0.8, average is about a meter."  The 0.8 is the book's 78 cm.  A companion short, "Coastal Erosion: How Old is the Earth?" (17 October 2022, https://www.youtube.com/watch?v=RPh2bLzVnA0), says the same thing more plainly — "the average erosion rate according to Google at the moment" — and runs a continental-shelf-width clock the podcast does not, dividing an east-coast shelf width of 131.8 km by 1.25 m/yr to get 105,440 years and a west-coast figure of 105,158.7 years, presenting the 281-year agreement as significant.

Two quantities each known to two significant figures cannot yield a quotient good to four, so that agreement is an artefact of rounding rather than a result.  But the videos deserve credit on one count: Fleming flags the provenance and the instability of his own number himself, twice, a year apart.

Low long-term erosion rates and extreme continental stability documented by ancient (U-Th)/He dates

Flowers, Rebecca M.; Bowring, Samuel A.; and Reiners, Peter W.  Geology, 34(11):925-928  (2006).  doi 10.1130/G22670A.1

topic: thermochronology, ancient-surfaces camp: mainstream type: measurement access: paywalled timescale: 1.7 Gyr direction: undercuts

Key figures: long-term integrated unroofing of ≤2.5 µm/yr — that is, ≤2.5 m/Myr — since about 1.7 Ga, on the western Canadian shield.

Footnote 12 of Fleming's Chapter 2, and the one place he engages the primary literature on continental denudation.  He cites it as an example of authors assuming long ages and therefore positing "extremely, impossibly low" erosion rates.  The paper says something different: these rates are far below what earlier thermochronology suggested for continental interiors, "but are within the range of short-term erosional estimates."  The locality is a Precambrian shield, the flattest and most tectonically dead terrain on the planet, so the low rate is the relief-dependence result arriving from a source the young-earth author selected himself.

Creation in-depth: Erosion and the post-Flood boundary

Oard, Michael J.  Journal of Creation, 27(2):62-70  (August 2013).

topic: denudation-global camp: creationist type: review access: paywalled timescale: n/a direction: supports

Key figure: applying strict actualism, "the average depth eroded from the continents in 4,500 years is only 0.275 m."

The most quantitative living creationist treatment of continental erosion, and the only one that answers Glenn Morton's regional-unconformity objection: the kilometres of erosion recorded inside the sedimentary sequence happened during the receding phase of the Flood, not at present rates.  It is also the only creationist engagement with cosmogenic dating this survey located, and it is a refusal rather than a rebuttal: "Some of them are indirect, such as chemical and cosmogenic isotopic methods, which rely on radiometric dating.  Since these methods are based on uniformitarian assumptions, other more scientific methods will be used."

The answer has a cost its author does not price.  If present rates are unrepresentative by three orders of magnitude, they cannot also be used to date the continents, and the erosion argument depends on their being representative.

Helble's critique of creationist research on the Coconino Sandstone fails

Whitmore, John H. and Garner, Paul A.  In Whitmore, J.H. (ed.), Proceedings of the 2025 New Scholars International Conference on Creationism, p. 32.  Cedarville, Ohio  (2025).

topic: sediment-budget camp: creationist type: review access: OA timescale: n/a direction: mixed

A one-page conference abstract, not a paper: no figures, no equations, four references, and described by its authors as "an initial critical response."  Its lead author edited the volume it appears in and wrote the preface; its second author co-hosts Let's Talk Creation.  The New Scholars conference publishes abstracts only.

It is listed here because of what it concedes.  On Helble's sediment-transport arithmetic it contests nothing — not the 42,000 km³, not the 1,600 km front, not the twelve-day window, not the transport rate, and it offers no counter-calculation.  What it says instead is: "Helble is correct that the processes he outlines could not have formed the Coconino during the Flood, but the Flood model requires extrapolations beyond conditions that prevail in the present-day.  If a Flood model is correct, we will never be able to observe the depositional processes that made the Coconino in the laboratory or in nature."  That is an honest position and it should be quoted rather than caricatured — but it concedes that no observable process will do the job, which is difficult to hold alongside an argument built on present rates being a reliable guide to the past.  Its real dispute with Helble is over whether the Coconino is wind-blown or underwater; on that it lists eleven lines of evidence and argues none, promising fuller work that has not appeared.

Stories About Earth's History: A Geologist's Dissent From Deep Time

Fleming, Monte.  Independently published, 121 pp.  ISBN-13 979-8770378399  (12 December 2021).

https://www.amazon.com/dp/B09NGYC67T

topic: denudation-global, coastal camp: creationist type: book access: book timescale: annual gauging direction: supports

Key figures, all taken from reviews that quote the book, because the book's own text was not obtained: Atlantic coast retreating over two feet per year (78 cm/yr), Pacific coast 8 inches per year (20 cm/yr), the two fronts meeting near Salt Lake City in 4.2 Myr; world rivers delivering 15 Gt/yr = 5.6 km³/yr, levelling all land above sea level in 22 Myr; the Amazon alone, given all the land on earth, taking 372 Myr; dissolved load about 4 Gt/yr; reservoirs trapping about 50 km³/yr, enough to bury a continent's worth of rock in 2.5 Myr; Everest rising 50-62 mm/yr against 2 mm/yr of erosion, which over a conventional 40 Myr would put the summit 1,920 km up.

The fullest written statement of the modern young-earth erosion argument, and the source everything in the current exchange traces back to.  Fifteen chapters carry nine arguments, and the last chapter is a Seventh-day Adventist theological appeal that two of the four reviewers found jarring.  The river arithmetic is correct and should be conceded without hedging.  15 Gt/yr ÷ 2.7 t/m³ = 5.556 km³/yr, and 1.180 × 10⁸ km³ of land above sea level ÷ 5.556 km³/yr = 21.2 Myr, against his stated 22.  He is dividing by rock density rather than sediment density, which is the right choice for a denudation argument, and his figure reproduces this project's independent ETOPO calculation.  As a lowering rate it is 37.4 m/Myr (= 37.4 mm/kyr).  The publisher's own framing is the claim to test: the book "contains nine independent arguments from different areas of geology, each of which constrains the age of the earth to a tiny fraction of 4.543 billion years."

Recorded so nobody has to find it later.  The book itself could not be obtained for this catalogue — no Look Inside without a login, no Google Books record, no Open Library record, not on archive.org.  Every number above is second-hand, and three of the four reviews are by non-scientists.  Reported page counts differ across five sources: 109, 120, 121, 125 and 160.  Whether a second edition exists could not be confirmed or excluded.  The companion webpage "Chapter 2 Update" — chapter 2 being the erosion chapter — now returns 404 at repositoryofmeaning.com and the Wayback Machine holds no snapshot of it.  The claim that "Mount Everest can at most be 4.6 Ma," attributed to the book in the Journal of Creation review, appears in no source that could be checked; height divided by 2 mm/yr reproduces it arithmetically, but that is a guess about his reasoning and should not be printed as his.

The Long Ages Aren't Real

Read, David.  Fulcrum7, 27 May 2022.

https://www.fulcrum7.com/apologetics/2022/5/27/the-long-ages-arent-real

topic: denudation-global, coastal camp: creationist type: review access: blog timescale: n/a direction: supports

Key figures, all attributed to the book: Atlantic coast eroding at over two feet per year, Pacific at 8 inches per year; the two coasts meeting near Salt Lake City in 4.2 Myr; rivers depositing 15 billion tons, or 5.6 km³, of the earth's crust into the oceans annually, levelling all land in 22 Myr; the Amazon carrying about 900 million tons per year of sediment and dissolved minerals; reservoirs losing about 1% of capacity per year, 50 km³/yr worldwide, filling with a continent in 2.5 Myr; Everest rising 50-62 mm/yr, eroding at 2 mm/yr, and therefore standing 1,920 km — 1,193 miles — high after 40 Myr at the net 48 mm/yr.

The best available proxy for the book's own numbers, and the reason it is usable is that it quotes the book directly rather than paraphrasing.  Verbatim: "The Atlantic coast is eroding away at the rate of over two feet per year, whereas the Pacific coast is wearing down at a much slower 8 inches per year."  That single sentence settles the units question — the 20 cm and 78 cm figures are metres of horizontal shoreline retreat per year, not totals — and the arithmetic confirms it independently.  The two coasts close at 0.20 + 0.78 = 0.98 m/yr, and the conterminous United States is about 4,180 km wide, so 4.18 × 10⁶ m ÷ 0.98 m/yr = 4.27 Myr.  The Pacific front would travel 853 km inland, and Salt Lake City sits roughly 900-1,100 km from the California coast.  Fleming's numbers hold together internally.

Two disclosures and one warning.  Read is an Adventist lawyer-apologist, not a geologist, and a personal friend of the author: "I met Monte Fleming at the 2015 San Antonio General Conference Session."  His is the only post on Fulcrum7 tagged with Fleming's name.  The warning concerns Everest.  The stated arithmetic is exact — 0.048 m/yr × 40 × 10⁶ yr = 1,920 km — but an uplift rate of 50-62 mm/yr for the Everest massif is roughly an order of magnitude above published geodetic surface-uplift rates and sits very close to the India-Eurasia horizontal convergence rate.  If a horizontal plate rate has been substituted for a vertical one, the whole Everest chapter collapses.  No geodetic source was obtained to test this, so the suspicion is recorded here and not asserted.

A Dissent from Deep Time

Goldstein, Clifford.  Adventist Review, 14 June 2024.

https://adventistreview.org/cliffs-edge/a-dissent-from-deep-time/

topic: denudation-global camp: creationist type: review access: blog timescale: n/a direction: mixed

Key figures: quotes the Kindle edition with page numbers at pp. 11, 18, 21, 75, 78-79 and 83 — the 15 Gt/yr and 22 Myr river figures at p. 18; the ancient-biomolecule list at p. 75, including a full genomic sequence from a horse bone dated 560-780 kyr, termite DNA in amber at 25-30 Myr, and squid-ink biomolecules at 160 Myr; living organisms recovered from ancient rock, salt and ice at pp. 78-79.

The most citable of the four reviews, because it is the only one that gives page numbers, and it is tagged mixed because it hands over the argument's weakest point in the argument's own words.  Fleming, p. 21: "I have seen rare mentions of the topic in scientific literature, but the authors assume that long ages are true, and therefore conclude that past erosion rates were extremely, impossibly low.  We know from actual measurements that this isn't the case."  At p. 11 he says the textbooks he studied for his MS and PhD "never mentioned this issue, and I've never seen a scientific article on the topic."  That is the hinge of the whole dispute.  He concedes the mainstream literature does contain the answer — lower past denudation rates — and dismisses it as question-begging.  But the pre-agricultural rates are measured, not assumed: cosmogenic nuclide inventories in quartz, apatite fission-track exhumation, and dated alluvial stratigraphy each give a background rate independently of any age model.  A claim about one's own reading is also the kind of claim a hostile reader tests in ten seconds.

Goldstein discloses his own limits — "I'm no earth scientist.  I can't testify to the validity of Monte Fleming's arguments" — and he is the editor of the Adult Bible Study Guide, not a geologist.  One internal discrepancy to note: his version of the Everest calculation gives 1,400 miles where Fulcrum7 gives 1,193 miles.  The Fulcrum7 figure reproduces from the stated inputs and Goldstein's does not, so prefer Fulcrum7 wherever the two differ.

The Trouble with Deep Time

Sheffield, Marcus.  ADvindicate, 22 April 2022.

http://advindicate.com/articles/2022/4/22/the-trouble-with-deep-time

topic: denudation-global camp: creationist type: review access: blog timescale: n/a direction: supports

Key figures: the book has "a helpful introduction and fifteen fascinating chapters"; the 4.2 Myr North America claim; a Himalayan claim that current erosion "advances 50 times too fast to accommodate millions of years of erosion"; two rhetorical illustrations — a worm eating 0.25 g/day for 4.5 Gyr consumes enough topsoil to cover 2,007 city lots, and a badger digging 3 ft/min for 4.5 Gyr digs 1.34 trillion miles.

Useful for one thing only: it is the source for the book's structure, and it supplies the "50 times" Himalayan figure that appears nowhere else.  That figure cannot be reconciled with the book's own Everest numbers, which imply a ratio near 8.7× (40 Myr conventional against a 4.6 Myr erosion-limited maximum), so one of the two is a misreport and there is no way to tell which without the book.  Sheffield is a retired Professor of English at Southern Adventist University and says so plainly: "I am not a scientist."  The worm and badger illustrations are rhetoric, not argument, and quoting them would be unfair to the book's better chapters.

The role of wind in sediment removal from potholes in semiarid environments

Fleming, Monte and Brand, Leonard.  Geomorphology, 329:194-205  (2019).  doi 10.1016/j.geomorph.2018.12.025

topic: method camp: creationist type: measurement access: paywalled timescale: n/a direction: background

Key figures: 50 potholes in the Navajo Sandstone on Rock Window Mesa, Chinle Valley, Arizona; a high-resolution DEM plus orthophoto; wind measured inside and above 11 representative potholes; the wind-speed threshold for sand removal in dry conditions about 50 km/h, independent of pothole size; helical vortices several metres tall observed lifting sand out.

Included to settle who Fleming is, because the answer cuts in his favour and the article should say so before a reader discovers it elsewhere.  He is an Assistant Professor in the Department of Earth and Biological Sciences at Loma Linda University, with a PhD from Loma Linda in 2018, an MS in geology there in 2014, and an MA in music from UC Riverside in 2004.  He publishes original field and laboratory data in mainstream journals with co-authors from outside the creationist world: besides this paper, "A 3-D-printed precision sediment feed," Journal of Sedimentary Research 93(8):541-551 (2023), doi 10.2110/jsr.2021.069, describing a stepper-motor flume feeder built for about $300 in materials; and Fleming, Bootsman, Moll and Brand, "Pothole formation on Rock Window Mesa, Chinle valley, AZ," Geomorphology 468:109498 (2025), doi 10.1016/j.geomorph.2024.109498.  He runs a flume at Loma Linda built to vary each depositional parameter independently.

The disclosures belong in the same breath.  Loma Linda is a Seventh-day Adventist institution, Leonard Brand is a long-standing creationist author, and the book closes with an Adventist theological appeal.  None of that touches the geology.  Only abstracts were obtained for all three papers; the 2019 paper is closed access and has been cited once.  The right standard is the obvious one: hold the book's arithmetic to exactly the level his own peer-reviewed papers meet.

The sources Fleming's numbers come from

Chemical Composition of Rivers and Lakes

Livingstone, Daniel A.  Data of Geochemistry, 6th ed., Chapter G.  U.S. Geological Survey Professional Paper 440-G, 64 pp.  (1963).

https://pubs.usgs.gov/pp/0440g/report.pdf

topic: denudation-global camp: mainstream type: measurement access: OA timescale: mid-20th-century river gauging direction: mixed

Key figures: world rivers deliver 3,905,000,000 metric tons of dissolved material to the sea each year (p. G40); mean world river water carries 120 ppm total dissolved solids, of which bicarbonate 58.4, calcium 15.0, silica 13.1, sulfate 11.2, chloride 7.8, sodium 6.3, magnesium 4.1, potassium 2.3, nitrate 1.0 and iron 0.67 (Table 81, p. G41); chemical denudation ranges from 6 long tons per square mile for Australia to 110 for Europe.  At 2.7 t/m³ that flux is 1.446 km³/yr of rock-equivalent volume, or 9.74 m/Myr (= 9.74 mm/kyr) spread over 148.5 million km² of land, levelling the continents in 81.6 Myr on chemical denudation alone.

The source of Fleming's "4 billion tons of dissolved solids per year," and the rounding is faithful — no misquotation.  What matters is what Livingstone does not do.  He states his method himself: "By multiplying the average chemical denudation of the continents by their size, one arrives at a figure of 3,905,000,000 metric tons for the total amount of mineral material carried in solution each year."  That is straight multiplication with no subtraction step of any kind.  A case-insensitive search of the full 64 pages finds the word "cyclic" zero times.  Sea spray appears only in regional discussions of Nova Scotia and Ireland, never as a global deduction.  Pollution appears only as regional commentary on the Cuyahoga and the Ohio Valley, never as a global correction.

So the number is inflated as a measure of rock removal, and the inflation runs in the young-earth argument's favour.  The 120 ppm mean includes 7.8 ppm chloride and much of the sodium and sulfate, which are recycled marine aerosol — salt that fell as rain and is on its way back to the sea, not continental rock being dissolved away.  It also carries mid-century industrial and agricultural loading.  Livingstone is candid about his inputs: "The principal gaps in geochemical data for lakes and rivers are long-term downstream averages for the general composition of large tropical rivers and trace-element analyses for large rivers everywhere."  Any honest treatment applies a stated cyclic-salt correction or cites a later global river-chemistry budget that does, rather than leaning on an uncorrected 1963 total.

Present and Future Losses of Storage in Large Reservoirs Due to Sedimentation: A Country-Wise Global Assessment

Perera, Duminda; Williams, Spencer; and Smakhtin, Vladimir.  Sustainability, 15(1):219  (2023).  doi 10.3390/su15010219

https://uploads.water-energy-food.org/sustainability-15-00219.pdf

topic: sediment-budget camp: mainstream type: model access: OA timescale: decades direction: undercuts

Key figures: 47,403 large dams in 150 countries; initial global storage 6,316 km³ falling to 5,282 by 2022, 5,104 by 2030 and 4,665 km³ by 2050, a 26% loss; "Annually, the globally averaged losses amount to approximately 0.36% of initial global storage capacity," which is 22.7 km³/yr; older literature quoted in the paper puts the range at 0.5-1%/yr, which would be 31.6 to 63.2 km³/yr; Brazil about 0.35%/yr and Asia-Pacific 0.36%/yr.

This is the paper the reservoir chapter cites, and it does not contain the number the chapter uses.  It never states an absolute global km³/yr of trapped sediment — only a percentage.  Fleming's 50-60 km³/yr corresponds to 0.79-0.95% of initial capacity, which is the top of a range Perera and colleagues report from a 1987 World Bank paper and a 2016 secondary study, and 2.2 to 2.6 times the paper's own central estimate.  The Fulcrum7 paraphrase, "Reservoirs lose about 1% of their capacity per year," confirms that 1% is the multiplier in play.  Correcting to the cited paper's own figure turns "everything behind the dams in 2.5 Myr" into 5.2 Myr.  As lowering rates over the world's land area, 50 km³/yr is 337 m/Myr and 22.7 km³/yr is 153 m/Myr.

The density problem is worse than the rate problem, and it is decisive.  A reservoir's storage loss is a volume of deposited sediment at reservoir bulk density — roughly 1.0-1.6 t/m³ for clay, silt and sand with pore water — not a volume of solid rock at 2.7 t/m³.  Taking 1.35 t/m³, Fleming's 50 km³/yr of deposit is 67.5 Gt/yr of solids, or 25.0 km³/yr of parent rock, giving 4.7 Myr rather than 2.4.  Perera's own 22.7 km³/yr becomes 11.4 km³/yr of rock, giving 10.4 Myr.  And there is a cross-check the book appears not to run: 67.5 Gt/yr trapped behind dams is 4.5 times the 15 Gt/yr the same book uses for total river delivery to the world ocean.  Dams intercept material that would mostly have been stored on floodplains, in deltas and in channels without ever reaching the sea, and they are built preferentially on high-yield rivers.  Presenting the dam figure as an independent constraint on continental denudation both double-counts and over-counts.  This is the weakest of the nine arguments and the article should concede it outright.  A podcast concession on this exact point is credited to Fleming elsewhere in this catalogue; the 12 September 2025 video was checked and no verified transcript could be obtained from it, so nothing is quoted here.

National Assessment of Shoreline Change Part 3: Historical Shoreline Change and Associated Coastal Land Loss Along Sandy Shorelines of the California Coast

Hapke, C.J.; Reid, D.; Richmond, B.M.; Ruggiero, P.; and List, J.  U.S. Geological Survey Open-File Report 2006-1219, 79 pp.  (2006).

https://pubs.usgs.gov/of/2006/1219/of2006-1219.pdf

topic: coastal camp: mainstream type: measurement access: OA timescale: decades to ~150 yr direction: undercuts

Key figures: long-term change from 14,562 transects, average net +0.2 m/yr, an accretional trend, with 40% of transects eroding; averaged over the eroding transects alone, −0.2 m/yr.  Short-term change from 16,142 transects, average net −0.2 m/yr with 66% eroding; over the eroding transects alone, −0.8 m/yr.

The most likely ancestor of Fleming's Pacific figure, and if it is the source then the argument has a selection problem at its foundation.  His 8 inches per year is 20.3 cm, matching the erosion-only long-term average of −0.2 m/yr exactly.  But that statistic is computed over the eroding 40% of transects with the accreting 60% thrown away.  The coast-wide number is the opposite sign: "The average net rate of long-term shoreline change for California was 0.2 m/yr, an accretional trend."  California, on the best available survey, is gaining shoreline over the long term, not losing it.  Using an erosion-only subsample as though it were a coast-wide average is exactly the move that turns a real measurement into a misleading one.

Two honest caveats.  Whether this report is in fact Fleming's citation could not be determined, because the book's text was not obtained — the match is on the number and the region, not on a footnote anyone has read.  And this is a report about sandy shorelines, so it is not the whole California coast; the cliffed sections are covered by different products.  The point stands regardless: a horizontal shoreline retreat rate is not a vertical continental denudation rate, and 0.20 m/yr of retreat is 200 km/Myr of horizontal migration, a quantity that cannot be converted to m/Myr of lowering without an assumed shore profile.

The cosmogenic nuclide literature

Dating by Cosmogenic Nuclides

Bierman, Paul R.; Bender, Adrian M.; Christ, Andrew J.; Corbett, Lee B.; Halsted, Christopher T.; Portenga, Eric W.; and Schmidt, Amanda H.  In Encyclopedia of Geology, 2nd ed., Elsevier, pp. 101-115  (2021).

https://www2.oberlin.edu/faculty/aschmidt/papers/Bierman%20et%20al%202021.pdf

topic: method camp: mainstream type: review access: OA timescale: 10^3^-10^6^ yr direction: mixed

Key figures: in-situ ¹⁰Be half-life 1.4 Myr with sea-level high-latitude production of 4-5 atoms g⁻¹ yr⁻¹ in quartz; ²⁶Al 0.7 Myr and 35 atoms g⁻¹ yr⁻¹; in-situ ¹⁴C 5.7 kyr and 18-20; ³⁶Cl 0.3 Myr, 70 in calcite and 200 in K-feldspar.  Himalayan erosion "can be as high as 1 km/My" (1,000 m/Myr), where ¹⁰Be integrates over only a few millennia; central Australia integrates over millions of years; Australian inselbergs erode at under 0.5 m/Myr and "had changed little since the Miocene time, some 20 Ma"; Pildappa Rock on the Eyre Peninsula erodes at 40-50 cm/Myr; mid-latitude bedrock runs at tens of m/Myr; detrital sampling uses the 0.25-1 mm grain fraction and assumes sediment transport times of 10²-10⁴ yr.

The best single source for explaining the method to a lay reader, and the fairest document in this whole section, because its authors list their own failure modes.  Bierman and Portenga wrote the 2011 global compilation; this is the same people twenty years later.  They give the assumptions behind a basin-averaged rate in their own words: uniform quartz distribution across the watershed, well-mixed river sediment, rapid sediment transport from erosion to sampling, and erosion that is gradual rather than pulsed by deep-seated landslides.  Anyone attacking the method should start from that list rather than inventing one.  The published list is misnumbered, jumping from (3) to (5).

Three of its cautions cut against a lazy mainstream reading, and one of them cuts toward the young-earth side.  Individual samples replicate poorly where erosion is dominated by landslides, steep topography, concentrated farming or intense rainfall, even though the dataset as a whole replicates well — a Dominica time series before and after Hurricane Maria in 2017 shows the correlation between erosion rate and steepness changing across the storm.  Outcrop rates are minimum limits by construction: "Bedrock outcrops exist on the landscape because they erode more slowly than the hillslopes and river basins around them."  And in deeply weathered tropical settings the measured rate can be too low, not too high — in Cuba, chemical weathering rates exceed those inferred from ¹⁰Be by an order of magnitude, so there the isotope rate "may only represent minimum estimates of erosion."  That last one means true denudation in some settings is faster than the isotope number, which nudges toward the young-earth conclusion rather than away from it, and it should be reported for that reason.

Time scale bias in erosion rates of glaciated landscapes

Ganti, Vamsi; von Hagke, Christoph; Scherler, Dirk; Lamb, Michael P.; Fischer, Woodward W.; and Avouac, Jean-Philippe.  Science Advances, 2(10):e1600204  (2016).  doi 10.1126/sciadv.1600204

http://geomorph.geog.ucsb.edu/wp-content/uploads/2017/11/ganti-etal-2016c-science-adv.pdf

topic: timescale-bias camp: mainstream type: model access: OA timescale: decades to 10^7^ yr direction: mixed

Key figures: in glaciated landscapes, inverse power-law scaling of rate against averaging window with exponents from −0.71 to −0.11; their stochastic model of heavy-tailed erosional hiatuses gives a mean slope of −0.50 ± 0.004; an exponent of exactly −1 is what pure spurious correlation would produce, so the observed values being shallower than −1 is the result; in fluvially dominated landscapes, no timescale bias detected across decades to millions of years.

The strongest legitimate methodological objection available to a young-earth writer, and it is better raised here than left for a critic to spring.  The Sadler effect is real and well documented: measured rates rise systematically as the measurement window shrinks, so extrapolating a short-window rate backwards overstates the long-term rate.  Ganti and colleagues find exactly that in glaciated terrain and offer it explicitly as an explanation for why ¹⁰Be modern denudation rates run above fission-track and helium exhumation rates.  A creationist writer could take the same result and argue the opposite way — that a rate integrating over 10⁴ years is inflated relative to the true 10⁷-year rate, so the "continents should be flat" arithmetic is even stronger than it looks.

The same paper defuses that move for the terrain that matters.  Verbatim: "In contrast, we find no evidence for a time scale bias in spatially averaged erosion rates of landscapes dominated by river incision."  The cratonic and passive-margin measurements the argument turns on — Australia, Namibia, the southeastern United States — are fluvial, and Portenga & Bierman's global compilation explicitly excluded basins with extensive recent glacial cover.  So the bias is real, bounded, named by the people who found it, and largely outside the datasets in question.  That is the honest summary; it is not "there is no bias."

Spatially Averaged Long-Term Erosion Rates Measured from in Situ-Produced Cosmogenic Nuclides in Alluvial Sediment

Granger, Darryl E.; Kirchner, James W.; and Finkel, Robert.  The Journal of Geology, 104(3):249-257  (1996).  doi 10.1086/629823

topic: cosmogenic, method camp: mainstream type: measurement access: paywalled timescale: 16 kyr direction: undercuts

Key figures: two catchments; the ¹⁰Be-inferred erosion rate compared against measured alluvial-fan deposition over the past 16,000 years; the two independent estimates "agree within one standard error."  Cited 689 times.

The calibration receipt, and the first thing to show a reader who suspects the isotope method is circular.  Bierman & Steig proposed basin-averaging as a hypothesis the same year; Granger, Kirchner and Finkel tested it against a physical pile of sediment whose volume can be measured with a shovel and a survey, and it passed.  In their own words, the agreement demonstrates "that cosmogenic nuclide signatures of stream sediment can be used to measure spatially averaged long-term erosion rates."  The paper also established that long-term erosion rates are an exponential function of average hillslope gradient — the empirical relation that Willenbring and colleagues later extrapolated across every pixel of a global DEM, with the consequences recorded elsewhere in this catalogue.

Two samples is a small test, and only the abstract was obtained; the full text was not read.  The much larger replication is Kirchner et al. 2001, which agreed with debris-fan accumulation within 20% and with apatite fission-track exhumation within a factor of two.

Estimating rates of denudation using cosmogenic isotope abundances in sediment

Bierman, Paul R. and Steig, Eric J.  Earth Surface Processes and Landforms, 21(2):125-139  (1996).  doi 10.1002/(SICI)1096-9837(199602)21:2<125::AID-ESP511>3.0.CO;2-8

topic: method camp: mainstream type: model access: paywalled timescale: 10^3^-10^6^ yr direction: background

Key figures: none — this is the theory paper.  Cited 439 times.

The paper that turned cosmogenic nuclides from an outcrop curiosity into a landscape tool, by proposing that a handful of river sand carries the area-weighted average dose of the whole catchment above it.  It is listed here mainly for its stated conditions, because those conditions are the method's real assumptions and they come from its own authors rather than from a critic.  The approach works "if a basin is in isotopic steady state and if sampled sediments are well mixed."  Note also that the authors offered it "as a testable hypothesis" and that it was in fact tested, twice, within five years.  Abstract only obtained; the full text was not read.

Global analysis of the stream power law parameters based on worldwide 10Be denudation rates

Harel, Maud-Alix; Mudd, Simon M.; and Attal, Mikaël.  Geomorphology, 268:184-196  (2016).  doi 10.1016/j.geomorph.2016.05.035

https://www.pure.ed.ac.uk/ws/files/25670599/25669129._Harel_et_al_2016_author_accepted_manuscript_002_.pdf

topic: cosmogenic camp: mainstream type: compilation access: OA timescale: 10^3^-10^6^ yr direction: background

Key figures: 1,457 basin-averaged ¹⁰Be denudation rates from 77 publications in 59 study zones, every one recalculated on a single scheme; sea-level high-latitude ¹⁰Be production 4.3 atoms g⁻¹ yr⁻¹, revised down from Stone's 5.1 — a 19% cut; ¹⁰Be half-life 1.387 Myr, revised from the 1.5 Myr used in older work including Kirchner et al. 2001; Lal/Stone latitude-altitude scaling; Braucher et al. four-exponential muon production; global concavity m/n = 0.51 ± 0.12; stream-power slope exponent n greater than 1 on average.

The third of the three global ¹⁰Be compilations, after von Blanckenburg 2005 and Portenga & Bierman 2011, and the cleanest published statement of the constants that everything else rests on.  Its value to a skeptical reader is its candour about why every serious compilation recalculates rates from raw isotope concentrations rather than reusing published numbers: "Comparison of denudation rates from 10Be across multiple studies is a challenge because a wide range of methods are used to calculate these rates."  Five production schemes are in common use, shielding conventions differ between labs, and production by muons is, in their words, highly uncertain.  The practical consequence is blunt — any pre-2010 published ¹⁰Be erosion rate quoted without recalculation is a number that has since moved by roughly a fifth, and that includes rates quoted second-hand in review articles on both sides of this argument.

The paper's own purpose is different from this catalogue's: it fits the stream-power incision law by channel-profile analysis for 59 regions.  Its analysis scripts are public.

OCTOPUS database (v.2)

Codilean, Alexandru T.; Munack, Henry; Saktura, Wanchese M.; Cohen, Tim J.; Raczkowski, Wojciech; and others.  Earth System Science Data, 14:3695-3713  (2022).  doi 10.5194/essd-14-3695-2022

https://octopusdata.org

topic: cosmogenic, method camp: mainstream type: compilation access: OA timescale: 10^3^-10^6^ yr direction: background

Key figures: 4,425 in-situ ¹⁰Be data points and 403 ²⁶Al data points in the CRN Denudation collection, 1,083 and 164 of them added in version 2; the Global sub-collection holds "about 77% of all published 10Be data (75% of publications)"; a partner glacial-exposure collection of 16,009 observations, 2,229 with both ¹⁰Be and ²⁶Al; mean sample number per contributing study about 20 (version 1, Codilean et al., Earth System Science Data, 10(4):2123-2139, 2018).

The reason nobody has to take any cosmogenic number in this catalogue on trust.  OCTOPUS ships not just the isotope concentrations but the entire computational chain — basin outlines, DEM, gradient raster, flow-direction and flow-accumulation rasters, atmospheric pressure raster, production-scaling and topographic-shielding rasters — and recalculates every rate with CAIRN, "an automated, open-source method for calculating basin-averaged denudation rates so that inferred denudation rates are reproducible."  The web interface was live on 26 July 2026 and there is an OGC feature service at geoserver.octopusdata.org.

The authors state the sampling skew themselves rather than burying it: most measurements come from Northern Hemisphere basins "clustering around distinct, mostly tectonically active regions, such as the Pacific coast of the United States, the Appalachians, the European Alps."  That is a real bias and it inflates the global mean.  It is also an open invitation.  A craton-only median denudation rate — the statistic the young-earth argument actually needs, and one nobody in the literature appears to have published — is computable from this database by anyone willing to filter on a shield polygon or on peak ground acceleration.  The data being free removes any excuse on either side for not engaging them.

Earth is (mostly) flat: Apportionment of the flux of continental sediment over millennial time scales

Willenbring, J.K.; Codilean, A.T.; and McElroy, B.  Geology, 41(3):343-346  (2013).  doi 10.1130/G33918.1

topic: cosmogenic camp: mainstream type: model access: paywalled timescale: 10^3^-10^6^ yr direction: mixed

Key figures, all of them superseded: 990 calibration basins; fitted relation D = 11.9·exp(0.0065 · slope), D in m/Myr and slope in m/km, from 90 m SRTM basin slopes extrapolated on a 1 km global DEM; total continental denudation flux 5.5 Gt/yr, of which 4.4 Gt/yr reaches the ocean; land gentler than 20 m/km is 67% of the surface and supplies 54% of the flux; the steepest 50%, 20% and 10% supply 60%, 33% and 22%; median calibration basin 27 km².

Recorded so that nobody cites it by accident, because its numbers would help the old-earth side enormously if they stood and they do not stand.  A global denudation flux of 4.4 Gt/yr reaching the ocean would make the "continents planed flat" arithmetic run to roughly 87 Myr instead of 14, and the low-slope apportionment would say that mountains barely matter.  The same authors corrected the flux upward by a factor of five the following year and inverted the apportionment result, tracing the error to a DEM-resolution mismatch and a log-transformation bias.  The corrected paper is catalogued separately and is the one to cite.

Only the abstract was obtained, and the indexed copy of that abstract has lost its less-than signs, so one sentence in it reads nonsensically — every figure above is quoted through the 2014 reanalysis's description of the original, not from the original itself.  A published Comment by Warrick and colleagues, Geology 42(1):e316, doi 10.1130/G34846C.1, and the authors' Reply at 42(1):e317, doi 10.1130/G35326Y.1, both exist and neither could be obtained.

Slow Rates of Rock Surface Erosion and Sediment Production across the Namib Desert and Escarpment, Southern Africa

Bierman, Paul R. and Caffee, Marc.  American Journal of Science, 301(4-5):326-358  (2001).  doi 10.2475/ajs.301.4-5.326

topic: ancient-surfaces, cosmogenic camp: mainstream type: measurement access: paywalled timescale: 10^5^-10^6^ yr direction: undercuts

Key figures: 66 paired ¹⁰Be/²⁶Al analyses; 47 bedrock samples on three transects from coast to escarpment to highlands, plus 3 small-stream samples, 4 large-river samples and 12 desert-surface clasts.  Bedrock inland of the escarpment 3.2 ± 1.5 m/Myr (n = 9); seaward of it 3.6 ± 1.9 m/Myr (n = 38); statistically indistinguishable.  Basin rates 16 m/Myr for an escarpment basin, 5 for the highlands, 8 for the coastal plain.  Four large rivers 3-9 m/Myr, averaged over 10⁵ yr and 10⁴-10⁵ km².  Desert-pavement clasts with simple ¹⁰Be model ages to 1.8 Myr and minimum total exposure-plus-burial histories exceeding 2.7 Myr.  Rainfall across the transect ranges from under 25 mm/yr at the coast to over 400 mm/yr inland with no detectable effect on the rate.

A second continent, a different tectonic setting, and the same answer as Australia: stable-interior bedrock erodes at single-digit m/Myr.  At 3.4 m/Myr a 100 m hill lasts 29 Myr and 840 m of mean continental elevation takes 247 Myr to remove even before isostasy is considered.  The paired-nuclide result is the part that does not depend on any assumed erosion rate at all — individual desert pebbles carry ²⁶Al/¹⁰Be ratios demanding more than 2.7 million years of combined surface exposure and burial.  The authors also report a clean negative result: "the Namibian cosmogenic data do not support the model of significant and on-going escarpment retreat."

A discrepancy to record before someone else finds it.  Bierman & Nichols (2004) attributes 40-90 cm/Myr for hyperarid Namibian gneiss and quartz to this paper, but this paper's own bedrock averages are 3.2-3.6 m/Myr — nearly an order of magnitude apart.  The lower figure is presumably a slowest-lithology subset rather than the study average.  Do not quote 0.4-0.9 m/Myr as "the Namibian rate."  Only the abstract of the 2001 paper was obtained; the body text was not read, so the discrepancy is unresolved here.

Cosmic ray labeling of erosion surfaces: in situ nuclide production rates and erosion models

Lal, Devendra.  Earth and Planetary Science Letters, 104(2-4):424-439  (1991).  doi 10.1016/0012-821X(91)90220-C

topic: method camp: mainstream type: model access: paywalled timescale: n/a direction: background

Key figures: 2,642 citations as of July 2026 — by a wide margin the most-cited work anywhere in this catalogue.

The theoretical foundation of everything in this section.  Lal supplies the latitude- and altitude-dependent scaling of cosmogenic production rates, still called the Lal/Stone scheme after Stone's 2000 revision, and the steady-erosion model that converts a measured isotope concentration into an erosion rate.  A skeptic asking whether the method has a physics basis or is curve-fitting lands here, which is why it is listed.

The text was not obtained and nothing here is quoted from it.  What it does is described only from the way the papers that were obtained use it.  Someone should get the actual paper.  The technique's first demonstration is older still and is also unread: Nishiizumi, K.; Lal, D.; Klein, J.; Middleton, R.; and Arnold, J.R., "Production of 10Be and 26Al by cosmic rays in terrestrial quartz in situ and implications for erosion rates," Nature, 319(6049):134-136 (1986), doi 10.1038/319134a0, reporting a handful of samples from the Mojave Desert.  The dates matter for framing the argument fairly: this is a forty-year measurement programme with thousands of published analyses, not a recent claim invented to answer creationists.

The control mechanisms of erosion and weathering at basin scale from cosmogenic nuclides in river sediment

von Blanckenburg, Friedhelm.  Earth and Planetary Science Letters, 237(3-4):462-479  (2005).  doi 10.1016/j.epsl.2005.06.030.  Erratum doi 10.1016/j.epsl.2005.10.007; corrected republication EPSL 242:224-239, doi 10.1016/j.epsl.2005.11.017

topic: cosmogenic camp: mainstream type: review access: paywalled timescale: 10^3^-10^6^ yr direction: background

Key figures: none.  The text was not obtained and no number in this catalogue comes from it.  Cited 637 times.

Listed because of what it is and because of what could not be done with it.  Harel and colleagues name it as one of only three global-scale ¹⁰Be compilations, alongside Portenga & Bierman 2011 and Willenbring et al. 2013, and Portenga & Bierman cite it for the tectonic control on basin denudation rates.  It is closed access with no repository copy anywhere, the publisher has elided the abstract, and every retrieval route tried returned an error.  This is the largest single gap in the cosmogenic material assembled here, and any figure attributed to it in a secondary source should be treated as unchecked until someone reads the paper.

Two cautions for whoever gets it.  Cite the 2006 corrected republication rather than the 2005 original, and check which of the two any secondary source is quoting, because they are separate DOIs with separate page ranges and the correction is easy to miss.

Geology and Young Earth Creationism: The Misguided Erosion Argument

Morton, Glenn R.  Morton YEC archive blog, April 2016.

https://morton-yec-archive.blogspot.com/2016/04/the-misguided-erosion-argument.html

topic: denudation-global, isostasy camp: counter-apologetics type: blog access: blog timescale: n/a direction: undercuts

Key figures: restates the Nevins calculation as 383 × 10¹⁵ t ÷ 27.5 Gt/yr = 14 Myr; Gulf of Mexico sediment volume 2.41 × 10¹⁶ m³ accumulating at 1.75 × 10⁸ m³/yr, taking 137 Myr to fill; Nile canyon fill 2 × 10¹⁴ m³ at 5.05 × 10⁷ m³/yr, taking 3.96 Myr; isostatic rebound taken as "about a third of the extra height," so 3,692 m of erosion yields about 2,400 m of net elevation loss.

The same Glenn Morton who called continental denudation "the best geological argument" for a young earth at the 1986 creationism conference, thirty years later and on the other side.  That trajectory is worth stating openly rather than letting a reader discover it: he is a former young-earth creationist and an oil-industry geophysicist, and his change of position is part of why the piece is worth reading.  His Gulf of Mexico basin-filling calculation is the most useful thing in it, because it approaches the sediment budget from the receiving end rather than the eroding end.

Two reasons not to lean on it.  It contains no cosmogenic nuclide data at all, so it engages none of the strongest evidence bearing on the question.  And its isostasy fraction does not match the Airy value — a crust-to-mantle density ratio near 2.7/3.3 gives 0.82 of rebound and 0.18 m of net lowering per metre of rock removed, a multiplier near 5.5, not the roughly 3 his "about a third" implies.  Redo his arithmetic before citing any of it.

Five Global Evidences for a Young Earth

Institute for Creation Research.

https://www.icr.org/article/five-global-evidences-for-a-young-earth

topic: denudation-global camp: creationist type: popular access: blog timescale: n/a direction: supports

Key figures: quotes Dott, R.H. Jr. and Batten, R.L., Evolution of the Earth, McGraw-Hill, p. 136 (1971) — "North America is being denuded at a rate that could level it in a mere 10 million years, or, to put it another way, at the same rate, ten North Americas could have been eroded since middle Cretaceous time 100 m.y. ago."  Cites Nevins 1973 for the 14 Myr planation figure.  It supplies no tonnage of its own and runs no calculation of its own.

The form in which most readers meet the argument, and a fair sample of what is wrong with the genre.  Its strongest sentence is a quotation from a secular 1971 textbook, and that is a legitimate move — the observation is not a creationist invention, and a skeptic who assumes it is has not read the source.  What follows is the problem.  The article engages nothing published after 1985, and in particular nothing from the forty years of cosmogenic nuclide measurement that bears directly on the claim it is making.

A finding that belongs on this page rather than buried.  Repeated targeted searches of Journal of Creation, Answers Research Journal and Creation Research Society Quarterly for cosmogenic-nuclide denudation content returned nothing at all — not a rebuttal, not a critique of the method, not a citation.  That is a search result and not a proof of absence; CRSQ in particular is poorly indexed by general search engines, and its OAI-PMH archive should be queried directly before the claim is made in print.  But so far as could be determined, the single strongest dataset bearing on the erosion argument has never been addressed by the side that makes the argument.  That is a weakness in the young-earth case, not a point in its favour, and pretending otherwise would defeat the purpose of this catalogue.

Ancient landforms: the antiquity argument and where it comes from

On the Survival of Paleoforms

Twidale, C.R.  American Journal of Science, 276(1):77-95  (1976).  doi 10.2475/ajs.276.1.77

https://ajsonline.org/article/59816

topic: ancient-surfaces camp: mainstream type: review access: OA timescale: 10^6^-10^8^ yr direction: mixed

Key figures: the silcreted land surface of central Australia given "perhaps 20 m.y." of exposure; the laterite surface of the South Australian Gulfs region given "some 200 m.y. of epigene attack."  Rates he compiles, converted to m/Myr: Arcoona Plateau valley widening 0.15 mm/yr = 150 m/Myr lateral; southern Mount Lofty laterite stripping 1 cm per 200 yr = 50 m/Myr; Grand Canyon scarp retreat 0.06 cm/yr = 600 m/Myr; Colorado sandstone cliff rainwash 0.005 mm/yr = 5 m/Myr; Oahu 40-50° slopes ~8 cm/yr = ~80,000 m/Myr; New Guinea slopes 1 cm in 7-10 yr = 1,000-1,430 m/Myr, and 1 cm in 44 yr = 227 m/Myr; central Japan catchment 0.1-10 cm/century = 10-1,000 m/Myr; Shaur Anticline gorge, western Iran, 4 m in 1,200 yr = 3,333 m/Myr of channel incision.

The source of every "even mainstream geologists admit it" quote in the landform-antiquity argument, and the quotes are real.  Twidale writes that "there has surely been ample time for the very ancient features preserved in the present landscape to have been eradicated several times over," and then, opening his section on landscape-evolution models, that "the survival of these paleoforms is in some degree an embarrassment to all of the commonly accepted models of landscape development."  Neither line is torn out of context.  Both were checked against the original Scholastica scan rather than against a creationist quotation of them.

The same paper supplies its own answer twice over, and any use of it that omits both is quote-mining.  First, on pages 77-78, Twidale sets aside the exhumed Precambrian surfaces of Canada, Greenland, India and the Great Artesian Basin: "they have been preserved as a result of burial and have only recently been reexposed to subaerial attack.  Their survival thus poses few problems."  Second, on page 80, he classifies most of his remaining puzzling cases as etch surfaces — "the present surface represents either an exposed indurated horizon of the weathering profile, or the weathering front."  A laterite "surface" is the hard floor of a rotted soil profile, not the ground people once walked on.  He then lists five long-age survival mechanisms and defends each with field data: aridity, repeated uplift, resistant structure, Crickmay's unequal activity of rivers, and armouring feedbacks.  His own Iranian and New Guinean numbers make the point better than his prose does — a gorge can cut at 3,333 m/Myr while the interfluve two kilometres away loses almost nothing, which is an argument against applying a continental average to any particular hilltop.

Tagged mixed, leaning undercuts.  Twidale works entirely inside the conventional timescale and nowhere questions it; his conclusion is only that "the hills are not everlasting… but they persist for much longer periods than has been generally conceded."  One caveat on verification: the open-access PDF is a scan with no text layer, so the text behind these quotes was obtained by local OCR.  The quoted sentences were read back character by character; unquoted paraphrase from this paper carries a small residual OCR risk.

Geological Changes and Time

Roth, Ariel A.  Origins (Geoscience Research Institute), 3(2):106-108  (1976).

https://www.grisda.org/origins-03106

topic: ancient-surfaces camp: creationist type: review access: OA timescale: n/a direction: supports

Key figures: Gage 1970 reports about 0.1 inch/yr of erosion in steep western Southern Alps catchments = 2.54 mm/yr = 2,540 m/Myr, or 2.54 km per million years.

The two-page review that put Twidale into the creationist literature, and it is more careful than its descendants.  Roth quotes the "ample time… eradicated several times over" passage and the "embarrassment" sentence accurately, and he does not claim Twidale supports a young earth.  He states the opposite: "This is a concept which Twidale and most other geomorphologists might vigorously reject."  He also concedes the weakness of his own case in the same breath — "Unfortunately the proposed explanations given are not quantitatively evaluated and are not very useful in answering the time questions raised in this review."  That is a fair complaint about 1976 geomorphology and it has since been answered by cosmogenic and thermochronological measurement, which is the point of the next section.

Two flags.  Gage 1970 was read only through Roth; the Southern Alps figure has not been checked against the Journal of Geology original.  And a conversion error is worth recording because it is easy to make: 2.54 mm/yr is 2,540 m/Myr, not 2,540,000 — a rate that fast would strip a kilometre of rock in 400,000 years.

The age of the lateritized summit surface on Kangaroo Island and adjacent areas of South Australia

Daily, B.; Twidale, C.R.; and Milnes, A.R.  Journal of the Geological Society of Australia, 21(4):387-392  (1974).  doi 10.1080/00167617408728860

https://api.openalex.org/works/doi:10.1080/00167617408728860

topic: ancient-surfaces camp: mainstream type: measurement access: paywalled timescale: 10^8^ yr direction: mixed

Key figures: the Kangaroo Island laterite is bracketed as older than Middle Jurassic and younger than Early Permian, with the Triassic — about 200 Ma — the preferred age; the island is about 140 by 60 km; Roth's inference from it is that at 3 cm/1000 yr = 30 m/Myr the island should have lost 6,000 m in 200 Myr.

The primary source under the most-repeated version of the argument, and it does not date what the argument needs it to date.  The abstract is explicit that the dated object is the laterite — a weathering profile bracketed stratigraphically — and that "palaeoclimatic and palaeontological considerations suggest the Triassic as the most likely age."  That is an inference, not a measurement, and Twidale himself classified exactly these South Australian summit surfaces as etch type two years later.  What is exposed on Kangaroo Island today is the indurated horizon that sat below the Triassic ground surface, not the Triassic ground surface.

Roth's arithmetic is correct and his input rate is not.  Thirty m/Myr is a continental-average sediment-yield figure applied to one of the most stable cratonic interfluves on the planet.  At the rates later measured on the neighbouring Eyre Peninsula, 0.3-5.7 m/Myr, 200 Myr of the same process removes 60 to 1,140 m rather than 6,000 m.  That is still a great deal of rock, which is precisely why the mainstream reading is progressive lowering through a thick regolith rather than a frozen Triassic plain.  Only the publisher abstract was obtained; the full paper was not fetched, and no modern re-dating of this surface was found.

Antiquity of landforms: objective evidence that dating methods are wrong

Oard, Michael J.  CEN Technical Journal (now Journal of Creation), 14(1):35-39  (2000).

https://creation.com/antiquity-of-landforms

topic: ancient-surfaces camp: creationist type: review access: OA timescale: n/a direction: supports

Key figures: continents levelled "in roughly 10 to 20 million years… or possibly up to 33 million years"; a claimed 10 m of Shoalhaven gorge-wall retreat in 30 Myr.

The form the argument travels in today, written by a creationist meteorologist for a creationist journal.  Oard's thesis is stated plainly: "the existence of these old erosion surfaces is more an indictment of uniformitarian dating methods."  Credit where it is due — he flags the exhumation distinction himself, restricting his claim to surfaces "which were not exhumed from beneath other rocks," and he reports honestly that Twidale's preferred answer, Crickmay's unequal activity, "only diminishes the problem without solving it."

Three problems a hostile reader will find, recorded here so nobody has to find them later.  The 10 m in 30 Myr figure for the Shoalhaven could not be verified against the source it is attributed to; the verified Shoalhaven numbers are ratios, not a retreat distance.  The article places "western Arnhem Land" in Queensland, and Arnhem Land is in the Northern Territory.  And the levelling times omit the isostatic term entirely, which is the correction that moves the answer from tens of millions of years to hundreds.

Antiquity of landforms: an "extremely unlikely" concept vindicated

Twidale, C.R.  Australian Journal of Earth Sciences, 45(5):657-668  (1998).  doi 10.1080/08120099808728422

topic: ancient-surfaces camp: mainstream type: review access: paywalled timescale: n/a direction: mixed

Listed because it is cited constantly and because its text was not obtained.  Taylor & Francis returns 403 to every route tried, Unpaywall reports it closed, and both Semantic Scholar and OpenAlex have the abstract elided.  Title, author, journal, volume, issue, pages, year and DOI are confirmed; nothing else is.  The sentence usually quoted from it — "compelling evidence adduced pointing to their great antiquity, not only in Australia and Africa but also, and in lesser measure, in the Americas and Europe" — reaches this page only through Oard 2000's transcription, and must be cited as "quoted in Oard 2000" unless the original is obtained.  Twidale's own 1998 resolution of the paradox is therefore unknown here.  His near-contemporary "Landforms Ancient and Recent: the Paradox," Geografiska Annaler A 81(3):431-441, doi 10.1111/j.0435-3676.1999.00072.x, is likewise paywalled with no abstract available anywhere.

Quantitative resolution of the debate over antiquity of the central Australian landscape

Belton, D.X.; Brown, R.W.; Kohn, B.P.; Fink, D.; and Farley, K.A.  Earth and Planetary Science Letters, 219(1-2):21-34  (2004).  doi 10.1016/S0012-821X(03)00705-2

topic: ancient-surfaces camp: mainstream type: measurement access: paywalled timescale: unknown direction: background

The largest gap in this section, listed rather than hidden.  This is the paper aimed most directly at the Australian landform-antiquity claim, and no abstract could be obtained from Crossref, OpenAlex, Semantic Scholar or Europe PMC; ScienceDirect returns 403.  The citation is verified and the content is not.  Nothing in this article rests on it, and nothing should be attributed to it until a copy is read.

How fast the "unchanging" surfaces actually erode

Cosmogenic exposure and erosion history of Australian bedrock landforms

Bierman, Paul R. and Caffee, Marc.  GSA Bulletin, 114(7):787-803  (2002).  doi 10.1130/0016-7606(2002)114<0787:CEAEHO>2.0.CO;2

https://api.openalex.org/works/doi:10.1130/0016-7606(2002)114%3C0787:CEAEHO%3E2.0.CO;2

topic: ancient-surfaces, cosmogenic camp: mainstream type: measurement access: paywalled timescale: 10^5^-10^6^ yr direction: undercuts

Key figures: 10 sites on a 2,000 km transect from the Eyre Peninsula at 33°S to Kakadu at 14°S; 61 exposed samples plus 7 shielded samples from 3.5-13 m depth; maximum limiting model erosion rates 0.3 ± 0.1 to 5.7 ± 1.0 m/Myr; minimum limiting model exposure ages 105 ± 16 to 1310 ± 190 ka; ²⁶Al/¹⁰Be ratios 3.89 ± 0.19 to 6.29 ± 0.82; assumed sea-level high-latitude ¹⁰Be production 5.17 atoms g⁻¹ yr⁻¹.

The single most important measurement on this question, and it cuts both ways at once — which is why it belongs to nobody's talking points.  It confirms the creationist's best fact: "exceptionally high activities of in situ-produced 10Be and 26Al indicate that Australian bedrock surfaces are some of the most stable in the world."  At 0.3 m/Myr a hundred-metre inselberg survives 333 million years of subaerial attack.

Then the isotopes settle the question the argument actually turns on.  "Nuclide activities are not saturated in any of the 61 exposed samples we analyzed, mandating that the surfaces we sampled are losing mass, albeit slowly, over time.  Such mass loss indicates that inselbergs are dynamic landforms, confounding the assignment of discreet ages to inselberg surfaces."  Saturation is what a surface that had stopped eroding would look like, and no sample shows it.  These landforms are old in form while being continuously lowered.  The authors are willing to say the domes "may well be direct descendents of Cenozoic and perhaps even Mesozoic forms," and they also report that stability varies inversely with mean annual precipitation, which is the aridity mechanism Twidale proposed in 1976 turned into a measurement.  Only the publisher abstract was obtained; GeoScienceWorld returns 403 and the full text was not read.

¹⁰Be and ²⁶Al Evidence for Exceptionally Low Rates of Australian Bedrock Erosion and the Likely Existence of Pre-Pleistocene Landscapes

Bierman, Paul R. and Turner, J.  Quaternary Research, 44(3):378-382  (1995).  doi 10.1006/qres.1995.1082

https://api.openalex.org/works/doi:10.1006/qres.1995.1082

topic: ancient-surfaces, cosmogenic camp: mainstream type: measurement access: paywalled timescale: ≥5 × 10^5^ yr direction: undercuts

Key figures: 20 isotopic analyses of 10 samples from granite domes on the Eyre Peninsula — Mt Wudinna, Little Wudinna and Yarwondutta Rock; erosion "at rates as low as 0.6 m/myr"; mean model rate less than 0.7 ± 0.1 m/Myr (n = 6); minimum exposure at least 0.5 Myr.

The pilot study behind the 2002 paper, and useful for one extra observation the larger dataset repeats: some samples do not fit a simple history.  "Samples collected from the flanks of some domes and the tops of other domes have isotopic abundances suggestive of complex exposure histories including burial and reexposure."  Burial and re-exposure is the mechanism the whole antiquity debate turns on, and it is visible in the isotope ratios themselves rather than being assumed.  Abstract only; the full text was not obtained.

Quantifying regolith erosion rates with cosmogenic nuclides ¹⁰Be and ²⁶Al in the McMurdo Dry Valleys, Antarctica

Morgan, D.J.; Putkonen, J.; Balco, G.; and Stone, J.  Journal of Geophysical Research: Earth Surface, 115:F03037  (2010).  doi 10.1029/2009JF001443

https://api.openalex.org/works/doi:10.1029/2009JF001443

topic: ancient-surfaces, cosmogenic camp: mainstream type: measurement access: paywalled timescale: 10^6^ yr direction: undercuts

Key figures: measured erosion rates 0.1-4 m/Myr; Arena Valley ash dated by ⁴⁰Ar/³⁹Ar to 4.33 Ma with overlying material eroding at 3.5 ± 0.41 × 10⁻⁵ g cm⁻² yr⁻¹, about 0.19 m/Myr averaged over ~4 Myr; Hart ash dated by K-Ar to 3.9 Ma with erosion at 4.8 ± 0.21 × 10⁻⁴ g cm⁻² yr⁻¹, about 2.6 m/Myr over ~1 Myr.

The cleanest statement anywhere of the mainstream resolution, and it names the puzzle before answering it.  "The McMurdo Dry Valleys, Antarctica (MDV) are among the oldest landscapes on Earth, and some landforms there present an intriguing apparent contradiction such that millions of years old surface deposits maintain their meter-scale morphology despite the fact that measured erosion rates are 0.1-4 m Ma⁻¹."  That is a mainstream author conceding, in a mainstream journal, that the young-earth objection points at something real.

The answer occupies one sentence: "These results indicate that the slow, steady lowering of the surface without vertical mixing may allow landforms to maintain their meter-scale morphology even though they are actively eroding."  A metre-scale ridge survives not because nothing is removed but because everything is removed evenly and nothing stirs the deposit.  The independent ash dates are what make this a measurement rather than a story — the ⁴⁰Ar/³⁹Ar and K-Ar ages come from different minerals and different physics than the cosmogenic rate they are paired with.  Abstract only; the bronze open-access PDF at Wiley was not retrieved.

Cosmogenic isotope data support previous evidence of extremely low rates of denudation in the Dry Valleys region, southern Victoria Land, Antarctica

Summerfield, M.A.; Sugden, D.E.; Denton, G.H.; Marchant, D.R.; Cockburn, H.A.P.; and Stuart, F.M.  Geological Society, London, Special Publications, 162:255-267  (1999).  doi 10.1144/GSL.SP.1999.162.01.20

https://api.openalex.org/works/doi:10.1144/GSL.SP.1999.162.01.20

topic: ancient-surfaces, cosmogenic camp: mainstream type: measurement access: paywalled timescale: 10^6^-10^7^ yr direction: undercuts

Key figures: maximum denudation on rectilinear slopes "may be only c. 1 m Ma⁻¹," falling to <0.2 m/Myr on low-relief surfaces above about 1,500 m in the western Dry Valleys; higher rates near the Ross Sea coast.

The floor value for subaerial erosion in a cold desert, and the source of the claim that "landforms with a relief of a only a few metres have survived with minimal modification since the mid-Miocene."  That sentence is the strongest single statement of landform antiquity in the verified literature here, and it is worth quoting for the young-earth side rather than against it — before pairing it with Morgan et al., who measured the same landscape and found it losing mass throughout.  Abstract only.  The companion ²¹Ne paper in Geomorphology 27(1-2):113-129, doi 10.1016/S0169-555X(98)00093-2, was not fetched and nothing is quoted from it.

Long-Term Landscape Evolution: Early Miocene and Modern Rivers in Southern New South Wales, Australia

Young, R.W. and McDougall, I.  The Journal of Geology, 101(1):35-49  (1993).  doi 10.1086/648195

https://api.openalex.org/works/doi:10.1086/648195

topic: ancient-surfaces camp: mainstream type: measurement access: paywalled timescale: since the Early Miocene, ~20 Myr direction: undercuts

Key figures: plateau surfaces lowered by "only about 2 to 5 m/Ma"; major streams incised "only about 5 to 18 m/Ma, with a maximum incision of 30 m/Ma," all since the Early Miocene.

The measurement that turns Twidale's unequal-activity hypothesis into a number, using dated basalts that fill and cap old valleys as the datum.  Divide lowering and channel incision differ by a factor of two to fifteen within one landscape.  That is the direct answer to using a continental-average denudation rate on a hilltop: the hilltop is not eroding at the average and never was.  The authors also note that "despite the passage of 20 Ma, stream gradients have not waned, nor have major breaks in profiles been eliminated" — landscapes do not march smoothly toward a peneplain, which is the assumption the levelling arithmetic requires.  Abstract only.

Wearing Down, Wearing Back, and Gorge Extension in the Long-Term Denudation of a Highland Mass: Quantitative Evidence from the Shoalhaven Catchment, Southeast Australia

Nott, J.; Young, R.; and McDougall, I.  The Journal of Geology, 104(2):224-232  (1996).  doi 10.1086/629816

https://api.openalex.org/works/doi:10.1086/629816

topic: ancient-surfaces camp: mainstream type: measurement access: paywalled timescale: 10^7^ yr direction: undercuts

Key figures: headward advance of the Shoalhaven Gorge runs at about 15× the rate of major escarpment retreat, 250× the average rate of summit lowering, and over 500× the rate of interfluve consumption.

One catchment, four processes, a 500-fold spread.  This is the quantitative reason a single global denudation number cannot be applied to any particular surface, and it is measured rather than modelled.  It also supplies the verified replacement for Oard's unsourced claim of 10 m of Shoalhaven gorge-wall retreat in 30 Myr — the published Shoalhaven results are ratios between process rates, not a retreat distance, and the ratio result is the stronger argument anyway.  Abstract only.

Neogene fluvial landscape evolution in the hyperarid core of the Atacama Desert

Ritter, B.; Binnie, S.A.; Stuart, F.M.; Wennrich, V.; and Dunai, T.J.  Scientific Reports, 8:13952  (2018).  doi 10.1038/s41598-018-32339-9

https://www.nature.com/articles/s41598-018-32339-9

topic: ancient-surfaces, cosmogenic camp: mainstream type: measurement access: OA timescale: 10^6^-10^7^ yr direction: undercuts

Key figures: single-clast ¹⁰Be and ²¹Ne exposure dating puts the end of enhanced fluvial erosion at ~19, ~14 and ~9.5 Ma, with large-scale fluvial modification finished by ~2-3 Ma; clasts recording pre-Miocene exposure at ~28 and ~34 Ma; average Early and Mid-Miocene erosion "could have been well below 1 m/Myr"; reverse-fault scarps up to 300 m high formed since >15 Ma with slip <20 m/Myr; the Adamito fault shows ≥150 m of displacement since ≤9 Ma, "at least 15 m/Myr."

The natural lower bound on how slowly a land surface can erode, measured in the driest place on Earth and available in full open access.  Below 1 m/Myr is thirty to sixty times slower than the global-average denudation rate the young-earth argument extrapolates, and it establishes that the extrapolation is not merely uncertain but wrong by orders of magnitude in the settings where old surfaces actually survive.  The fault-scarp numbers are the useful counterpart: 300 m of relief built and kept for 15 Myr, because the tectonic supply term outruns a near-zero removal term.

Oligocene-Miocene age of aridity in the Atacama Desert revealed by exposure dating of erosion-sensitive landforms

Dunai, T.J.; González López, G.A.; and Juez-Larré, J.  Geology, 33(4):321-324  (2005).  doi 10.1130/G21184.1

https://api.openalex.org/works/doi:10.1130/G21184.1

topic: ancient-surfaces, cosmogenic camp: mainstream type: measurement access: paywalled timescale: 10^7^ yr direction: undercuts

The standard citation for Atacama surface antiquity, listed with a warning.  The OpenAlex abstract is truncated mid-sentence at "Measurement of cosmogenic," Unpaywall reports the paper closed, and GeoScienceWorld returns 403.  Everything after that word is unverified.  ²¹Ne exposure ages of 25, 20 and 14 Ma with individual values at 37 and 9 Ma circulate in search-engine summaries of this paper; they are recorded here as an unchased lead and must not be printed as figures.  Ritter et al. 2018, above, covers the same desert with the same conclusion and was read in full.

Mid-Miocene cooling and the extinction of tundra in continental Antarctica

Lewis, A.R.; Marchant, D.R.; Ashworth, A.C.; and eleven others.  PNAS, 105(31):10676-10680  (2008).  doi 10.1073/pnas.0802501105

https://pmc.ncbi.nlm.nih.gov/articles/2495011/

topic: ancient-surfaces camp: mainstream type: measurement access: OA timescale: 10^7^ yr direction: undercuts

Key figures: an in situ ashfall layer 3 cm thick dated to 14.07 ± 0.05 Ma; cooling of at least 8 °C between 14.07 ± 0.05 Ma and 13.85 ± 0.03 Ma; polar desert conditions in the western Olympus Range continuously "since at least 13.85 Ma."

The independent age control under the Dry Valleys erosion rates, and the reason those rates are believable over millions of years rather than only over the cosmogenic integration window.  A 3 cm ash layer that fell 14 Myr ago is still in place and undisturbed, and the glass shards "show no evidence of either hydration or authigenic clay mineral formation" — chemical weathering has effectively not happened.  A 3 cm layer is also the most direct possible refutation of the idea that these landscapes are being scoured; nothing that erodes at metres per million years leaves a centimetres-thick ash bed intact.  Verification caveat: these fragments came back through a summarising fetch of the PMC page rather than a raw text extraction, and each must be re-checked against the article before being printed as a quotation.  A related and often-cited claim, buried Miocene glacier ice in Beacon Valley reported by Sugden et al. 1995 in Nature 376:412-414, was not obtained from any source and the widely repeated 8 Ma figure is not verified here.

Buried, exhumed, or etched: the mechanism the argument has to get past

Stratigraphic landscape analysis, thermochronology and the episodic development of elevated, passive continental margins

Green, P.F.; Lidmar-Bergström, K.; Japsen, P.; Bonow, J.M.; and Chalmers, J.A.  Geological Survey of Denmark and Greenland Bulletin, 30, 150 pp.  (2013).  doi 10.34194/geusb.v30.4673

https://geusbulletin.org/index.php/geusb/article/download/4673/10344

topic: ancient-surfaces, thermochronology camp: mainstream type: review access: OA timescale: 10^7^-10^8^ yr direction: undercuts

Key figures: the Sub-Cambrian Peneplain extends 700 km along the east coast of Sweden and 450 km across the central Swedish lowlands, and can be followed as an inclined surface in boreholes and seismic profiles for more than 400 km and down to over 4,000 m below sea level offshore.  Lidmar-Bergström's relief classes: pediplains under 20 m, inselberg plains and hilly relief about 20-200 m.

The best single source on this question, 150 open-access pages by the leading Fennoscandian geomorphologist and the apatite-fission-track group, and it answers the antiquity argument from inside mainstream geology rather than around it.  The rule they state is the one the whole debate turns on: "epigene surfaces older than Miocene are unlikely to be preserved unless they have been carved in extremely resistant rocks or preserved in a landscape of high aridity… Re-exposed surfaces can be very old and they are easy to date approximately."  And: "A palaeoplain only exists for a certain time, which varies with different climatologic, tectonic and lithological settings.  Palaeoplains (uplifted epigene or re-exposed surfaces) are therefore transient features."

The Fennoscandian case makes the mechanism concrete.  The Sub-Cambrian Peneplain is roughly 540 Myr old as a surface of formation, was buried under Palaeozoic cover, and is being exhumed now — and once exhumed it begins to be destroyed at once: "the once-horizontal, Sub-Cambrian Peneplain of the South Swedish Dome has been uplifted, tilted and partly re-exposed long after its formation.  It has been incised by valleys and is now being destroyed."  Its extraordinary flatness is evidence of a long-lasting cover, not of a young earth.  Glaciation helps, because ice "easily evacuates saprolites" and strips covers fast.

The most useful thing in it for an honest article is that these authors think Twidale over-dates surfaces: "high-level peneplains are often assigned very old ages even far away from the contact between the basement and its cover rock (e.g. Twidale 2007).  It is argued that they must be older than the cover but without checking if it really is the same surface as at the cover contact."  The dispute over landform antiquity is an internal mainstream argument about which surfaces were buried and when.  In no published source read for this page is it an argument about the age of the earth.

Age and Duration of Weathering by ⁴⁰K-⁴⁰Ar and ⁴⁰Ar/³⁹Ar Analysis of Potassium-Manganese Oxides

Vasconcelos, P.M.; Renne, P.R.; Brimhall, G.H.; and Becker, T.A.  Science, 258(5081):451-455  (1992).  doi 10.1126/science.258.5081.451

https://api.openalex.org/works/doi:10.1126/science.258.5081.451

topic: ancient-surfaces, method camp: mainstream type: measurement access: paywalled timescale: 10^6^-10^7^ yr direction: undercuts

Key figures: supergene cryptomelane from deeply weathered pegmatites in southeastern Brazil gives ⁴⁰K-⁴⁰Ar dates from 10.1 ± 0.5 to 5.6 ± 0.2 Ma; laser-probe ⁴⁰Ar/³⁹Ar step-heating of the two most disparate samples gives plateau dates of 9.94 ± 0.05 and 5.59 ± 0.10 Ma.

The method that dates a weathering profile directly instead of bracketing it between cover units, which is the independent check Twidale's stratigraphic age estimates never had.  A single profile records more than four million years of continuous mineral growth, and the profiles "do not reflect present climatic conditions but are the result of a long-term process that was already advanced by the late Miocene."  The authors also report a time lag between weathering and the sediment pulses offshore, which matters for the sediment-budget half of the erosion argument: rock can be rotted in place for millions of years before any of it is exported.

The Australian applications of this method are the highest-value unread material bearing on the landform-antiquity claim — Li & Vasconcelos 2002 on Mt Tabor, Queensland (doi 10.1016/S0012-821X(02)00594-0), Vasconcelos & Conroy 2003 on the Dugald River valley (doi 10.1016/S0016-7037(02)01372-8), Spier et al. 2006 on the Quadrilátero Ferrífero (doi 10.1016/j.chemgeo.2006.04.006), and the review, Vasconcelos & Carmo 2018, Earth-Science Reviews 179:411-435 (doi 10.1016/j.earscirev.2018.01.003).  Citations for all four are verified; no abstract could be obtained for any of them, and nothing may be quoted from them.

Mountains of southernmost Norway: uplifted Miocene peneplains and re-exposed Mesozoic surfaces

Japsen, P.; Green, P.F.; Bonow, J.M.; and Erlström, M.  Journal of the Geological Society, 175:721-741  (2018).  doi 10.1144/jgs2017-157

topic: thermochronology, ancient-surfaces camp: mainstream type: measurement access: paywalled timescale: 10^7^-10^8^ yr direction: background

Listed to document that the antiquity question is genuinely contested inside mainstream geology, and to be honest that neither side of the contest was read.  Japsen and colleagues argue Norwegian summit surfaces are Miocene peneplains plus re-exposed Mesozoic surfaces, following burial after Late Jurassic rifting.  The opposing case is Hall, A.M.; Ebert, K.; Kleman, J.; Nesje, A.; and Ottesen, D., "Selective glacial erosion on the Norwegian passive margin," Geology 41:1203-1206 (2013), doi 10.1130/G34806.1.  Unpaywall marks the Japsen paper open access but the only deposit found is on figshare, the Lyell Collection returns 403, and GeoScienceWorld is blocked; neither abstract was obtained.  What can be said from Green et al. 2013, which was read in full, is that the two camps disagree about whether passive margins were buried and re-exhumed or eroded continuously — and that both work within the same timescale.

The flat-contact argument, as its advocates make it

'Flat gaps' in sedimentary rock layers challenge long geologic ages

Roth, Ariel A.  Journal of Creation, 23(2):76-81  (August 2009).

topic: hiatus camp: creationist type: review access: OA timescale: n/a direction: supports

Key figures: continental denudation "about 60 mm per thousand years" = 60 m/Myr, halved to 30 m/Myr on the grounds that agriculture doubled the rate.  At 30 m/Myr the expected erosion at each gap: Dead Horse Point 300 m and 600 m, against a canyon 600 m deep; Grand Canyon's three gaps 180 m, 420 m and 3,000 m, against a canyon about 1,600 m deep; Chinle to Bidahochi, 190 Myr, 5,700 m; Morrison to Dakota, 40 Myr, 1,200 m; Bulli Coal, 6 Myr, 180 m; Palo Duro, 200 Myr, ~6,000 m and 16 Myr, 480 m.  Over an assumed 2.5 Gyr for the continents, they "could have been eroded to sea level over 100 times."  The Dakota Formation averages about 30 m thick and covers some 800,000 km².

The technical version of the 1988 paper, and the one to cite.  The argument is stated as a per-gap arithmetic rather than a general complaint, which makes it testable gap by gap.  Two new mainstream admissions are printed here, both second-hand: Stanley Beus on the sub-Redwall unconformity, that even though it represents "more than 100 million years" it "may be difficult to locate"; and Ronald Blakey on the Manakacha–Wescogame boundary, that its location "can be difficult to determine both from a distance and from close range."

The denudation rate is not the weak link.  This project's own flux-derived figure — 15 Gt/yr over 148.48 × 10⁶ km² of land at 2.7 t/m³ — is 37 m/Myr, which sits inside Roth's 30-60 m/Myr bracket.  A skeptic who wants to break this argument has to break it somewhere other than the rate.

Flat gaps in the rock layers challenge long geologic ages

Roth, Ariel A.  Geoscience Research Institute, 1 April 2021.  Adapted from Design and Catastrophe, chapter 47.

topic: hiatus camp: creationist type: review access: OA timescale: n/a direction: mixed

Key figures: uses ~60 m/Myr, not the 30 m/Myr of the 2009 paper, and so gets ~360 m, ~840 m and ~6 km of expected erosion at the Grand Canyon's 6, 14 and 100 Myr gaps — double the 2009 figures for the same gaps.

Tagged mixed for two reasons, one of them a problem on our own side.  The good half is the most useful concession in the whole tradition: where channels are cut at a gap, Roth measures them and reports that the "maximum depth of erosion there is only about 0.5% of the overall average erosion expected" over the assigned interval.  That is the argument in its strongest and most falsifiable form — not "there is no erosion" but "there is two orders of magnitude too little."

The bad half is that the same author, running the same argument, uses denudation rates differing by a factor of two in two papers twelve years apart, and the 2021 version quietly drops the halving-for-agriculture correction that the 2009 version made a point of applying.  Nothing in the 2021 text explains the change.  A reader is entitled to know which number Roth thinks is right.

Sand injectites at the base of the Coconino Sandstone, Grand Canyon, Arizona (USA)

Whitmore, John H. and Strom, Raymond.  Sedimentary Geology, 230(1-2):46-59  (2010).  doi 10.1016/j.sedgeo.2010.06.019

https://isgenesishistory.com/wp-content/uploads/Whitmore-and-Strom-2010-sand-injectites-at-the-base-of-the-Coconino-Sandstone-reduced.pdf

topic: hiatus camp: creationist type: measurement access: OA timescale: n/a direction: mixed

Key figures: cracks examined at 30 locations in the Grand Canyon region under permit GRCA-2005-SCI-0011; the deepest, on the Bright Angel Trail, "at least 15 m deep and 0.25 m wide"; a structureless homogenised zone at the base of the Coconino up to 2 m thick containing bedded Coconino clasts more than 25 cm long; preferred crack orientation about 143°, roughly perpendicular to the Bright Angel and Eminence fault zones.

Original creationist field research published in a mainstream Elsevier journal, and the site should say so plainly rather than burying it.  Six independent lines of evidence — flow banding at both scales, sandstone-clast breccias, lateral sand bodies connected to the vertical cracks, a crack-depth distribution zoned about the Bright Angel Fault, insufficient clay in the Hermit to crack by drying, and the preferred orientation — argue that the sand-filled cracks at the Hermit–Coconino contact are injectites forced up from below, not desiccation cracks.  Thin sections show matrix-supported clasts and almost no porosity, so the sand did not trickle in passively.  That reinterpretation stands on its own and is independent of the age question.

The young-earth inference is the authors' own stated difficulty, not a settled result, and the entry records it in their words: "It is unlikely the Coconino could have remained uncemented in excess of 250 million years" is preceded by the phrase "a problem with no readily apparent solution."  Whitmore and Strom then offer two conventional escapes themselves — cement removal shortly before Laramide faulting, or previously unrecognised early movement on the Bright Angel Fault — and the second is fully compatible with deep time.  Presenting this paper as a demonstrated time conundrum overstates what its authors wrote.

Lithostratigraphic Correlation of the Coconino Sandstone and a Global Survey of Permian 'Eolian' Sandstones

Whitmore, John H.  Answers Research Journal, 12:275-328  (2019).

https://answersresearchjournal.org/coconino-permian-sandstones/

topic: hiatus camp: creationist type: review access: OA timescale: n/a direction: mixed

The clearest statement of the Hermit–Coconino problem, written by the creationist who has done the most fieldwork on that contact.  Whitmore reports that "the contact between the formations is everywhere flat," and sets out why a conventional geologist must infer missing time: 125 km south of the Grand Canyon, about 300 m of Schnebly Hill Formation sits between the Hermit and the Coconino, so on the standard column something on the order of 5-10 Myr has to be absent at the canyon.

What makes the paper honest is that Whitmore states the load-bearing assumption out loud instead of leaving it implicit.  Time must have passed at the Grand Canyon to allow the Schnebly Hill to accumulate to the south, "assuming the top of the Hermit is isochronous."  That single clause is where the argument turns, and Ron Blakey's mapped intertonguing of the Schnebly Hill with the Coconino is direct field evidence that the assumption fails.  Whitmore also concedes the marine half of the problem: "it is not surprising that we see flat contacts between marine layers," which are most of the rocks in the Grand Canyon, "even from a conventional perspective."

The New Geology

Price, George McCready.  Pacific Press, pp. 285-287  (1923).

https://archive.org/download/newgeologytextbo00pric/newgeologytextbo00pric_djvu.txt

topic: history camp: creationist type: book access: book timescale: n/a direction: supports

Key figures: Cretaceous resting on Devonian around Lake Athabaska "for some 150 miles," with the whole Jurassic, Triassic, Permian and Carboniferous absent; at Ust-Waga on the Dwina in northern Russia, Pleistocene shell beds on "absolutely conformable superposition on the horizontal Permian sediments," quoting Suess.

The origin of the entire flat-gaps line of reasoning, and public domain, so the full text can be read by anyone.  Price's point is the same one Roth made sixty-five years later: "there is nothing even in the way of erosion to give a hint of the alleged long ages of time intervening between the successive beds."  His conclusion from it is blunter than any modern creationist would put it — "Judged by these localities, these alleged ages never existed."

There is an irony on the same two pages that this article should not let pass.  Price also reproduces United States Geological Survey photographs of Pennsylvanian clay and sandstone that is still unconsolidated, and notes that Cambrian beds near Petrograd and in Wisconsin are "clays that are scarcely indurated, and sands still incoherent."  So the founder of the argument already knew in 1923 that very old sediment is often still loose — which is precisely the fact that defeats the modern claim that Coconino sand could not have stayed soft for 250 Myr.  Price's own Lake Athabaska example is the same unconsolidated Cretaceous McMurray sand that oil-sands engineers core today.

What is actually at the contacts

Stratigraphy of Grand Canyon National Park, chapter 3

National Park Service / Colorado Plateau Cooperative Ecosystem Studies Unit.  Author and year not recorded in the source audit.

https://irma.nps.gov/DataStore/DownloadFile/637945

topic: hiatus camp: mainstream type: compilation access: OA timescale: n/a direction: undercuts

Key figures: the paleovalleys filled by the Surprise Canyon Formation are "as much as 120 m (400 ft) deep," and the formation forms a dendritic drainage system traceable across the whole park, with fills inside caves in the Mooney Falls and Horseshoe Mesa members of the Redwall Limestone.  The Devonian Temple Butte Formation in eastern Grand Canyon is discontinuous, "filling channels cut into the underlying Cambrian rocks."  The Coconino is 210 m (700 ft) thick in the east and thins "to practically nothing" in western Grand Canyon.

The single most damaging document in this group, and it undercuts the flat-contact argument with mapped physical evidence rather than with theory.  Roth's case is that the great hiatuses show no gullies, no canyons, no valleys and no karst.  At the top of the Redwall, in the same canyon, there is a dendritic river network with 120 m of relief cut into a karst terrain, with caves, and it is filled by a named formation that geologists have been mapping since the 1980s.  At the Ordovician-Silurian gap that Roth uses as a flagship flat contact, the Devonian sits in channels cut into Cambrian rock.

Neither Roth 1988, nor Roth 2009, nor Roth 2021, nor the 2024 podcast presentation mentions the Surprise Canyon Formation.  That omission is the strongest single objection to the compilation method, and the article has to state it in its own voice rather than waiting for a critic to.

Grand Canyon: layered Paleozoic rock, and the missing time

National Park Service.  Grand Canyon National Park geology pages  (undated).

https://www.nps.gov/articles/000/grcatime-layered-paleozoic-rock.htm and https://www.nps.gov/articles/000/grcatime-missing-time.htm

topic: hiatus camp: mainstream type: review access: OA timescale: n/a direction: undercuts

Key figures: Permian ages given as Hermit 285 Ma (284-290), Coconino 280 Ma (276-282), Toroweap 275 Ma (273-278), Kaibab 270 Ma (269-273).  The Great Unconformity is described as "approximately 1.2 billion years (or about 30 percent) of Earth history missing"; the Great Nonconformity as up to 500 Myr; the Great Angular Unconformity as up to 750 Myr.  The Vishnu basement rocks "were exhumed by erosion from 12 mile (20 km) depths."

Included for one narrow but important reason: the Park Service's own account of the four Permian formations says deposition was "nearly continuous" between the Hermit, Coconino, Toroweap and Kaibab, and names no unconformity and no missing interval among them.  The "6 million year gap" displayed in the popular presentations of this argument is not something the standard literature asserts for the Grand Canyon Hermit–Coconino contact.  It is derived by correlation to the Schnebly Hill Formation 125 km to the south, and Whitmore, who does the correlating, writes it with a question mark.

Note the second page cuts the other way and is worth reading beside the first.  Where mainstream geology does claim a huge hiatus in the Grand Canyon, it also claims 20 km of exhumation to go with it, which is the opposite of a surface that sat still doing nothing.

Pennsylvanian and Permian geology of Arizona: the Schnebly Hill Formation

Blakey, Ronald C.  Northern Arizona University course pages (GLG 240), restating Blakey, R.C., GSA Bulletin, 102(9):1189-1217  (1990).

https://jan.ucc.nau.edu/~nrr/GLG240/blakey2a.htm

topic: hiatus camp: mainstream type: review access: OA timescale: n/a direction: undercuts

Key figures: the Schnebly Hill Formation is middle Leonardian, has six members, and thickens from a zero edge in the western Mogollon Rim and in the subsurface east of the Grand Canyon to over 600 m in the Holbrook Basin.  It sharply overlies the Hermit.  It "grades upward and commonly intertongues with the overlying Coconino."  In the Sedona area "no major unconformity is present" between the Schnebly Hill and the Coconino.  There are "no rocks similar to the Schnebly Hill in Grand Canyon."

The mainstream answer to the Hermit–Coconino gap, and it is supplied by the very stratigraphy the creationist presentation cites.  Because the Schnebly Hill intertongues with the Coconino, the base of the Coconino is diachronous: the Coconino at the Grand Canyon is the younger part of a sand sea whose older equivalents lie to the south.  The Grand Canyon contact is then a facies boundary that climbs in time, not one surface that lay exposed for 6 Myr.  The "missing" interval is not a period when nothing happened; it is a period when several hundred metres of sediment were accumulating 125-200 km away.

State the residual honestly.  The Grand Canyon still has no Schnebly Hill at all, so something there was neither receiving sediment nor being deeply eroded while up to 600 m piled up to the south.  "Zero edge" means depositional pinch-out on a very low-gradient surface near base level.  That is coherent and standard, but it is an inference from regional geometry, not an observation of non-erosion.  The peer-reviewed source behind this page, Blakey 1990, sits behind GeoScienceWorld and was not obtained; the facts above come from Blakey's own restatement on a university course site.

The Shinarump Member of the Chinle Formation

Evensen, C.G.  New Mexico Geological Society Field Conference Guidebook, 9:95-97  (1958).

topic: hiatus camp: mainstream type: measurement access: paywalled timescale: n/a direction: undercuts

Key figures: the largest Shinarump channel cut into the Moenkopi, on the west side of Monument Valley, is reported at 3,000 ft (910 m) wide and 275 ft (84 m) deep.

Listed with an explicit warning: the text was not obtained.  The primary PDF at nmgs.nmt.edu refused connection on three attempts, and the channel dimensions above are Wikipedia's rendering of Evensen, not a reading of Evensen.  They must be checked before the article prints them.

If they hold, they matter a great deal.  The Moenkopi-Shinarump contact is presented in the popular version of this argument as a 15-40 Myr gap with "no real evidence of time passage," and its advocate concedes only that "occasionally we'll find troughs carved into the Moenkopi."  An 84 m deep, 910 m wide paleovalley system is not a trough.  Two further details reported from the same source point the same way: the channel form suggests the Moenkopi was still unconsolidated while the de Chelly Sandstone was already well cemented, and the upper Moenkopi beds are bleached where the Shinarump rests on them — which is weathering, at a contact said to show none.

Microstructural characterization of a Canadian oil sand

Author list not confirmed.  arXiv:1301.2658  (2013).

https://arxiv.org/pdf/1301.2658

topic: hiatus camp: unaffiliated type: measurement access: preprint timescale: 120 Myr of burial direction: undercuts

Key figures: bitumen is hosted in "the unconsolidated quartz sands of the 32 m thick McMurray Formation" at depths from 0 to 700 m; the Formation was deposited about 120 Myr ago in an Early Cretaceous paleovalley system; the samples studied came from 75 m depth, with porosity 31-35% and in-situ permeability 3-4 darcy.

A neutral, quantitative refutation of the strongest young-earth inference from the Coconino injectites.  Whitmore and Strom argue that sand injected during Laramide faulting must still have been loose, and that sand cannot stay loose for 250 Myr.  The Athabasca oil sands are a 120 Myr old sand that has been buried to depth, heated, and re-exhumed, and it is still unconsolidated today — loose enough that the industry mines it with shovels.  Cementation depends on pore-fluid chemistry, temperature and burial history, not on elapsed time.

The paper is a rock-mechanics study with no stake in this debate, which is what makes it useful.  Note the coincidence recorded in the Price entry above: Lake Athabaska is the locality Price used in 1923 as a flat-contact example, and it is the same sand.  Caveat: the dossier could not confirm the author list for this preprint, so the citation above is incomplete and should be completed before print.

First-order and super bounding surfaces in eolian sequences — bounding surfaces revisited

Kocurek, Gary.  Sedimentary Geology, 56:193-206  (1988).  doi 10.1016/0037-0738(88)90054-1

topic: hiatus camp: mainstream type: review access: paywalled timescale: n/a direction: undercuts

Listed because it is the biggest hole in this group's evidence, and hiding that would defeat the purpose of the page.  Nothing from this paper was obtained.  ScienceDirect returned 403, no repository copy was found, and OpenAlex holds no abstract; only the DOI, journal, volume, pages and year were confirmed through Crossref.  Nothing is quoted from it here and nothing should be quoted from it in the article.

Its subject is the mechanism that would make a flat Hermit-Coconino contact the expected result rather than an anomaly.  An advancing sand sea can be deflated by wind down to the capillary fringe of the water table, which is by definition a near-horizontal datum, producing a planar "super surface" without any need for a peneplain.  If that account is right, the flatness the argument treats as impossible is what eolian sequence stratigraphy predicts.  Until someone reads Kocurek 1988 and Havholm & Kocurek 1994 (Sedimentology 41:913-934, doi 10.1111/j.1365-3091.1994.tb01432.x) in full, this remains an inference the article cannot document, and it should be labelled as one.

How much time hides in a bedding plane

The Influence of Hiatuses on Sediment Accumulation Rates

Sadler, Peter M.  GeoResearch Forum, 5:15-40, Trans Tech Publications  (1999).

https://tectonics.caltech.edu/meetings/journal_club/session1/Sadler1999TransTech.pdf

topic: hiatus camp: mainstream type: compilation access: OA timescale: hours to 10^8^ yr direction: undercuts

Key figures: the slope of log rate against log time span decodes directly to completeness — 0 means steady accumulation and a complete section, −0.5 is a purely random accumulation process, and −1 is the limit of incompleteness.  Compilation sizes given in the figure captions: deltas 2,988 rates; shelf seas 22,636; continental slopes 6,421; rises and abyssal plains 10,821; abyssal red clays 2,215; peritidal platform and reef 27,029; periplatform apron 7,693; calcareous ooze 68,315; manganese nodules 85 on the shelf and 1,054 abyssal; fluvial and alluvial 12,338.  For carbonate platforms, average rates fall to 10-20% of their shorter-term values across the Milankovitch band, so the bounding surfaces of metre-scale cycles record 80-90% of the time in each cycle.

Sadler's own open restatement of his 1981 result, and the source to use instead of the closed 1981 paper.  The mechanism is stated in one sentence that converts the abstract slope into something a reader can hold: if the average rate for ten-thousand-year intervals is four times the rate for hundred-thousand-year intervals, then only one in four of the ten-thousand-year intervals is represented by sediment and the other three quarters are hiatus.  His opening claim is the mainstream position in eleven words — "Hiatuses pervade the stratigraphic record at all scales from grain boundaries to inter-regional unconformities."

He also answers a common objection before it is made.  A flat contact need not be erosion or deposition: "Much of the time, sediment does not accumulate."  Sadler notes that a short-term observation of most modern floodplains or shelf seas would measure no accumulation at all.

What it does not do, and the article should say so.  This is a statistical description of the record as a whole, not a measurement of the duration of any particular contact.  It shows the record must be riddled with gaps.  It cannot tell anyone that a given flat surface in the Grand Canyon represents six million years.  Read the other way it is a large concession: most of geologic time is not represented by rock anywhere in a given section, which is Roth's premise rather than his conclusion.

Sediment Accumulation Rates and the Completeness of Stratigraphic Sections

Sadler, Peter M.  The Journal of Geology, 89(5):569-584  (1981).  doi 10.1086/628623

topic: timescale-bias camp: mainstream type: compilation access: paywalled timescale: hours to 10^8^ yr direction: background

Key figures: "A compilation of nearly 25,000 rates of sediment accumulation shows that they are extremely variable, spanning at least 11 orders of magnitude."  Completeness is defined as the ratio of median long-term to median short-term accumulation rate, and "expected completeness deteriorates as finer time scales are considered."

The founding paper of the whole timescale-bias literature, cited constantly and read rarely.  Only the publisher-deposited abstract was obtained; journals.uchicago.edu returned HTTP 403 and the full text was not read, so nothing above the two quoted sentences should be attributed to it.  Sadler's stated cause for the trend is unsteady, discontinuous sedimentation rather than measurement error or compaction, both of which he says contribute.

One correction worth making loudly, because it is repeated everywhere.  There is no single "Sadler exponent" in this paper.  Sadler reports gradients that vary by depositional environment, and the 1981 abstract states no one number.  Anyone who cites the paper for a specific exponent has got it from somewhere else.

Time Not Our Time: Physical Controls on the Preservation and Measurement of Geologic Time

Paola, Chris; Ganti, Vamsi; Mohrig, David; Runkel, Anthony C.; and Straub, Kyle M.  Annual Review of Earth and Planetary Sciences, 46:409-438  (2018).  doi 10.1146/annurev-earth-082517-010129

https://www2.tulane.edu/~kmstraub/Publications/Paola_etal_2018.pdf

topic: hiatus camp: mainstream type: review access: OA timescale: hours to 10^6^ yr direction: mixed

Key figures: the worked example is the number to remember.  Take 20 m of cross-bedded fluvial sandstone with a long-term deposition rate of 1 km/Myr = 1,000 m/Myr, so the body represents 20,000 yr on the million-year timescale.  At a mean cross-set thickness of 0.2 m that is 100 sets.  Using measured fluvial sand-dune sediment fluxes of 10⁻⁵ to 10⁻² m²/s, the passage time of each dune face is under four hours, so 100 sets record under 400 hours in total.  Temporal completeness ≤ 2 × 10⁻⁶, and smaller still if turbulence and grain avalanching are resolved.

The strongest statement of the flat-contact problem in the mainstream literature, and the strongest answer to it, in the same paper.  A typical fluvial sandstone body records about two parts in a million of its own elapsed duration; the authors write that "recorded time is best thought of as a kind of dust."  A creationist who quotes only that line has quoted the field's own leading review accurately.

The answer is in where the missing time went.  Paola and colleagues conclude that most of it was not eroded out but represents inactivity — "whatever time gaps are hidden in conformable strata, at any scale, must mainly represent stasis."  A long, quiet, non-erosive, non-depositional surface is exactly what a flat unweathered contact would look like, so the creationist observation and the uniformitarian claim are not in direct contradiction under this model.  They also reject strong catastrophism explicitly: a record with completeness ε would naively require preserved events with recurrence intervals of order 1/ε years, and paleotransport estimates say that is not the case.  The abstract calls the record "strangely ordinary."

The open question they leave, and where the argument actually lives.  Paola et al. themselves flag that nobody knows how much time is represented by ordinary bedding planes showing no sign of stasis.  Stasis is testable — paleosols, bioturbation, karst, hardgrounds, cosmogenic accumulation — and how often those markers are present against absent at claimed paraconformities is a question this audit could not answer from any fetched source.

Real and apparent changes in sediment deposition rates through time

Schumer, Rina and Jerolmack, Douglas J.  Journal of Geophysical Research: Earth Surface, 114:F00A06  (2009).  doi 10.1029/2009JF001266

https://sseh.uchicago.edu/doc/Schumer_and_Jerolmack_JGRF_2009.pdf

topic: timescale-bias camp: mainstream type: model access: OA timescale: 10^0^-10^7^ yr direction: mixed

Key figures: if hiatus durations are drawn from a heavy-tailed distribution with tail parameter γ, where 0 < γ ≤ 1, then the observed accumulation rate decays as t^(γ−1) — so the measured slope is γ minus one.  Fitted values by environment: alluvial channel deposits γ = 0.17, the most incomplete, giving a slope of −0.83; river floodplains γ = 0.48, slope −0.52.  Carbonate platforms decay as t^(−0.39), giving γ = 0.61.  Eastern Alps denudation decays as t^(−0.28), giving γ = 0.72; observed volumetric rates there are 2,000-20,000 km³/Myr against a model-corrected "true average" of about 40,000 km³/Myr — a correction of two- to twenty-fold, not orders of magnitude.

The paper that turned the Sadler effect from a curve into a mechanism, and the reason it cannot simply be dismissed as an artefact of dividing thickness by time.  The derivation is analytic and the tail parameter is measured independently per environment.

It also contains the honest bound.  The best fit to the empirical curves is a truncated Pareto distribution of hiatus lengths, meaning there is an upper limit past which "accumulation rates are (nearly) independent of time."  Geologic and geometric constraints set that limit.  Whoever wants to use timescale bias as an unlimited deflator of measured erosion rates is arguing against this paper, not with it.

Sediment Accumulation Rates and High-Resolution Age-Depth Models

Hagen, Cedric J.  The Sedimentary Record, 22(1)  (2024).  doi 10.2110/001c.124125

https://thesedimentaryrecord.scholasticahq.com/article/124125-sediment-accumulation-rates-and-high-resolution-age-depth-models.pdf

topic: hiatus camp: mainstream type: measurement access: OA timescale: 1 Myr intervals direction: mixed

Key figures: Ediacaran regional composites are "~22% and ~37% complete at a 1-Myr interval" for Oman and China respectively.  Quoting Sommerfield 2006 for Holocene ocean margins: tidal estuary sections about 20-48% complete, sediment-starved slope sections about 85-100% complete.

Real completeness percentages for real sections, which is rarer in this literature than one would expect.  Its value here is the tension it exposes, and both sides should feel it.  Completeness at the one-million-year scale can be anywhere from 22% to 100%; completeness at the hour scale, in Paola's dune example, is about two parts in a million.  Both numbers are correct.  They answer different questions.  Any completeness figure quoted without its timescale is meaningless, and that rule bites creationist and mainstream writers alike.

Two flags.  The completeness equation is rendered as an image in the PDF and could not be read from the extracted text, so the algebraic form should not be reproduced from this source.  The Sommerfield percentages are Hagen's characterisation; the primary (Continental Shelf Research 26:2225-2240, doi 10.1016/j.csr.2006.07.015) was not obtained.

The importance of doing nothing: stasis in sedimentation systems and its stratigraphic effects

Tipper, John C.  Geological Society, London, Special Publications, 404:105-122  (2015).  doi 10.1144/SP404.6

topic: hiatus camp: mainstream type: model access: paywalled timescale: n/a direction: mixed

Abstract only — the full text was not obtained, and nothing below should be attributed to the body of the paper.

The source of the "stasis" concept that Paola et al. build on, and the entry that most complicates easy use of completeness numbers by anybody.  Four claims from the abstract.  Sedimentation systems are generally in stasis.  Stasis is readily preserved in stratigraphic successions.  Successions from high-stasis systems are more complete than those from comparable systems, which is the opposite of what intuition suggests.  And the proportion of stasis in a system cannot be reliably estimated from the successions it produces.

That fourth point deserves to be stated as plainly as Tipper states it, because it is a parameter that cannot be recovered from the rock: the succession a system leaves behind "is necessarily a systematically biased and partial record."  It cuts against both sides equally.  A creationist cannot read absence of erosion off a contact as absence of time, and a uniformitarian cannot read a specific duration off it either.

The Nature of the Stratigraphical Record

Ager, Derek V.  Macmillan, 1st ed. 1973; 2nd ed. 1981; Wiley, 3rd ed. 1993.

topic: hiatus camp: mainstream type: book access: book timescale: n/a direction: mixed

The book was not obtained and nothing in it is quoted here.  The archive.org copy of the second edition (natureofstratigr0000ager) is lending-restricted, and both search-inside endpoints failed during this audit.  Two sentences from it — one about the gap being more important than the record, one comparing earth history to a soldier's life of long boredom and short terror — circulate constantly in young-earth writing, with page numbers that vary by edition and by citer.  Neither wording has been checked against any edition, and until someone borrows the book, neither should appear in the article.

The reason to list an unread book at all is that the quotation is load-bearing for the young-earth case and the context reverses its force.  Ager's thesis is catastrophist about processes and gradualist about time: individual beds are laid down in hours or days, and most of the elapsed time sits in the gaps between them.  The young-earth reading needs the beds to be fast and the gaps to be absent.  Ager thought the gaps were real, long, and dominant.  He also, by second-hand report, added a preface disclaimer to a later edition rejecting creationist use of his work — a disclaimer that reaches this audit only through a critic's page, not through the book.  Printing Ager without that context is the textbook quote-mine, and this site should not do it.

Whether the timescale bias is large enough to matter

Limits to timescale dependence in erosion rates: quantifying glacial and fluvial erosion across timescales

Wilner, Joshua A.; Nordin, Byron J.; Rogers, Emily R.; McDannell, Kalin T.; Palucis, Marisa C.; and Keller, C. Brenhin.  Science Advances, 10(51):eadr2009  (20 December 2024).  doi 10.1126/sciadv.adr2009

https://pmc.ncbi.nlm.nih.gov/articles/PMC11661439/

topic: timescale-bias camp: mainstream type: compilation access: OA timescale: days to 10^8^ yr direction: mixed

Key figures: more than 400 glacial erosion-rate measurements, about ten times any previous open compilation, plus more than 6,000 non-glacial measurements, spanning days to about 100 Myr, from all continents and five method families.  Log-mean erosion rates: glacial 0.51 mm/yr = 510 m/Myr; fluvial 0.067 mm/yr = 67 m/Myr; non-fluvial subaerial 0.00032 mm/yr = 0.32 m/Myr.  Cosmogenic-derived slopes in log rate-timescale space map to about −1 and cluster around a thickness of 600 mm, which is the cosmic-ray attenuation depth.  Volumetric-derived slopes are near zero regardless of glacier type.

The most recent and most decisive paper in this group, and it postdates every creationist and most apologetic treatment of the subject.  It decomposes what everyone calls the Sadler effect into three separate phenomena, of which two are instrument artefacts.  The measurement-thickness effect forces a slope of exactly −1 on any technique that measures the time to erode a fixed thickness, and in this compilation only cosmogenic methods show it — the 600 mm figure is the instrument, not the earth.  The erosion-redeposition effect is one-dimensional Brownian motion, giving a slope of −0.5, and it is the only physical one; it applies to back-and-forth transport below base level, not to unidirectional continental lowering.  The non-observation effect appears only when quiescent intervals are dropped from the average.  Include them, and "unidirectional processes show no timescale dependence at all."

That removes the timescale bias as a general-purpose deflator for continental erosion rates, which is the most common reply to the young-earth argument.  The consistency check is strong: global log-mean fluvial rates from cosmogenic and volumetric methods agree with each other and with the overall 0.067 mm/yr despite differing in timescale by four orders of magnitude, "suggesting the existence of a true long-term log-mean erosion rate."

Tagged mixed, because the same paper cuts back.  Its non-fluvial subaerial log-mean of 0.32 m/Myr is about 120 times below this article's global figure of 37 m/Myr, and it says that most of the land surface, most of the time, barely erodes at all.  A skeptic can use that number as fairly as a creationist can use the removal of the Sadler escape.  One flag: the exact author string could not be confirmed — search results also list Alexander Getraer, whose position in the byline the audit could not verify.

A signature of transience in bedrock river incision rates over timescales of 10⁴-10⁷ years

Finnegan, Noah J.; Schumer, Rina; and Finnegan, Seth.  Nature, 505:391-394  (2014).  doi 10.1038/nature12913

https://finneganlab.org/wp-content/uploads/2015/10/finnegan-et-al-2014-nature-a-signature-of-transience-in-bedrock-river-incision-rates-over-timescales-of-104e28093107-years.pdf

topic: timescale-bias camp: mainstream type: measurement access: OA timescale: 10^4^-10^7^ yr direction: mixed

Key figures: 155 independent river-incision measurements from 14 sites, each spanning at least one order of magnitude in time, including the Indus, Clearwater, Umpqua, Siletz, Yellow, Blue Nile and Grand Canyon rivers, plus Sierra Nevada and Mammoth caves.  11 of the 14 sites show a negative power-law dependence of incision rate on measurement interval.  The mean exponent relating cumulative incision to interval is about 0.8, so the rate exponent is about −0.2.  No upper bound on hiatus length appears in this dataset.

The cleanest measurement of the effect in a landscape process, and the one with the best method.  To dodge the objection that plotting a rate against its own denominator forces a negative slope, Finnegan and colleagues regress cumulative incision against interval and subtract one from the exponent.  Any timescale-bias plot that does not handle that self-correlation is weak evidence, and this one does.  Their conclusion is that long-term incision rates are "apparently inseparable from the interval over which they are measured."

Now do the arithmetic that decides whether it rescues an old earth, because the paper does not do it.  A −0.2 exponent sustained across the full 10⁴ to 10⁷ range buys a factor of about 16.  Stretching a 21 Myr continental levelling time to 4.5 Gyr needs a factor of about 214, which would require an exponent near −0.9 held over six orders of magnitude.  Sadler himself says slopes that steep are rare in the empirical data and persist over limited ranges.  The bias is real; it is roughly an order of magnitude short.

Scaling laws for aggradation, denudation and progradation rates: the case for time-scale invariance at sediment sources and sinks

Sadler, Peter M. and Jerolmack, Douglas J.  Geological Society, London, Special Publications, 404:69-88  (2014/2015).  doi 10.1144/SP404.7

topic: timescale-bias camp: mainstream type: compilation access: paywalled timescale: months to 10^7^ yr direction: supports

Key figures, from the abstract only: upland denudation rates from specific sediment yield show "little or no dependence of rate on time interval."  In the transfer zone between source and sink, erosion and deposition rates balance at all scales.  The product of progradation and aggradation is independent of timescale, implying that global sediment flux into the world's oceans has been constant at the order of 10⁰ m²/yr "from scales of months to tens of millions of years."

The man the effect is named after, retracting its application to the measurement this article depends on.  If upland denudation from sediment yield carries little or no timescale dependence, then the most common rebuttal to the erosion argument — that measured modern rates are Sadler-inflated and cannot be projected backwards — fails at exactly the point it is needed.  Global denudation flux, spatially integrated, is treated in this paper as a time-invariant quantity.

The constraint runs both ways, and the article has to accept both halves.  The same result forbids a creationist from inflating a short-interval sediment yield into a long-term rate, and Sadler and Jerolmack's own caveat is that global rates are invariant only "with appropriate spatial averaging," while "site-specific application remains a daunting challenge."  Access flag: confirmed closed by Unpaywall with no repository copy, and lyellcollection.org returned 403; only the abstract was read.  Willenbring & Jerolmack 2016 restates the same figure independently and is open, so use that as the citable route.

The null hypothesis: globally steady rates of erosion, weathering fluxes and shelf sediment accumulation during Late Cenozoic mountain uplift and glaciation

Willenbring, Jane K. and Jerolmack, Douglas J.  Terra Nova, 28(1):11-18  (2016).  doi 10.1111/ter.12185

https://escholarship.org/content/qt29p0s9d9/qt29p0s9d9.pdf

topic: timescale-bias camp: mainstream type: review access: OA timescale: years to 10^8^ yr direction: mixed

Key figures: sediment flux expressed as cross-sectional growth rate in m²/yr "does not vary systematically from timescales of years to one hundred million years," from a database of thousands of empirical aggradation and progradation values.  ¹⁰Be/⁹Be ratios are invariant across ocean basins over the last 10 Myr, and whole-ocean δ⁷Li is essentially steady over the same interval.

The open route to the Sadler & Jerolmack result, and a useful demonstration that a headline erosion trend can be an artefact.  The authors diagnose the widely reported late-Cenozoic acceleration of mountain erosion as a combination of the Sadler effect and a thickness-limit selection bias in thermochronology, where only fast rates can be sampled for recent intervals because a rock has to be exhumed through a finite depth to be measured at all.

Tagged mixed, and both halves are real.  A genuinely steady global denudation rate over the last 10 Myr is what a uniformitarian needs, and this paper argues for it.  It also destroys the reply that measured erosion rates are meaningless because everything is timescale-dependent.  If the flux really has been constant from years to 10⁸ years, then the levelling arithmetic is a fair question and has to be answered on uplift, isostasy and recycling rather than on measurement bias.

Geomorphic and tectonic process rates: effects of measured time interval

Gardner, Thomas W.; Jorgensen, D.W.; Shuman, C.; and Lemieux, C.R.  Geology, 15(3):259-261  (1987).  doi 10.1130/0091-7613(1987)15<259:GATPRE>2.0.CO;2

topic: timescale-bias camp: mainstream type: compilation access: paywalled timescale: 10^0^-10^7^ yr direction: mixed

Key figure, from the abstract: "a 10⁴ or 10⁵ yr variation in the measured time interval accounts for approximately one order of magnitude change in the measured process rate."

The paper that generalised the effect from sedimentation to landscape and tectonic processes, and the source of the number most often quoted when the effect is invoked against erosion rates.  Its finding is stated flatly: "process rate is not independent of measured time interval."

Quantify what that licenses before using it either way.  One order of magnitude of rate change per four to five orders of magnitude of time is a real bias and far too small to close the gap the erosion argument opens.  Turning a 21 Myr levelling time into 4.5 Gyr needs roughly five orders of magnitude of correction, which on Gardner's own scaling would demand something like twenty orders of magnitude of elapsed time.  Access flag: GeoScienceWorld was unreachable and only the abstract was obtained; the dataset behind the paper — its size, which processes, which regions — is unknown to this audit.

The Institute for Creation Research

Four Geological Evidences for a Young Earth

Clarey, Tim.  Acts & Facts 48(6), Institute for Creation Research  (31 May 2019).

https://www.icr.org/content/four-geological-evidences-young-earth

topic: cosmogenic camp: creationist type: popular access: OA timescale: 10^3^-10^6^ yr direction: supports

Key figures: outcrops erode "at an average rate of about 40 feet every one million years" — that is 12 m/Myr, a correct rendering of Portenga & Bierman's outcrop mean; "the time needed to completely erode most continents would be less than 50 million years"; much of Canada and the eastern United States has had no significant uplift in over 250 Myr and stands less than 1,000 ft above sea level; "1,000 feet of elevation would erode away in just 25 million years."

The arithmetic is right.  A thousand feet at 39.4 ft/Myr is 25 Myr, and the conversion from the source is faithful.  The mismatch is the population.  Portenga & Bierman report an outcrop median of 5.4 m/Myr, which gives 56 Myr for the same 1,000 ft, and they state in the paper that "by collecting from the tops of bedrock outcrops, geologists sample the most stable portions of the landscape."  Clarey takes a bare-rock rate and applies it to a whole continent.  The strongest part of the piece is not the arithmetic at all but the freeboard observation behind it: the Canadian Shield and the eastern United States are low, old and tectonically quiet, and that is the hard case any coupled uplift-and-erosion answer has to meet.

One internal tension deserves naming.  Clarey's third evidence, "There's a Lack of Time Between Layers," argues that megasequence boundaries record too little erosion for the time assigned to them.  His first evidence argues that erosion is too fast for deep time.  Both can be true at once, but the article never puts them side by side, and a reader is entitled to ask which it is.

Continents Should Have Eroded Long Ago

Thomas, Brian.  Institute for Creation Research, Creation Science Update  (22 August 2011).

https://www.icr.org/articles/view/6309/264/

topic: cosmogenic camp: creationist type: popular access: OA timescale: 10^3^-10^6^ yr direction: supports

Key figures: outcrop mean "40 feet every one million years" (12 m/Myr); mean continental elevation 623 m = 2,044 ft, taken from Roth 1986; 2,000 ft at 40 ft/Myr gives "only 50 million years"; basins erode 18× faster than outcrops (218 ÷ 12 m/Myr); basins would level the continents "in less than three million years"; the stated range is "between three and 50 million years"; continents "leveled at least 70 times over."

Every arithmetic step checks.  The problems are elsewhere, and one of them is bad enough that it has to be said plainly, because our own side wrote it.  Thomas describes the isotope as "an isotope of the element beryllium that becomes radioactive with exposure to the sun."  Beryllium-10 is made by cosmic-ray spallation of oxygen and silicon inside quartz, mostly by galactic cosmic rays rather than by the sun, and it is radioactive whether or not anything is exposed to anything.  The paper he is summarising says that "cosmic rays interact with Earth's surface, producing ¹⁰Be."  A skeptic will find that sentence in under a minute.  It should be disowned, not explained.

Two quieter problems sit under it.  Thomas uses only the means, 12 and 218 m/Myr, and never mentions the medians of 5.4 and 54 m/Myr, which would move his stated range from 3-50 Myr to about 11.5-115 Myr.  And "70 times over" appears nowhere in the study; it is his own division of 3.5 Gyr by 50 Myr, introduced with the words "A new study indicates."

Scientific Creationism

Morris, Henry M. (ed.), prepared by the technical staff of the Institute for Creation Research.  Master Books, 1974 and 1985, pp. 155-156.

https://archive.org/download/scientific-creationism/Scientific-Creationism_djvu.txt

topic: denudation-global camp: creationist type: book access: OA timescale: annual river gauging direction: supports

Key figures: Nevins reproduced wholesale — 27.5 Gt/yr of sediment in, 820 × 10¹⁵ t of ocean sediment, "30 million years as the maximum age of the ocean," 383 × 10¹⁵ t of continent above sea level, "383/27.5, or 14, million years."  One figure is added: with subduction removing less than a tenth of the input, "all processes combined therefore can still account for an earth-age of only about 75 million years at the most."

No new data anywhere.  Its value is provenance.  These two pages in the textbook ICR sold for five decades are a compression of one 1973 Impact article, whose entire sediment budget comes from ten pages of a single 1971 textbook — Garrels & Mackenzie, Evolution of Sedimentary Rocks, pp. 102-111.  That ten-page span is the least examined node in the whole chain and could not be located on archive.org under any title search; nobody in the transmission appears to have re-read it, and its caveats are unknown.

Two details worth recording, since a hostile reader will list them.  ICR's own flagship book calls the result settled: "Geologist Stuart Nevins has apparently shown this fact conclusively in a recent study."  And footnote 30 spells the pen name "Nivens" and dates the Impact to October 1973 rather than 1 November.  This passage is also the exact text that TalkOrigins CD501 answers.

Global Stratigraphy Supports a Progressive Worldwide Flood

Clarey, Tim.  Institute for Creation Research  (29 September 2023).  With Grappling with Megasequences, Acts & Facts 44(4)  (31 March 2015).

https://www.icr.org/content/global-stratigraphy-supports-progressive-worldwide-flood and https://www.icr.org/content/grappling-megasequences

topic: hiatus camp: creationist type: model access: OA timescale: n/a direction: supports

Key figures: the ICR Column Project covers over 3,000 stratigraphic columns across five continents, grown from over 500 drilling-core and outcrop samples reported in 2015; the Tejas megasequence holds 32.5% of the total volume of rock deposited; sediment volume peaks in the Zuni, read as Flood day 150 near the end of the Cretaceous; Moshier & Hill 2016 are cited for 160 Myr missing at the Redwall-Muav contact and 1 Gyr at the Tapeats-basement contact.

The inverse of the denudation argument, and in some ways the better half of ICR's case.  "Megasequences are defined as packages of sedimentary rock bounded top and bottom by erosional surfaces."  Clarey's claim is that those bounding surfaces are too planar, across too great a lateral extent, to represent the tens or hundreds of millions of years assigned to them, given the very erosion rates ICR's other articles insist upon.  That is internally consistent: fast erosion should gully a surface, not planate it.

The weakness is not in either argument but in the fact that ICR never states them together.  One set of ICR articles says present erosion is too fast for deep time; another says the erosional surfaces record too little erosion for the time assigned.  A serious treatment has to put both on one page and say how they fit, and the young-earth literature has not done it.

Mount St. Helens, Living Laboratory for 40 Years

Clarey, Tim and Sherwin, Frank.  Acts & Facts 49(5), Institute for Creation Research  (30 April 2020).

https://www.icr.org/content/mount-st-helens-living-laboratory-40-years

topic: method camp: creationist type: popular access: OA timescale: hours to 40 yr direction: mixed

Key figures: up to 400 ft of new strata since 1980; a 25-ft finely laminated unit laid down in a matter of hours; the 19 March 1982 mudflow "carving a new canyon up to 140 feet deep," described as a 1/40th-scale Grand Canyon; the 1980 landslide moved over 3.3 billion cubic yards at more than 150 mph, with 680 million cubic yards going into Spirit Lake; about a million trees sheared; more than 19,000 upright logs mapped on the lake floor in the 1985 survey.

Tagged mixed, and the reason is the whole point of including it.  Mount St. Helens is a rate demonstration, not a rate measurement, and what it demonstrates is that erosion is wildly episodic — that one afternoon can accomplish what a long-run average would allot centuries.  That is good evidence and it should be conceded as such.  It is also the principle that, run backwards, dissolves the denudation argument.  If observed averages do not constrain past events, then today's 60 m/Myr cannot be extrapolated backwards to cap the age of a continent either.

The article repeats the 40 ft/Myr outcrop figure and the "less than 50 million years" conclusion without noticing that its own headline undercuts them.  Both claims appear in the same piece.

Answers in Genesis, and the popular young-earth books

Does Sand Prove Long Ages?

Snelling, Andrew A.  Answers in Genesis feedback article  (9 July 2010).

https://answersingenesis.org/age-of-the-earth/does-sand-prove-long-ages/

topic: method camp: creationist type: popular access: OA timescale: days direction: mixed

Key figures: the Glen Canyon Dam spillway cavitation failure of 28 June 1983 — 32,000 cubic feet per second through a 40-ft-diameter tunnel, implosion pressures estimated as high as 440,000 psi, a hole 32 ft deep by 150 ft long by 40 ft wide cut through 3 ft of steel-reinforced concrete and into sandstone bedrock, 63,000 cubic feet of concrete needed to fill it, and material removed at peak "in excess of 1,000 cubic feet per second."  Surtsey grew sand beaches within a few months of the 1963-64 eruptions ceasing.

The most useful single item in the young-earth corpus for anyone trying to be honest about it, and the reason is structural rather than factual.  Snelling here rejects, in principle, the exact inference his own Best Evidence #1 and #9 depend on: "That statement assumes that erosion to produce the sand on beaches has always occurred at the rate that it is observed to occur today."  The general rule he states is broader still — challenge anyone who converts a present-day process rate into an age about the assumptions doing the work.

Same author, same site, opposite methodological rule, applied according to which direction the answer comes out.  This is not a cheap gotcha.  The Glen Canyon and Surtsey observations are real, they are well chosen, and they are good evidence that catastrophic erosion happens and is fast.  The inconsistency is real too, and any article that wants to use the sediment or salt clocks has to answer it before it uses them.

No Slow and Gradual Erosion

Snelling, Andrew A.  Answers magazine, Flood Evidence Number Five  (1 January 2009).

https://answersingenesis.org/geology/grand-canyon-facts/no-slow-and-gradual-erosion/

topic: hiatus camp: creationist type: popular access: OA timescale: n/a direction: supports

Key figures: about 4,500 ft (1,372 m) of sedimentary layers exposed in the Grand Canyon walls; the Muav Limestone dated 500-520 Ma, the Temple Butte 350-400 Ma and the Redwall 330-340 Ma, so roughly 100 Myr of claimed gap with no erosional relief; the Schnebly Hill Formation, almost 2,000 ft (610 m) thick in central and eastern Arizona, entirely absent in the Canyon with no erosion surface where it should be.

A better argument than the rate arithmetic, and a different one — this is about missing erosion, not excess erosion.  It rests on Snelling's own field photographs of four contacts and it makes a falsifiable claim: that there are no weathering profiles, palaeosols or topographic relief at surfaces which conventional dating separates by tens to hundreds of millions of years.  "The boundary between the Coconino Sandstone and the Hermit Formation is flat, featureless, and knife-edge from one end of the Grand Canyon to the other."

The mainstream answer is that a flat contact is what a low-relief epeiric sea followed by a planing transgression produces, and that the Grand Canyon does contain major erosional unconformities — the Great Unconformity, and the Redwall-Surprise Canyon channels Snelling himself mentions.  This entry is tagged supports because the observations are real and checkable in the field, which makes them a better thing to feature than the sediment-budget arithmetic, and because settling them requires going to the outcrops rather than to a spreadsheet.

Mountain-Building Quiescence in Earth's "Middle Age"

Snelling, Andrew A.  Answers in Depth, Answers in Genesis  (19 March 2021).

https://answersingenesis.org/geology/mountain-building-quiescence-earths-middle-age/

topic: uplift camp: creationist type: review access: OA timescale: n/a direction: mixed

Key figures: none of its own.  It comments on Ming Tang and colleagues in Science, who use europium in detrital zircons as a crustal-thickness proxy and find that mountains stopped growing through the "boring billion," from 1.8 to 0.8 Gyr ago, while erosion wore existing peaks down to stumps.

Snelling reinterprets that history as creation week, then the roughly 1,650 years from creation to the Flood, then the Flood year.  Read what the rates have to be.  An entire supercontinent's mountain belts are eroded inside three days — "as the supercontinent and its mountains rose from under the water, they were eroded rapidly to mature the landscape before he vegetated it all" — then there is near-zero net landscape change for sixteen centuries, then a second global episode of erosion and orogeny inside one year.

No rate calculation appears anywhere in the article.  That is the honest problem, and it is the mirror of the problem in "Does Sand Prove Long Ages?"  A model that can accommodate any erosion rate whatever cannot be tested by erosion rates — and it cannot be used to test anything else by them either.

Very Little Salt in the Sea

Snelling, Andrew A.  Answers magazine, October-December 2012; evidence #9 of 10.

https://answersingenesis.org/evidence-for-creation/9-very-little-salt-in-the-sea/

topic: sediment-budget camp: creationist type: popular access: OA timescale: modern gauging direction: supports

Key figures: 458 million tons of sodium enter the ocean each year (Meybeck 1979); 122 million tons, or 27%, are removed (Sayles & Mangelsdorf 1979); present ocean saltiness is reached from zero in 42 Myr; the most old-earth-friendly assumptions still cap it at 62 Myr (Austin & Humphreys 1990, whose abstract puts the present rate of increase at 3 × 10¹¹ kg/yr).

Included because it is the same logical move as the erosion argument — measure a flux, divide a reservoir, call the quotient an age — and because it fails in a way the erosion argument should be checked against.  The failure is self-inflicted and printed inside the article: the assumptions behind an old ocean, Snelling writes, "fail to take into account the likelihood that God created a saltwater ocean for all the sea creatures He made on Day Five."  A clock whose own author denies its zero point cannot bound anything at all.

The disputed quantity is the output side, and the treatment of it is thin.  Sodium uptake into altered basalt on the ridge flanks is dismissed in one sentence, footnoted to a personal communication on a 1996 mailing list.  That is the weakest citation in AiG's ten-best list, and it is carrying the single most important contested term in the argument.

Evidence for a Young World

Humphreys, D. Russell.  Answers in Genesis, 1 June 2005; adapted from Impact 384, Institute for Creation Research, and also hosted by Creation Ministries International.

https://answersingenesis.org/astronomy/age-of-the-universe/evidence-for-a-young-world/ and https://creation.com/en/articles/evidence-for-a-young-world

topic: sediment-budget camp: creationist type: review access: OA timescale: modern gauging direction: supports

Key figures, AiG's copy: 20 Gt/yr eroded into the ocean, under 400 m of average sediment, 1 Gt/yr subducted, 19 Gt/yr accumulating, "less than 12 million years"; sodium at over 450 Mt/yr in and 27% out, giving 42 Myr and a 62 Myr maximum; total magnetic field energy decaying with a half-life of 1,465 ± 165 yr, so the field "could not be more than 20,000 years old."  CMI's copy of the same author's list: 25 Gt/yr, 24 Gt/yr accumulating, the same "less than 12 million years," and a magnetic field cap of 10,000 years rather than 20,000.

Humphreys deserves credit for a framing his imitators drop.  He states outright that the numbers in bold are "maximum possible ages set by each process, not the actual ages."  Computing a maximum-age bound is a legitimate exercise and saying so is honest.

The problem is that two ministries host two numerically different versions of one man's list.  Walker's CMI article "The mud is missing" uses 20 Gt/yr and nets 19; Humphreys' CMI copy uses 25 and nets 24; both land on "less than 12 million years."  They cannot both follow from one sediment inventory, since 24 Gt/yr over 12 Myr is 2.88 × 10¹⁷ t while 19 Gt/yr over 12 Myr is 2.28 × 10¹⁷ t — a 26% gap that nobody in the chain has remarked on.  It cannot be closed without Hay, Sloan & Wold 1988, which is closed access and was not obtained here.

In the Beginning: Compelling Evidence for Creation and the Flood

Brown, Walt.  Center for Scientific Creation, 8th edition  (2008).  Item 76, "Continental Erosion," p. 40; endnote p. 103.

topic: denudation-global camp: creationist type: book access: book timescale: annual river gauging direction: supports

Key figures: item 76 — "The continents are eroding at a rate that would level them in much less than 25 million years," set against fossils at high elevations that have "somehow avoided this erosion for more than 300 million years."  Item 75 — "More than 27 billion tons of river sediments enter our oceans each year," giving 30 Myr for the seafloor inventory.  Item 74 — volcanoes eject nearly a cubic mile a year, which over 4.5 Gyr would produce ten times the whole volume of Earth's sedimentary rock, though only about 25% of sediments are volcanic.  In the Grand Canyon chapter, Clarence Dutton's Great Denudation swept at least 2,000 cubic miles of Mesozoic rock off about 10,000 square miles of Kaibab Limestone, before 800 cubic miles were excavated to make the canyon.

The research value here is the footnote chain, which Brown prints in full and nobody else does.  Item 76 rests on Nevins pp. 80-81, on George Kennedy's 1959 American Scientist paper, on Roth 1986 pp. 65-67, and on the Dott & Batten sentence about North America being levelled in a mere 10 million years.  Three creationist relays and two secular textbook citations, and no primary measurement anywhere in the chain.

The book also prints its own strongest internal objection on the facing half of a single page.  Item 74 says volcanoes have produced ten times too much material for an old earth; item 76 says erosion should have removed the continents.  TalkOrigins CD501 names that pair as mutually undercutting and is right to.  The page numbers here were derived from running heads in a text extraction of a two-column layout rather than from the printed page.

It's a Young World After All

Ackerman, Paul D.  Baker Books  (1986); web edition with 2002 updates at creationism.org.

topic: history camp: creationist type: book access: book timescale: n/a direction: background

The most-promoted popular young-earth book of the 1980s contains no erosion-rate argument at all.  Every chapter of the on-disk copy was read, and every hit for erosion, erode, sediment, denudation, salt, Niagara, Mississippi and delta was checked in context.  Its clocks are moon dust, a "fossil" meteorite, solar cleanup, the shrinking sun, radiohalos, a decaying speed of light, polystrate fossils and the magnetic field.  Chapter 9, "Back Down to Earth," is where an erosion clock would have gone, and instead runs radiometric anomalies, the Lompoc fossil whale and polystrate fossils.

That is a useful negative datum about how central the erosion argument really was in the popular literature, and the book contains one sentence worth quoting against itself: erosional features "present numerous mysteries and dilemmas for the concept of recent creation as described in the Book of Genesis."  Note also that its case against radiometric dating leans on Setterfield's decaying speed of light, which Creation Ministries International has since placed on its own list of doubtful and inadvisable arguments.

Dismantling 9 "Proofs" of Millions of Years

Ham, Ken.  Answers in Genesis blog  (30 June 2026).

https://answersingenesis.org/theory-of-evolution/millions-of-years/dismantling-9-proofs-millions-of-years/

topic: denudation-global camp: creationist type: blog access: blog timescale: n/a direction: supports

Included for one reason: it is the most recent statement from the largest creationist ministry, and it confirms the argument is live.  The ninth of its nine listed "proofs" is, verbatim, "Erosion rates prove an old earth."  The post runs about six hundred words, contains no data and no citations, points readers to a video, and carries the disclosure "This item was written with the assistance of AiG's research team."

The video it links to, "This Video OBLITERATES All of the Geological 'Proof' for Evolution" (YouTube 4Kk7MVJKDoQ, confirmed as an Answers in Genesis upload through the oEmbed API), is AiG's current spoken answer on erosion rates and is presented by Andrew Snelling.  Its transcript was not obtained.  Nothing anywhere in this article is attributed to it, and nothing should be until somebody pulls it.

Creation Ministries International

Enormous erosion of the continents during the Recessive Stage of the Flood

Oard, Michael J.  Journal of Creation 37(1):10-12  (April 2023); posted 9 August 2024.

https://creation.com/en/articles/enormous-erosion-recessive-stage

topic: sediment-budget camp: creationist type: model access: OA timescale: 1 yr vs 4,500 yr direction: mixed

Key figures: global ocean sediment volume 3.37 × 10⁸ km³ at an average 927 m, after Straume et al. 2019, a 29.7% increase on earlier estimates; about 33% of continental sediment supply attributed to coastal erosion and 10-20% to wind and ice; Roth's present average continental erosion rate of 61 mm/ka = 61 m/Myr; deep-ocean average sediment 404 m, of which half is assigned to the Flood by stipulation; Flood-runoff sediment volume 2.8 × 10⁸ km³ over a continental area of 1.49 × 10⁸ km², giving about 1,900 m of average continental erosion in the Flood's recessive stage; regional totals compiled at 6,000 m for the central Appalachians, 2,400 m for Namibia, 2,500-5,000 m for the Colorado Plateau and 6,000 m for the Flinders Ranges.

Tagged mixed for one sentence Oard writes against his own side: "Applying strict actualism, the average depth eroded from the continents in 4,500 years is only 0.275 m."  That is 61 m/Myr multiplied by 4,500 years, and it is correct.  It is also a concession that present-day erosion is geologically trivial on a young-earth timescale, and it cuts the ground from under the continental-erosion clock.  A model that requires 1,900 m of erosion in one Flood year and produces a quarter of a metre in the 4,500 years since cannot then turn round and use present rates to cap the age of anything.

Two other things belong on the record.  The allocation is asserted rather than derived — none of the present ocean sediment is assigned to the Flood's first 150 days, and the deep-ocean split is a flat fifty-fifty by stipulation.  And Oard puts a third of continental sediment supply down to coastal erosion, where Walker's Eroding ages says the amount removed by winds, glaciers and waves is small next to rivers.  Both are CMI, published sixteen months and twenty-three years apart, and they cannot both be right.

Erosion and the post-Flood boundary

Oard, Michael J.  Journal of Creation 27(2):62-70  (August 2013); posted 25 August 2014.  With Tremendous erosion of continents during the Recessive Stage of the Flood, Journal of Creation 31(3):74-81 (December 2017), and Continental-scale erosion, Creation 35(3):44-47 (July 2013).

https://creation.com/en/articles/flood-boundary-erosion

topic: method camp: creationist type: model access: OA timescale: n/a direction: undercuts

Key figures: up to 850 m of valley- and basin-fill erosion in Wyoming, 1,520 m in Colorado and 1,000 m in New Mexico; Grand Staircase cliff-forming strata about 3,000 m thick, giving a minimum of 1,830 m of erosion, 2,430 m once the Marysvale volcanics are added, and a range of 2,430-3,600 m stripped from the Grand Canyon region before the canyon was cut; the 2017 companion gives about 6,000 m across the Blue Ridge and Piedmont and 2,400 m off the continent adjacent to the southwest African margin.

Tagged undercuts on method, not on conclusion.  Oard sets out the four techniques he uses across the whole series — heights of erosional remnants, projected dips on eroded anticlines, coal rank as a burial-depth proxy, and continental-margin sediment volume — and then rules out the one technique that measures catchment-averaged erosion directly: "Some of them are indirect, such as chemical and cosmogenic isotopic methods, which rely on radiometric dating.  Since these methods are based on uniformitarian assumptions, other more scientific methods will be used."

Cosmogenic nuclide accumulation is the only method that integrates erosion over 10³ to 10⁶ years at the catchment scale, and Kirchner and colleagues showed in 2001 that it agrees with debris-fan accumulation over 16 kyr and with fission-track exhumation over 10 Myr.  Excluding it on presuppositional grounds and then relying on coal rank and dip projection is a discipline-level weakness and should be named as one.  Note also that the erosion totals themselves are ordinary mainstream figures — kilometres of Appalachian denudation is uncontroversial geology — so the dispute in these three articles is about duration, not amount.

CMI presents geological 'misinformation'?

Walker, Tas and Sarfati, Jonathan, replying to Adam Shinabarger.  Creation Ministries International feedback, 31 March 2002; reposted 23 March 2006.

https://creation.com/cmi-presents-geological-misinformation

topic: uplift camp: creationist type: blog access: OA timescale: n/a direction: undercuts

The place where CMI answers the decisive mainstream objection in its own words, and the answer is one sentence long.  The critic, a geology student at Michigan State, wrote that "the rate of uplift that occurs in the various parts of the planet more than compensates for the rates of erosion around the world," citing GPS in the Himalaya and on the US west coast.  CMI's entire substantive reply: "Uplift does not solve the problem because sediments of all ages are found in mountainous regions.  If significant uplift and erosion had occurred, then only old sediments would be present."

No rate comparison, no isostatic calculation, no citation.  Walker's Eroding ages gives the same one-line answer.  Woodmorappe in 2024 gives a different one-line answer — that isostatic compensation would require removing "tens or perhaps hundreds of kilometre thickness of rock" from every continent — also with no calculation and no source for the figure.  Across twenty-four years and three articles, no CMI author has published a quantitative treatment of isostatic compensation against denudation.  That statement rests entirely on CMI's own pages rather than on a hostile source, and it is the largest single gap in the young-earth case.

One concession in the same reply is worth recording, because it is a fair statement of scale coming from the side making the argument: "The estimated average rates are slow — generally less than a millimetre per year."

The mud is missing — so the world is young

Walker, Tas.  Creation 32(3):52  (July 2010); posted 30 September 2011.

https://creation.com/en/articles/missing-mud

topic: sediment-budget camp: creationist type: popular access: OA timescale: modern river gauging direction: supports

Key figures: about 20 billion tonnes a year of sediment reaching the ocean, after Milliman & Syvitski 1992; average ocean-floor sediment under 400 m; 1 billion tonnes a year removed by subduction; 19 billion tonnes accumulating; the whole inventory in under 12 Myr.

The short popular form of the seafloor-sediment argument, and the clearest single place to watch the corpus fail to reconcile with itself.  Humphreys' CMI-hosted list uses 25 Gt/yr and nets 24; this article uses 20 and nets 19; both report "less than 12 million years."  Those cannot both follow from one sediment mass — 24 Gt/yr over 12 Myr is 2.88 × 10¹⁷ t, and 19 Gt/yr over 12 Myr is 2.28 × 10¹⁷ t.  At least one of the two is a conclusion carried forward without being recomputed.

Walker states his own epistemic limit better than almost anyone else in this literature: "So, does the lack of mud on the ocean floor prove the world is young?  No, evidence about past events can only be like forensic science, dealing in probabilities."  One citation problem should be flagged.  His support for regions of seafloor with no sediment at all is Rea et al., Geology 34:873-876 (2006), whose central claim drew a published comment from Glasby and a reply from Rea & Lyle in Geology 35 (2007).  CMI cites the original silently, and neither the comment nor the reply was obtained here.

Coastal erosion too fast for continents to be millions of years old

Author not established.  Journal of Creation 38(3)  (December 2024).

https://www.journalofcreation.com/journalofcreation/2024_volume_38_issue_3/MobilePagedArticle.action?articleId=2018321

topic: coastal camp: creationist type: review access: paywalled timescale: n/a direction: supports

Listed because the coastal version of the argument runs through it, and the text has not been obtained.  Seven routes were tried — a direct fetch with a browser agent, a Googlebot agent, a 2024 Wayback snapshot, the publisher's mirror, the publisher's table-of-contents endpoint, CMI's own client-rendered issue contents page, and seven creation.com slug guesses.  All returned a JavaScript shell, a 403, a 404, or an empty file.  The genre is unconfirmed and the author is unknown; type: review is a guess from the venue.

Nothing is quoted from it here.  One open question was not resolved: this may be the print headline of Woodmorappe's Fleming review in the same issue at pp. 33-35, whose Fleming figures are precisely coastal erosion rates yielding a continental age cap.  Settling that needs a printed copy of Journal of Creation 38(3) or a subscriber login.

The retirement lists, and the clock next door

Arguments we think creationists should NOT use

Creation Ministries International, multi-author, continuously updated.

https://creation.com/arguments-we-think-creationists-should-not-use

topic: history camp: creationist type: review access: OA timescale: n/a direction: mixed

Key figures: 34 to 37 entries under "Which arguments should definitely not be used?" and 13 under "What arguments are doubtful, hence inadvisable to use?"

Every heading on the page was extracted and read.  There is no erosion entry, no continental-denudation entry, no ocean-sediment entry, no salt entry, no Niagara entry, no delta entry and no planation-surface entry in either list.  The only occurrence of the word "erosion" on the whole page is incidental, under the Paluxy tracks.  The one neighbouring young-earth clock that has been retired is moon dust, and the reason given for retiring it is exactly the right one: "these early estimates were wrong, and by the time of the Apollo landings, NASA was not worried about sinking."

Both halves of that finding matter.  It shows the movement can retire an argument on evidence and has done so publicly about fifty times, which earns some credit that a hostile reader will not otherwise extend.  It also shows that the erosion arguments have been looked at and consciously kept.  Nobody writing about this subject can say creationists have moved on from the erosion argument.  The largest creationist ministry has not, and the weakness is therefore unacknowledged rather than superseded.

Arguments Christians Shouldn't Use, and the Arguments to Avoid index

Mitchell, Tommy, and the Answers in Genesis topic index.  Answers in Genesis, 27 July 2010; index spanning 1 September 1990 to 15 February 2026.

https://answersingenesis.org/creationism/arguments-to-avoid/arguments-christians-shouldnt-use/ and https://answersingenesis.org/creationism/arguments-to-avoid/

topic: history camp: creationist type: review access: OA timescale: n/a direction: mixed

Key figures: 21 articles in the index, across thirty-six years.

The same negative finding at the other ministry, and it is starker.  All 21 entries were listed: eight blood types on the Ark, NASA proving Joshua's long day, no rain before the Flood, the Second Law beginning at the Fall, "if humans evolved from apes, why do apes exist," the Japanese trawler plesiosaur, Darwin's deathbed conversion, women having more ribs, giant skeletons in the desert, geocentrism, the boat-shaped rock on Ararat.  Urban legends, hoaxes and Bible-trivia errors.  Not one quantitative age argument appears, and moon dust is not on the list either, though AiG concedes it elsewhere in a children's article.

The framing essay sets the standard the ministry holds itself to, and the standard is a good one: "We must subject creationist claims to the same high standards of peer review we would demand of evolutionists."  A skeptic is entitled to ask whether an erosion clock built on a 1964 river-gauging paper that nobody in the citation chain has read meets it.

Dating Niagara Falls, and Niagara Falls and the Bible

Morris, John D.  Impact 359, Institute for Creation Research  (1 May 2003).  With Pierce, Larry.  Creation 22(4):8-13  (September 2000).

https://www.icr.org/article/dating-niagara-falls/ and https://creation.com/en/articles/niagara-falls-and-the-bible

topic: method camp: creationist type: popular access: OA timescale: 10^2^ yr direction: mixed

Key figures: the gorge from Queenston to the falls is 7 miles = 35,000 ft, with walls 60-90 m high; recession measured since the mid-1800s averages 4-5 ft/yr (1.2-1.5 m/yr), giving 7,000-9,000 years; the notched-crest interval of 1875-1886 cut 60 m in 11 years, or 5.5 m/yr; Pierce's own reconstruction gives 4,400 years; Lyell in 1859 adopted 1 ft/yr against a resident's reported yard a year and got 35,000 years; the Niagara Parks Commission gives 1-1.5 m/yr over the last 560 years and a current rate near 1 ft/yr, possibly falling to 1 ft per decade; Hayakawa & Matsukura 2009 report a decline "from ca. 1 m y⁻¹ a century ago to ca. 0.1 m y⁻¹ at present," which they attribute to diverted discharge and a lengthening waterfall lip rather than to age.

The neighbouring clock most like the erosion argument, and the one place where both sides' arithmetic fails in public.  The historical charge against Lyell is documented and fair: he had an eyewitness rate of about a yard a year and substituted what he himself called a "conjecture" of one foot.  But Pierce's 4,400-year answer is reverse-engineered — he doubles the rate for the downstream section and uses 2 m/yr upstream, above every observed long-run rate, and lands on a figure matching his post-Flood Ice Age date.

The mainstream chronology is not a rate extrapolation either, which is what makes this case instructive.  It has the falls stalling for millennia after meltwater was rerouted and the Niagara River dropped to about a tenth of its flow, then full flow returning about 5,500 years ago, then the falls intersecting the buried St. David's Gorge in what the Niagara Parks Commission calls a geological moment of weeks or days.  Mainstream geology already asserts that the recession rate varied by an order of magnitude.  That concedes the creationists' methodological point and destroys their arithmetic in the same breath.  Two small errors to record: Lyell wrote of Bakewell's son and Pierce's article renders the name "Blackwell"; and the widely repeated "12,300 years" for the gorge does not appear in the Niagara Parks material, which gives deglaciation at about 12,500 years.

Planation surfaces, and what has been measured on them

Earth's Surface Shaped by Genesis Flood Runoff — the planation-surface chapters

Oard, Michael J.  Earth's Surface Shaped by Genesis Flood Runoff, Volume I: Tectonics and Erosion, chapters 32, 34 and 35.  Self-published, michael.oards.net  (2013).

http://michael.oards.net/GenesisFloodRunoff.htm

topic: ancient-surfaces camp: creationist type: book access: OA timescale: n/a direction: supports

Key figures: Oard's Table 35.1, attributed to Summerfield, Global Geomorphology (1991), p. 396, in m/Myr: mountainous with high precipitation 95-740; mountainous with low precipitation 45-370; moderate relief 30-110; low relief and dry 5-35; low relief and temperate 15-30; low relief and tropical 1.5-10.  Ages he lists for surfaces mainstream geology calls old: Arnhem Land plateau >100 Ma; Kangaroo Island >150 Ma; Argentine Andean foothill surfaces 300-400 Ma; Kimberley Plateau >540 Ma; Mojave pediments about 10 Ma.

The full argument, free, in 87 chapter PDFs and 23 appendices, downloadable without a login.  Chapter 32 defines the object: a planation surface is a flat erosion surface that bevels tilted hard and soft strata at the same plane.  Oard takes Melhorn and Edgar's three recognition criteria, rejects paleosols as unreliable, calls accordant summits "equivocal," and rests the case on beveling.  Chapter 34 argues that such surfaces are not forming today and that mainstream geomorphologists say so in print.  Chapter 35 is the hinge: "According to erosion rates today, there should not be any planation surfaces older than a few hundred thousand to a few million years within the uniformitarian timescale."  The load-bearing claim of the whole corpus is Chapter 32's: "Amazingly, planation surfaces are eroded independent of rock hardness on the regional scale."

That claim is an inference from one plane cutting several units, not a measurement.  And the arithmetic underneath it substitutes one rate for another: Summerfield's table is basin-average denudation derived from river solid and dissolved load, and Oard applies it to bare, resistant, low-gradient rock.  Cosmogenic nuclide measurements made directly on such surfaces return 0.2 to 5 m/Myr, which is seven to fifty times slower — see the three Bierman entries below.

Weak spots, recorded so nobody has to find them later.  The caption of Table 35.1 reads "millimeters per year" while the body text one paragraph earlier gives the same low-relief dry figure as "5-35 mm/1000 years"; the body text is right and the caption is wrong by a factor of 1,000, so anyone quoting the caption is out by three orders of magnitude.  Chapter 35 routes its conclusion through the RATE accelerated-decay hypothesis, which stops any reader who does not already accept RATE.  Arnhem Land is placed in Queensland twice, in the text and in the caption of Figure 35.2; it is in the Northern Territory, about 1,500 km away, and the same error appears in the Journal of Creation version.  Chapter 79 is missing from the book's own index, as are chapters 68, 69 and 88 onward.

Erosion rates and patterns in a transient landscape, Grand Staircase, southern Utah, USA

Riley, Kelsey E.; Rittenour, Tammy M.; Pederson, Joel L.; and Belmont, Patrick.  Geology, 47(9):811-814  (2019).  doi 10.1130/G45993.1

https://digitalcommons.usu.edu/geology_facpub/510

topic: cosmogenic, ancient-surfaces camp: mainstream type: measurement access: paywalled timescale: 10^3^-10^4^ yr direction: undercuts

Key figures: 52 alluvial and colluvial ¹⁰Be samples collected in nested fashion from five catchments.  Inferred erosion rates 20 to more than 3,500 m/Myr, a spread of more than 170×.  The authors state the unit identity outright: "20 to >3500 m/m.y. (or mm/k.y.)."  Hotspots are at cliff retreat along the White Cliffs near Mount Carmel Junction and at headward drainage expansion at the Pink Cliffs in Bryce Canyon.  The downslope Vermilion Cliffs near Kanab erode more slowly despite similar slope and lithology, read as a knick-zone that has already passed.

This one targets Oard's showpiece.  His Chapter 9 argument from the Grand Staircase is that soft rock — the Claron Formation under the Table Cliffs — survived while hard volcanic rock above was stripped, and that differential erosion cannot do that over millions of years.  Riley and colleagues measured the modern erosion field in that exact staircase and found it radically non-uniform, "characterized by cliff-bench topography caused by differential erosion of lithologic units," with erosion concentrated at retreating cliff faces and expanding knickzones while benches between them lower slowly.  Whether that pattern holds over the full inferred timescale is a separate argument.  But "erosion cannot leave a soft-rock plateau standing" is contradicted by direct measurement in the landscape Oard chose.

Obtained from the abstract only.  The full text is paywalled and was not read.

Up to 20,000 Feet of Erosion from the Continents, and Tall Erosional Remnants Demonstrate Catastrophic Erosion

Oard, Michael J.  Earth's Surface Shaped by Genesis Flood Runoff, chapters 8 and 9.  Self-published, michael.oards.net  (2013).

http://michael.oards.net/pdf/Book/Chapter09.pdf

topic: denudation-global, ancient-surfaces camp: creationist type: book access: OA timescale: n/a direction: mixed

Key figures, all Oard's, from mainstream sources: Grand Canyon area 1,830-3,000 m of section removed before the canyon formed; San Rafael Swell 4,200-5,100 m by trigonometry on a 6-8° dip; whole Colorado Plateau averaging 2,500-5,000 m over 337,000 km², which is 842,000 to 1,700,000 km³ of rock; Catskill coal rank implying 6.4 km of former burial; Appalachians 4-6.4 km; Vredefort 8-11 km and Sudbury 5 km.  Devils Tower stands 380 m above the Belle Fourche valley and 275 m above the general High Plains level, on phonolite porphyry dated 33-55 Ma; the visitor-centre sign says 2.4 km of sedimentary rock once lay above it and was removed over 50 Myr.  Continent-levelling times he collects: 10 Myr from Roth, 33 Myr from Schumm 1963, 10-25 Myr from Young 1983, 11 Myr for New Zealand from Twidale and Campbell.  He adopts "a maximum of 50 million years to level the continents" as his conservative figure.

The erosion amounts here are mainstream numbers and are not in dispute; what is in dispute is the rate.  Chapter 9 is Oard's most rhetorically effective, because the Devils Tower signage story is checkable: he photographs an old roadside sign saying the top quarter of the tower was exposed 40 Ma and a new one saying it was exposed only in the last 1-2 Myr, and comments, "This is a 95% reduction in the time for erosion!"  His conclusion is blunt: "Devils Tower should have been destroyed quickly in a timeframe of thousands or tens of thousands of years, certainly in less than 100,000 years."  And his general claim about the missing debris is the real argument: "Another important feature of these erosional remnants is that the debris eroded from their surroundings has been transported out of the area; it cannot be located."

Weak spots.  There is no isostatic term anywhere in the levelling arithmetic, and isostatic compensation returns roughly five-sixths of removed mass as rock uplift.  Cliff retreat and surface lowering are not distinguished, which is exactly the mechanism that lets a caprock butte survive while the plain around it lowers.  Oard rules out thermochronology on principle — "Some of them are indirect, such as chemical methods that rely on radiometric dating… I will not use these" — while accepting burial depths derived from coal rank, which is selective.  The sign-change story rests on his own two photographs and nobody else has documented it or checked the current National Park Service text.  And the chapter contains the sloppy phrase "the erosion rate would be around 16 million years," a rate stated as a time.

Erosion, isostatic response, and the missing peneplains

Phillips, Jonathan D.  Geomorphology, 45(3-4):225-241  (2002).  doi 10.1016/s0169-555x(01)00156-8

topic: isostasy, ancient-surfaces camp: mainstream type: model access: paywalled timescale: 10^6^-10^8^ yr direction: undercuts

The text was not obtained.  Only the Crossref and OpenAlex records were verified; no OA copy was reachable, so no figure or quotation from it appears here.  It is listed because the title states the mainstream answer to the continent-levelling arithmetic, because it has 92 citations, and because of one fact about how it has been used: Oard cites this paper in Chapter 50 only for the etymology of "peneplain," never for its thesis.  The one Phillips sentence in the Oard corpus — that Davis coined "peneplain" because a truly flat erosional surface would defy geophysical principles — is quoted second-hand through Oard and has not been checked against the original.

Anyone finishing this page should get the PDF.  Isostasy is the single largest missing term in the young-earth erosion argument, and the paper whose title names it has not been read by either side of this exchange.

Origin of the highly elevated Pyrenean peneplain

Babault, Julien; Van Den Driessche, Jean; Bonnet, Stéphane; Castelltort, Sébastien; and Crave, Alain.  Tectonics, 24:TC2010  (2005).  doi 10.1029/2004TC001697

https://access.archive-ouverte.unige.ch/access/metadata/f1475cf0-0a45-408d-9380-d5eb6fbf3dac/download

topic: ancient-surfaces, uplift camp: mainstream type: model access: OA timescale: 10^7^ yr direction: undercuts

Key figures: low-relief erosional surface remnants in the Pyrenees at more than 2,000 m, locally 2,000-3,000 m.  The southern flank was buried, back to and overlapping the southern margin of the axial zone, in up to 3 km of continental fluvial and alluvial deposits.  Subhorizontal top-wedge alluvial deposits crop out to 2,000 m where they merge into the axial-zone surfaces; their maximum elevation along the south-Pyrenean frontal thrust is about 1,100 m above sea level.  One paleosurface profile falls from 2,000 m to 1,000 m, of which recent erosion accounts for at most about 400 m.  Relief analysis from the SRTM90 digital elevation model.  The competing uplift interpretation would require about 12 km of Plio-Pleistocene crustal thickening, which is the reductio.

The mechanism Oard says does not exist, worked out in full and from a primary source that was read.  Because the foreland basins overfilled and buried the flanks of the range in kilometres of its own debris, the base level of the range itself rose.  That killed the erosive efficiency of the drainage network and let relief smooth out at altitude, with no sea-level peneplain to uplift and no flood.  The authors put it plainly: "we challenge this hypothesis by pointing out that relief applanation does not necessarily require elevation lowering."  And in the conclusion: "we suggest that relief subduing does not necessarily equate to surface elevation lowering, so that relief of mountain belts can be smoothed at high elevation."

It is testable, which is the point.  The model predicts thick, coeval, high-standing piedmont alluvium, and the Ebro basin has it.  The authors expect the case to generalise: "the example of the Pyrenees is probably not a single case."  Oard does not engage this paper anywhere in the chapters examined.

Flat-topped mountain ranges: Their global distribution and value for understanding the evolution of mountain topography

Calvet, Marc; Gunnell, Yanni; and Farines, Bernard.  Geomorphology, 241:255-291  (2015).  doi 10.1016/j.geomorph.2015.04.015

topic: ancient-surfaces camp: mainstream type: review access: paywalled timescale: 10^6^-10^8^ yr direction: mixed

The text was not obtained, and this entry quotes nothing from it.  Unpaywall, OpenAlex, Semantic Scholar, ScienceDirect and the Wayback Machine all failed.  It is listed because it is the modern mainstream survey of exactly the landform Oard says mainstream geology cannot explain, and because it is the highest-priority paper still missing from this page.

Everything currently known about it here comes at second hand, from Oard's two-page response — "The uniformitarian puzzle of mountaintop planation surfaces," Journal of Creation 30(2):9-10 (2016), https://creation.com/mountaintop-planation-surfaces-puzzle — which was fetched.  Oard reports that Calvet and colleagues call rangetop low-gradient tracts enigmatic, that they give a figure for the share of global denudation occurring on the steepest tenth of the continental surface, and that they append a parenthetical saying they ignore creationist views of planation surfaces and naming Oard's website.  If that parenthetical is accurate it is a fair and quotable observation about how the mainstream literature handles him.  It is not verified, so it is not reproduced as their words, and neither is the denudation figure.  Oard's own verdict is quotable because his article was read: "They truly have no idea how to explain mountaintop planation surfaces."

The remarkable African Planation Surface

Oard, Michael J.  creation.com feature article, about 8,500 words  (undated on the page).

https://creation.com/african-planation-surface

topic: ancient-surfaces camp: creationist type: review access: OA timescale: n/a direction: supports

Key figures: the Cypress Hills planation surface in Canada, capped by an average 23 m of well-rounded quartzite gravel and boulders carried more than 400 km from across the continental divide.  The Nullarbor Plain, 200,000 km² of limestone, across which the Transcontinental Railway runs almost 500 km without deviation, with about 60 m of section removed near the coastal margin and an age of about 10 Ma.  The Tibetan Plateau, 2.5 million km² of dissected planation surface at 4,500-5,000 m, cut by gorges 3-4 km deep.  The Grand Canyon area, 1,800-3,000 m removed.  Andean surfaces up to 2,000 km².  King's estimate that about 60% of Africa is planed erosion surface.

The best single entry point for a reader who wants Oard's global claim in one place, and the clearest statement of his definition: "A planation surface is eroded into hard rock or sometimes into unconsolidated sediment by some watery erosive mechanism, usually leaving behind a veneer of rounded rocks on top."  It also contains the one thing his critics rarely credit — a falsification criterion.  A flat depositional surface does not count, and he uses a 60-70 m thick Lake Missoula basalt gravel bar on the Snake River as the control case that his own definition excludes.  That is a fair distinction and it should be granted.

The article's evidential weight is thinner than its length suggests.  The Twidale quotation on the Nullarbor and the Wallace Hansen quotation on the Gilbert Peak surface truncating "rocks of all ages indiscriminately, from Bridger to Precambrian" both reach the reader through Oard; neither original was opened.  Oard's quotations have checked out where cross-checking was possible, which is worth saying, but that is not the same as verification.

Pediments rapidly carved by channelized Flood runoff

Oard, Michael J.  creation.com; restating Earth's Surface Shaped by Genesis Flood Runoff, chapters 64-67 (2013), and Oard, M.J., Journal of Creation, 18(2):15-27  (2004).

https://creation.com/pediments

topic: ancient-surfaces camp: creationist type: review access: OA timescale: n/a direction: supports

Key figures: Whitaker's 1973 bibliography maps pediments from more than 900 references worldwide, with hundreds in the western United States.  Pediment areas occasionally exceed 300 km².  A Tobacco Root Mountains example is 18 km long and 5 km wide, standing 300 m above the adjacent river.

The pediment is the strongest form of Oard's landform argument, because a pediment is a gently sloping bedrock surface cut across resistant rock at the foot of a mountain front, with a sharp junction that no obvious modern process maintains.  The mainstream authors he quotes do concede the difficulty: Dohrenwend on pediments as azonal worldwide phenomena, Selby on the risk that "the origin of the whole form is being confused with the process which is merely retouching the present surface," and M.F. Thomas on the climate muddle, in which pediments are read as markers of wetter conditions in arid zones and drier conditions in humid ones.  Every one of those quotations reaches this page through Oard, not from the originals.

The weakness is a date.  Oard's pediment bibliography effectively stops in 1994 — Whitaker 1973, Hadley 1967, Ritter 1978, Twidale 1978, Oberlander 1974 and 1989, Selby 1985, Dohrenwend 1994 — and Chapter 66 is titled "No viable Uniformitarian Hypothesis."  Three quantitative pediment models were published in 2006 and 2007 and none is engaged; see the next entry.  Oard's 2013 ebook also cites only the 1994 Dohrenwend chapter, not the expanded 2009 second edition with Parsons.  A claim about a literature stops being true when the author stops reading it.

Emergence of pediments, tors, and piedmont junctions from a bedrock weathering-regolith thickness feedback

Strudley, Mark W.; Murray, A. Brad; and Haff, P.K.  Geology, 34(10):805-808  (2006).  doi 10.1130/G22482.1

topic: ancient-surfaces, method camp: mainstream type: model access: paywalled timescale: 10^5^-10^6^ yr direction: undercuts

The text was not obtained — only the Crossref and OpenAlex records, and the abstract returned by OpenAlex is metadata boilerplate.  No figures are reported here.  It is listed because the title alone answers a claim the young-earth literature repeats: that mainstream geomorphology has no mechanism for pediments.  In this model the pediment, the tor and the sharp piedmont junction all emerge from a feedback between bedrock weathering rate and regolith thickness, with no exotic climate and no flood.

Two companion papers exist and were also not obtained: Strudley, Murray and Haff, "Regolith thickness instability and the formation of tors in arid environments," Journal of Geophysical Research: Earth Surface, 111 (2006), doi 10.1029/2005JF000405; and Strudley and Murray, "Sensitivity analysis of pediment development through numerical simulation and selected geospatial query," Geomorphology, 88:329-351 (2007), doi 10.1016/j.geomorph.2006.12.008.  A third line of work ties pediment form to base-level change from drainage integration: Larson and colleagues, "Pediment response to drainage basin evolution in south-central Arizona," Physical Geography, 35(5):369-389 (2014), doi 10.1080/02723646.2014.931089, also not obtained.

If this article ever says that no uniformitarian pediment mechanism exists, a hostile reader will produce these in one search.

Water gaps and transverse drainages

Water and wind gaps carved during late Flood runoff

Oard, Michael J.  Earth's Surface Shaped by Genesis Flood Runoff, chapters 75-86 (2013); popular restatement at creation.com.

https://creation.com/water-and-wind-gaps

topic: ancient-surfaces, uplift camp: creationist type: book access: OA timescale: n/a direction: supports

Key figures: 300 major water gaps in the Zagros Mountains, after Oberlander (1965); "around 1,700" in the Appalachians, given with no source; Himalayan gaps up to 6 km deep.  Lodore Canyon, where the Green River cuts a 701 m slot through the Uinta Mountains although it could have gone around about 3 km to the east.  Shoshone Canyon near Cody, Wyoming, a 760 m gorge with an alternate route about 3 mi to the south.  Union Gap near Yakima, where the lava ridge top stands at 500-600 m while Konnowac Pass 4 mi east is 200 m lower.  Cumberland Gap, a 300 m notch.  Hells Canyon, walls to 2,440 m over 145 km.  His observed catastrophic analogue is the Lake Missoula flood: 2,210 km³ emptied in two days from behind a 610 m ice dam, velocities over 100 kph, cutting Palouse Canyon from an 8-mile-wide head down to a 150 m deep meandering canyon and leaving the 150 m deep Devils Coulee wind gap beside it.

The argument is good where it is descriptive and weak where it is exhaustive.  Chapter 75 attacks the Glossary of Geology definition of a water gap for smuggling two hypotheses into what should be a descriptive term, and that complaint is sound — Lee's 2019 global database concedes the same point by classifying gaps "independent of any presumed method of origin," and counts more than 1,500 transverse drainages deeper than 100 m worldwide, which is the figure that should replace Oard's unsourced 1,700 for the Appalachians alone.  Chapters 80 to 83 then dismiss antecedence, superimposition and piracy in turn, and Chapter 84 supplies the Flood mechanism, in which notches on a ridge deepen as sheet flow channelises and the losers are left as wind gaps.

The fatal gap is a single sentence.  Overflow gets one line — "A fourth, the lake spillover hypothesis, has recently been revived, but it applies to very few water gaps" — and the best mainstream field study finds overflow the most common of the four mechanisms.  Oard's attack is aimed at the wrong target, and it is falsifiable from a paper he himself cites.  Smaller errors to flag: Devils Gate on the Sweetwater River is "350-foot (110 m)" in the text and "a 330-foot (100 m) gap" in the figure caption on the same page; the Atchafalaya River is placed in Mississippi, and it is in Louisiana; John Douglass becomes "Douglas" after first mention; the Geology paper "Preservation or piracy" is miscited as "Preservation of piracy"; and Chapter 80 states that antecedence "was first described in the early 1960s" two sentences after crediting Powell with applying it in 1869.

A criteria-based methodology for determining the mechanism of transverse drainage development, with application to the southwestern United States

Douglass, John; Meek, Norman; Dorn, Ronald I.; and Schmeeckle, Mark W.  GSA Bulletin, 121(3-4):586-598  (2009).  doi 10.1130/B26131.1

https://www2.paradisevalley.edu/~douglass/douglass_gsab_actual.pdf

topic: method, uplift camp: mainstream type: review access: OA timescale: 10^5^-10^7^ yr direction: undercuts

Key figures: 20 transverse drainages in the southwestern United States, selected by an objective random spatial procedure.  Result: 1 piracy, 4 antecedence, 7 superimposition, 8 overflow.  The mechanism could be assigned with high or very high confidence at only 25% of sites.  Half the drainage lines resolved on 1:250,000 digital elevation models, 8 at 1:100,000 and 2 at 1:24,000; every high-confidence assignment came from the two finer scales.  The Grand Canyon is assigned to overflow, developing after 5.9 Ma, and is the oldest overflow gorge in the study area.

The paper that answers the water-gap argument, and the reason to read it is the random sample.  The authors compiled every published discriminating criterion for the four mechanisms into a decision tree with five ordinal confidence levels, published an interactive tool with it, and then applied it blind: "The results are that one transverse drainage likely formed through piracy, four from antecedence, seven from superimposition, and eight from overflow."  Overflow leads, superimposition is second, and antecedence — the mechanism Oard spends the most effort demolishing — accounts for one gap in five.

State the concession as well.  Douglass and colleagues could not assign a mechanism with confidence at three sites in four, and they say so.  Oard's complaint that the evidence for the conventional mechanisms is thin is partly vindicated by their own confidence ratings.  What does not survive is the conclusion he draws from it, that no mechanism is left.

Numerical Evaluation of Post-Flood Formation of Transverse Drainages (Water Gaps)

Mogk, Nathan W.  Creation Research Society Quarterly, 61(3):202-212  (2025).  Accepted 5 June 2024.

https://crsq.creationresearch.org/id/eprint/1347/

topic: uplift, method camp: creationist type: model access: OA timescale: 4,300 yr post-Flood window direction: mixed

Key figures: the headline result is that "the maximal depth of an antecedent or superimposed transverse drainage is determined to be around 200 ft. without enhanced erosion."  Shoshone-Rattlesnake, near Cody, Wyoming: gorge depth 202 ft, slope 73.4 ft/mi, width 114 ft, fitted total erosion 5.09 × 10⁻² ft/yr, which is 15.5 mm/yr or 15,500 m/Myr.  Wheeler Ridge, south of Bakersfield: gorge depth 18 ft, slope 244 ft/mi, width 490 ft, 4.59 × 10⁻² ft/yr or 14,000 m/Myr.  Both sites score a transverse-drainage index of 1 with an alternate divide of 0 ft, meaning neither is spillover-compatible.  Required erosion boosts: Shoshone needs 22.9× under linear uplift or 410× under logarithmic; Wheeler Ridge needs only 3.6× or 68.6×.  Decision rule offered: gaps 2,200 ft deep or more can confidently be placed in the waning stage of the Flood; gaps under 220 ft have an "eminently reasonable" post-Flood origin; in between is ambiguous.

The most valuable single item in the creationist water-gap literature, and it is a partial rebuttal of Oard from inside the Creation Research Society.  Mogk built an uplift-against-incision competition model, calibrated the erosion coefficients on Whipple and colleagues' 1912 Novarupta channel, ran it over a 4,300-year post-Flood window, and concluded that some gaps need the Flood and some do not: "For the Shoshone-Rattlesnake transverse drainage, the necessary erosion strongly implies a late-Flood origin, whereas for Wheeler Ridge, the required erosion is of the same order of magnitude as modeled, and could plausibly be generated in post-Flood time by antecedence."

Two things follow.  Mogk explicitly declines the blanket claim — "Transverse drainages need to be analyzed individually for applicable evidences for all of the available mechanisms" — and he cites the mainstream overflow literature approvingly, where Oard dismisses it in a line.  And he sets an in-house standard that this article should adopt: "hypothetical erosion rates more than an order of magnitude higher should be justified by particular evidence of a more energetic catastrophic process."  That is the honest, defensible version of the water-gap argument.

Rates and processes of bedrock incision by the Upper Ukak River since the 1912 Novarupta ash flow in the Valley of Ten Thousand Smokes, Alaska

Whipple, Kelin X.; Snyder, Noah P.; and Dollenmayer, Kate.  Geology, 28(9):835-838  (2000).

topic: timescale-bias, uplift camp: mainstream type: measurement access: paywalled timescale: 85 yr direction: mixed

Key figures: minimum incision into intact rock of 0.01 to 0.10 m/yr, which is 10,000 to 100,000 m/Myr — three to four orders of magnitude above typical outcrop erosion rates.  Measured over 85 years in a channel of exactly known 1912 birth, created when the ash flow diverted the river.  Stream-power slope exponent n = 0.4 ± 0.2; erosion coefficients K = 2.4 × 10⁻⁴ to 9.0 × 10⁻⁴ m^0.2/yr.

The most useful single number available to a young-earth argument made honestly, and it cuts both ways in the same paper.  It is a direct, dated, mainstream measurement showing that bedrock gorges can be cut at metres per century when the conditions are right — fresh, jointed, weak rock, steep gradient, a newly established channel.  It also notes that despite the speed, "calibrated stream-power erosion coefficients for the Ukak River … are within the range of previously published estimates," so the fast rate is not evidence of exotic physics.

And it is the calibration Mogk used.  With these coefficients the maximum post-Flood antecedent gap comes out around 200 ft, far short of the 2,300-2,500 ft gaps Oard emphasises.  The same paper that shows fast incision is possible also bounds how much of the water-gap population can be explained without a flood — and how much can be explained with one.

Obtained from the abstract only; the full text is paywalled and was not read.

How Rivers Get Across Mountains: Transverse Drainages

Larson, Phillip H.; Meek, Norman; Douglass, John; Dorn, Ronald I.; and Seong, Yeong Bae.  Annals of the American Association of Geographers, 107(2):274-283  (2017).  doi 10.1080/24694452.2016.1203283

topic: method camp: mainstream type: review access: paywalled timescale: n/a direction: mixed

The compact statement of the four mechanisms, from the abstract, which was read in full: antecedence, where the river holds its course during mountain building; superimposition, where a river cuts down through cover onto buried bedrock; piracy, where a steeper path captures a gentler one; and overflow, where a basin fills with sediment and water until it breaches the lowest sill.

It is listed here for one clause, which is a concession to Oard and should be reported as such.  The article "notes a major misconception about the power of headward eroding streams that has dogged scholarship."  That misconception is the mechanism Oard ridicules in Chapter 82 — a phantom stream eroding headward from the Grand Wash Cliffs to the head of the Grand Canyon, up to 200 miles.  Mainstream geomorphology dropped it, and replaced it with overflow.  Oard's criticism landed.  His conclusion that no mechanism therefore remains did not.

Obtained from the abstract only.  The full text is paywalled.  Republished as a chapter in Mountains: Physical, Human-Environmental, and Sociocultural Dynamics, Routledge, pp. 40-48 (2018), doi 10.4324/9781315158914-4.

About Forty Last-Glacial Lake Missoula Jökulhlaups through Southern Washington

Waitt, Richard B., Jr.  The Journal of Geology, 88(6):653-679  (1980).  doi 10.1086/628553

topic: hiatus, method camp: mainstream type: measurement access: paywalled timescale: late Wisconsin direction: undercuts

Key figures: about 40 Touchet Beds rhythmites in the Walla Walla and lower Yakima valleys, and about 40 superposed rhythmites in glacial Lake Missoula bottom sediment in Montana.  Largest lake volume 2,130 km³.  A Mount St. Helens "set S" tephra couplet sits uncontaminated on top of one rhythmite as much as 220 m below the maximum backflood level.  Filled, semi-consolidated rodent burrows occur throughout the 30 m of the thickest section, along with dispersed mammal skeletons.

Included because it undercuts one specific Oard position, not the flood analogue itself.  The Lake Missoula flood as proof that catastrophic water cuts gaps in hard rock in days is uncontroversial and is fair for a young-earth article to use.  Oard's separate claim that there was one flood rather than dozens is his most isolated position, defended five times in print, and it is contradicted by physical evidence rather than by a model: "Evidence that a lengthy nonflood environment followed the emplacement of each of about 40 Touchet Beds rhythmite includes inferred eolian and slopewash sediment overlying many rhythmites, uncontaminated Mount St. Helens 'set S' tephra couplet atop one rhythmite as much as 220 m below the maximum level of backflooding, filled semiconsolidated rodent burrows … and dispersed skeletons of mammals."  Rodents do not burrow underwater and windblown silt does not settle between two beds of the same flood.

The discharge figures come from O'Connor, Jim E. and Baker, Victor R., GSA Bulletin, 104(3):267-279 (1992): peak discharge in the Spokane Valley over 17 ± 3 × 10⁶ m³/s, about 10 ± 2.5 × 10⁶ m³/s at Wallula Gap, flow duration on the order of several days, "the largest known terrestrial fresh-water flows."  That paper grants Oard something worth noting — it concludes the largest flood was a complete rupture of the ice dam, with the later ones smaller subglacial releases.  That is one big flood plus dozens of smaller ones, not one flood.  Both papers were read as abstracts only.

Creationists who criticise, qualify, or complicate the argument

Maintaining Creationist Integrity — A response to Kent Hovind

Wieland, Carl; Ham, Ken; and Sarfati, Jonathan.  Creation Ministries International, 11 October 2002; updated 16 December 2002 and 2 August 2006; further note added November 2009.

https://creation.com/en/articles/maintaining-creationist-integrity-response-to-kent-hovind

topic: history, method camp: creationist type: review access: OA timescale: n/a direction: background

The best-documented case of one creationist organisation forcing another creationist to abandon a clock argument, and it supplies the standard this page is written to.  The moon-dust exchange is reproduced in full, with Hovind's text and CMI's answers interleaved.  The test is stated in one sentence: "To use it in a way which talks of current influx rate 'x' (i.e. without any numbers, implying that x is large enough that there should be a huge thickness after 4.5 billion years) is a form of bearing false witness."  And then the reason: "Such words cannot get around the simple fact that rate 'x' has been measured directly as being much less than the traditional assumption (the error was made by an evolutionist, incidentally).  Until those measurements are found to be false… one cannot in honesty use the argument as it has been used."

Apply that test to erosion and the question is not whether present rates are real — they are — but whether the compensation term has been measured.  CMI also answers the "the verdict is not in yet" defence in advance: "even if Kent were right that 'the verdict is not in', surely this by itself is enough reason not to use it as if it were a refutation of evolution."  And on Paluxy, in words that transfer directly: "Our point was simply that this particular line of evidence should not be used in its present condition of weakness."

Arguments we think creationists should NOT use — and the two other retirement lists

Creation Ministries International (living document, editorial notes dated 2006, 2009 and 2014); Answers in Genesis, "Arguments to Avoid" (undated topic index); New Creation Staff, "Arguments Creationists Should NOT Use" (26 June 2024).

https://creation.com/en/articles/arguments-we-think-creationists-should-not-use, https://answersingenesis.org/creationism/arguments-to-avoid/ and https://newcreation.blog/arguments-creationists-should-not-use/

topic: history camp: creationist type: review access: OA timescale: n/a direction: background

Key figures, from full-text searches of all three pages: on CMI's list the word "erosion" appears exactly once, inside the Paluxy tracks entry, and "salt" appears exactly once, inside the Castenedolo skeleton entry.  There is no entry for continental erosion, denudation rates, sediment influx to the oceans, ocean salinity or Niagara Falls recession.  The Answers in Genesis index returns zero relevant hits.  The New Creation list, which retires the vapour canopy, hydroplate theory and "the Flood created the entire geologic record," also returns zero.

A negative finding, and it belongs on this page because it is the state of play.  No creationist body has ever advised against the erosion argument.  Answers in Genesis in fact runs the closely related ocean-salt argument as item 9 of its "10 Best Evidences from Science That Confirm a Young Earth."  The framing on CMI's own page is the right one and is worth quoting against them: "It is notable that some skeptics criticise creationists when they retract doubtful arguments, but these are also the same people who accuse creationists of being unwilling to change their minds!  Persisting in using discredited arguments simply rebounds—it's the truth that sets us free (John 8:32), not error."

A Little Flood Geology — Part III: A False Dilemma

Klevberg, Peter.  Creation Research Society Quarterly, 56(4):221-233  (Spring 2020).

https://crsq.creationresearch.org/id/eprint/1259/

topic: method camp: creationist type: measurement access: OA timescale: years to decades direction: mixed

Field documentation of slope failure, creep and rilling at Central Montana Rail bridge sites after the 2011 floods — Ross Fork Bridge, the Indian Creek viaduct, the Judith River — with slope inclinometers and horizontal drains installed to monitor movement, used as the vehicle for a methodological argument about how creationists reason about rates.

Klevberg is one of Oard's frequent co-authors and an illustrator of figures in Oard's ebook, which is exactly why this matters.  He is not writing about the erosion clock, but he names the failure mode the erosion clock invites: "Evolutionists tend to overlook or resist evidence for catastrophic, large-scale processes (especially diluvial processes), while creationists can be tempted to overlook or downplay the role of smaller-scale processes that are presently active.  This is a false dilemma."  And in the conclusions: "While many features of the Earth's crust indicate a catastrophic origin, not all do.  Some of the processes observed in the current environment are adequate to explain some of the geologic effects… However, they must not be ignored."

He aims the other barrel at the mainstream too, in the same paper — "A uniformitarian may attempt to explain regional strata and planation surfaces by extrapolating modern effects of rivers and mass wasting, with ludicrous results" — so this is not a defection.  It is the best available creationist warrant for saying, in a creationist article, that slow present-day processes must be given their due in the accounting.

RATE Study: Questions Regarding Accelerated Nuclear Decay and Radiometric Dating

Froede, Carl R., Jr. and Akridge, A. Jerry.  Creation Research Society Quarterly, 49(1):56ff  (Summer 2012).  Accepted 19 October 2011.

topic: method, history camp: creationist type: review access: OA timescale: n/a direction: undercuts

The second in-house retirement-grade critique in the Creation Research Society's own journal, aimed at the flagship creationist research programme of the 2000s.  The verdict: "the apparent lack of empirical data necessary for understanding accelerated nuclear decay seems to provide no practical use for young-earth creationists.  It also has a number of theoretical problems to solve before it can be accepted on even a theoretical basis.  Right now, it might best be described as an interesting hypothesis."  On the heat problem specifically, "no clear solution seems possible, given the unknowns at present."

It matters here for two reasons.  First, Oard's Chapter 35 routes its conclusion through accelerated decay, so this is a critique of the load-bearing assumption behind the planation-surface argument, published by creationists a year before that chapter.  Second, it establishes that CRSQ will print this kind of thing — which makes the absence of any comparable in-house critique of the erosion argument more telling, not less.

The exchange continued.  Marcus Ross, a young-earth paleontologist at Liberty University, replied in "Letter: A developing schism in Flood geology," Journal of Creation 27(3) (2013), granting that creation geology contains "two widely divergent groups seeking to reconstruct Earth history within a Biblical framework."  Only that first paragraph was obtained; the full-text PDF returns HTTP 403, so nothing else from the letter is reported.

The Origin of Grand Canyon Part II: Fatal Problems with the Dam-Breach Hypothesis, and A Dam Breach Unlikely for the Origin of Grand Canyon

Oard, Michael J.  Creation Research Society Quarterly, 46(4):290-307  (2010); and 57(4):206-222  (2021).

https://crsq.creationresearch.org/id/eprint/1262/

topic: method camp: creationist type: review access: OA timescale: n/a direction: background

Oard demolishing a creationist erosion model on evidential grounds, which is the precedent to hold him to.  The dam-breach hypothesis for the Grand Canyon is the most popular young-earth account of the canyon, and Oard rejects it because the physical traces a breached lake should leave are absent: "There are no diagnostic bottom sediments, shorelines, raised deltas, or other geomorphological features at the proposed sites of the breached lakes, despite their common occurrence at the sites of other Ice Age lakes."  The 2021 paper adds a further absence, "a huge cobble and boulder delta at the mouth of Grand Canyon."

The standard he applies is the right one and it is symmetrical.  A proposed catastrophic erosive event must leave physical evidence, and if the evidence is missing the model is rejected.  That same test applies to "catastrophic Flood runoff explains the discrepancy" when it is offered as a rescue for the continental-erosion argument — and to Mogk's rule that erosion rates more than 10× observed need particular evidence of a more energetic process.

A bibliographic conflict to resolve.  One source record gives the 2021 paper as CRSQ 57(3):1 and another as 57(4):206-222.  The second is used here because it came from the fetched PDF record, but it has not been confirmed against the printed issue.

Beyond Scientific Creationism

Reed, John K.; Klevberg, Peter; Bennett, Chris; Akridge, Jerry; Froede, Carl R., Jr.; and Lott, Thomas.  Creation Research Society Quarterly, 41(3):216ff  (December 2004).  Accepted 8 July 2004.

https://www.creationresearch.org/crsq-2004-volume-41-number-3_beyond-scientific-creationism

topic: method, history camp: creationist type: review access: OA timescale: n/a direction: background

Six senior creationist geoscientists arguing, in their own journal, that the whole "here is an empirical young-earth clock" strategy has failed and should be subordinated to philosophical argument: "Empirical arguments developed by an unfunded, outcast minority cannot penetrate the hidebound armor of modern Naturalism despite its many empirical flaws, because at its core Naturalism is not an empirical construct but an integrated worldview."  Their proposal follows: "We advocate the primacy of formal over empirical arguments."

The erosion argument is one of the clocks they are describing.  Recorded here because an article that keeps an empirical clock argument owes an answer to the six people best placed to say it will not work — and because this page's own position is the opposite one, that a measured number beats a worldview claim in both directions.

Why I Left Young-Earth Creationism

Morton, Glenn R.  Copyright 2000; republished by Peaceful Science, 7 September 2020, in memoriam.

https://peacefulscience.org/articles/glenn-morton/

topic: history camp: old-earth-christian type: blog access: OA timescale: n/a direction: undercuts

Key figures: sedimentary sections 30,000 ft thick; buried mountains showing thousands of feet of erosion; 20 or more items published by Morton in the Creation Research Society Quarterly before he left.

The companion piece to Morton's 1986 conference paper, and the irony is the point.  Morton believed the continental-erosion argument — he called it "the best geological argument" for a young earth in print — and he left young-earth creationism anyway, because the same erosion rates make the flood sequence impossible: "I would see buried mountains that had experienced thousands of feet of erosion, which required time.  Yet the sediments in those mountains had to have been deposited by the flood, if it was true."  And on karst inside the column: "It became clear that more time was needed than the global flood would allow."

He also records the reception, which is why the in-house criticism this page went looking for is so thin: "My last young-earth paper was entitled Geologic Challenges to a Young-earth, which I presented as the first paper in the First International Conference on Creationism.  It was not well received."  He lists, with attribution, the names he was called by named creationists, from "compromiser" to "intentionally deceitful."  An argument culture that responds this way to a working oil-industry geophysicist raising data problems will not generate its own corrections.

Devouring our own, and Letter to some young creationists

Wood, Todd C.  Todd's Blog, 9 November 2016 and 2 May 2018.

https://toddcwood.blogspot.com/2016/11/devouring-our-own.html and https://toddcwood.blogspot.com/2018/05/letter-to-some-young-creationists.html

topic: history camp: creationist type: blog access: blog timescale: n/a direction: background

Why the young-earth literature does not correct itself on this topic, from the creationist best placed to know.  In 2016: "Those creationists like me who honestly know that these issues haven't remotely been worked out (despite the protests of our fellow creationists) are becoming increasingly unwelcome in the larger world of young-age creationism.  This is very, very wrong."  And: "We do not have a perfect creationist interpretation of biology, geology, or astronomy.  We do not have all the answers."  In 2018, the cost: "I was taught in school to ruthlessly reject bad ideas.  That was what science was all about, but this world of young-age creationism can't abide that ruthlessness… I'm blacklisted, and I probably deserve it."

Wood went on to host the two 2024 Let's Talk Creation episodes that are the fullest public statement of the erosion argument, and he did not push back on it.  He did say one thing worth more than the pushback would have been.  Comparing the two sides in Episode 88, he observed that geologists defer to radiometric dating and treat erosion as a problem to be figured out later, and that "seems suspiciously familiar to me, like with creationists saying, well, we're pretty sure the Earth must be thousands of years old but I'm not quite sure how to solve this problem with the radiometric dating issue.  It's almost like we're kind of saying the same kinds of things."  That quotation comes from an automatic YouTube transcript, not an authored text, so the wording is approximate and the speaker attribution is inferred from context.

The old-earth Christian geologists

Leaping to Conclusions: What Coastal Erosion Actually Tells Us About Deep Time

Duff, R. Joel.  Naturalis Historia, 8 June 2026.

https://thenaturalhistorian.com/2026/06/08/leaping-to-conclusions-what-coastal-erosion-actually-tells-us-about-deep-time/

topic: coastal camp: old-earth-christian type: blog access: blog timescale: n/a direction: undercuts

Key figures: Cape Cod outer-cliff retreat "on the order of three feet per year" from National Park Service signage, or 0.91 m/yr of horizontal retreat, in Laurentide glacial sand and till about 20,000 years old; Highland Light rolled about 450 ft inland in 1996; mean elevation of North America "roughly 600 meters"; continental denudation "2 to 6 centimeters per millennium," which is 20-60 m/Myr; planation of North America in "ten to twenty-five million years"; last-glacial sea level about 120 m lower than today; Luijendijk et al. 2018 satellite record, 28% of sandy beaches accreting against 24% eroding.

The direct answer to the coastal version of the argument, written by a biology professor at the University of Akron who is Reformed and who wrote two chapters of The Grand Canyon: Monument to an Ancient Earth.  It is a point-by-point reply to Jake Ramgren's "Leaping Lighthouses!" and it makes five moves.  A beach is not a continent, and Cape Cod is a glacial sandpile rather than bedrock.  Continental erosion is vertical and coastal retreat is horizontal, so dividing the width of a continent by a shoreline retreat rate is a category error.  The "why aren't the continents flat" puzzle is a textbook problem, not an oversight.  The continental shelf is intact continental crust that stood above water at the glacial lowstand, not land that erosion removed, and the wide Atlantic shelf against the narrow Pacific one is the ordinary passive-margin versus active-margin contrast.  And continental interiors spent long intervals under shallow seas gaining mass rather than losing it.

Duff concedes the magnitude of the problem without argument, which is why the concession is worth quoting: "The mainstream geological literature openly states that if denudation were the only thing happening, the continents would have been planed flat many times over the span of geological history."  He calls it a standard chapter heading rather than a creationist gotcha.  His sharpest line is the one the main article builds a section on: "You cannot simultaneously argue that observed erosion rates prove the earth is young and that observed erosion rates are a misleading remnant of a vanished catastrophe."

Weak spots, recorded so nobody has to find them later: his 2-6 cm/millennium denudation bracket carries no citation at the point of use; his stated 10-25 Myr planation range does not reproduce from his own 600 m and 20-60 m/Myr inputs, which give 10-30 Myr; and his claim that isostatic uplift on the US east coast runs at "fractions of a millimeter to a few millimeters per year" is uncited and probably wrong in sign, since much of the mid-Atlantic margin is subsiding as the glacial forebulge collapses.

Flood Geology and the Grand Canyon: A Critique

Hill, Carol A. and Moshier, Stephen O.  Perspectives on Science and Christian Faith, 61(2):99-115  (June 2009).

http://www.csun.edu/~vcgeo005/Carol%202.pdf

topic: sediment-budget, hiatus camp: old-earth-christian type: review access: OA timescale: 6 Myr of canyon incision direction: mixed

Key figures: the young-earth claim they answer, quoted from Grand Canyon: A Different View, is that in 70 Myr the Colorado would have stripped more than five miles of rock off a 137,800 square-mile drainage — 8.05 km over 356,900 km², or 2.87 × 10⁶ km³, an implied basin-average denudation of about 115 m/Myr; a shallow Laramide proto-canyon at 40-50 Ma draining north; Basin and Range faulting at the Grand Wash Cliffs from about 16 Ma; eastern and western canyons joined at about 6 Ma; sediment destinations for the last 6 Myr are the Bouse Formation (5.5-5.3 Ma), the Imperial Formation (5.3-2.8 Ma) and the Gulf of California (since 2.8 Ma); 3,600 m of section above the Great Unconformity.

The crux entry for the Grand Canyon half of the subject, and it cuts both ways on the record.  Both authors are confessional Christians and working geologists; Hill's own canyon publications are on karst, speleothem dating and a proposed Laramide proto-canyon.  They dismantle the young-earth framing on timeframe rather than on rate: there was no Colorado River in the Grand Canyon before about 6 Ma, so extrapolating the modern river's sediment load back 70 Myr is arithmetic on the wrong river.  They also say where the last 6 Myr of sediment went, formation by formation.  Their verdict is blunt — young-earth creationism "promote[s] an erroneous and misleading interpretation of the geology of the Grand Canyon."

Then comes the concession, and it is the single most useful sentence in the old-earth Christian corpus for a young-earth writer: "actual measured incision rates are too low (not too high) to explain the carving of the entire Grand Canyon over the last 6 million years."  They lean on Pederson et al. (Geology 30(8):739-742, 2002), whose dated basalts and U-series ages give roughly 140 m/Myr of incision upstream of the Toroweap fault over the past 500 kyr.  At that rate 6 Myr yields 840 m against a canyon 900-1,800 m deep.  Hill and Moshier close the gap with an older proto-canyon or with faster erosion in the past.  That is a reasonable answer, but it is an answer, and a reader is entitled to see that the shortfall is real and admitted in print by the two geologists best placed to deny it.

Sediment Transport and the Coconino Sandstone: A Reality Check on Flood Geology

Helble, Timothy K.  Perspectives on Science and Christian Faith, 63(1):25-41  (March 2011).

https://www.csun.edu/~vcgeo005/Helble.pdf

topic: sediment-budget camp: old-earth-christian type: model access: OA timescale: 12 days direction: undercuts

Key figures: Coconino and correlated sandstones 10,000 mi³ = 42,000 km³, taken from Austin's own numbers, with an average thickness of 315 ft; northern source boundary 1,600 km long; time available for the Coconino inside Austin's 150-day early-Flood window, apportioned by thickness, is 11.8 days, rounded up to 12 in flood geology's favour; assumed transport rate 30 kg s⁻¹ m⁻¹, ten times the highest curve published by Rubin and McCulloch; mass delivered across the whole front in 12 days 5.0 × 10¹³ kg, or 26 km³; shortfall one part in 1,600; at 2.2 km³ per day the Coconino takes about 52 years.

The best arithmetic in the critical literature, and it inverts the whole subject.  Helble is a retired National Weather Service hydrologist and a co-editor of The Grand Canyon: Monument to an Ancient Earth.  His method is to take flood geology's own volume, its own timetable and a transport rate an order of magnitude more generous than anything measured, and then multiply.  Young-earth writers say there is too little sediment for an old earth.  Helble shows there is far too much for a young one, using their numbers.

He also produces the concession that matters most.  Six leading flood geologists wrote in 1994 that "[i]t may not be possible to derive all the Flood sediments from igneous and/or metamorphic precursors by physical and chemical processes in the course of a single, year-long Flood."  Helble reads that as an admission that the Flood could not erode enough pre-Flood rock to supply the section it must deposit.  Not yet answered here: a 2025 ICC New Scholars presentation is titled "Helble's Critique of Creationist Research on the Coconino Sandstone Fails," and it has not been read, so nothing on this page should claim Helble stands unanswered.

"Best evidences for a young Earth": Snelling and the ocean-sediment flux

Baker, Jonathan.  Age of Rocks, 30 May 2014.

https://ageofrocks.wordpress.com/2014/05/30/best-evidences-for-a-young-earth-snelling-and-the-ocean-sediment-flux/

topic: sediment-budget, crust-recycling camp: old-earth-christian type: blog access: blog timescale: Holocene to 10^9^ yr direction: undercuts

Key figures: Snelling's inputs are 20 Gt/yr of river sediment (Milliman & Syvitski 1992), 1 Gt/yr removed by subduction (Hay et al. 1988) and an average ocean sediment thickness of 400 m, giving a maximum age of 12 Myr.  Baker's corrections: Milliman and Syvitski's own text says the pre-agricultural discharge was probably less than half the present level, which moves Snelling's own calculation past 30 Myr; Willenbring's cosmogenic estimate of the long-term flux is about 5.5 Gt/yr, a quarter of Snelling's input; the 400 m average is deep-ocean pelagic sediment, about 74% biogenic calcite formed in surface waters and not delivered by rivers; total subduction recycles the entire crust in 1.8 Gyr (Clift et al. 2009); the Ganges-Brahmaputra delta took up to 70 m of sediment in the Holocene alone, about 6 mm/yr, from a system that carried nearly a tenth of the global flux before damming; fjords capture about 25% of recent riverine sediment.

The strongest technical objection to the whole "where did the sediment go" family, and it is a reservoir mismatch rather than a rate dispute.  Baker is a Reformed Protestant with a doctorate in paleoclimatology, and his point is that the numerator and the denominator describe different oceans.  River sediment lands on shelves, in deltas and in nearshore basins.  The 400 m figure describes the pelagic deep-sea floor, which is mostly plankton skeletons that never came down a river.  Dividing one by the other produces a number, but not an age.

He also states the general rule the erosion argument has to satisfy: "There is nothing scientific about extrapolating rates of any natural process blindly into the past."  He allows two escapes — establish a physical reason the rate cannot change, or make a testable prediction about past rates — and says the sediment argument does neither.  Two cautions. Baker's tone is accusatory, and the invective ("very dishonest research") should not be borrowed even if the numbers are.  And his attribution of the 5.5 Gt/yr figure to "Willenbring et al., 2014" is loose; the number is in Willenbring, Codilean and McElroy, Geology 41(3):343-346 (2013), where it carries an error range of 0.6 to 6.7 Gt/yr.

Grand Canyon Creation

Buchanan, Scott.  Letters to Creationists, standing page, about 173,000 characters.

https://letterstocreationists.wordpress.com/grand-canyon-creation/

topic: hiatus, uplift camp: old-earth-christian type: blog access: blog timescale: 10^8^ yr of stratigraphy direction: undercuts

Key figures: Grand Canyon Supergroup up to 12,000 ft thick; Tapeats Sandstone up to 300 ft; more than 1,000 ft average total thickness of limestone across the United States landmass; fossil-dated gap between the Muav (about 505 Ma) and the Redwall (about 335 Ma) of over 200 Myr, and a further gap of nearly 50 Myr at the top of the Redwall below the Supai (about 285 Ma); the deposition rate flood geology needs across much of North America, "something like a foot an hour," which is 0.3 m/hr.

Buchanan is an evangelical with a doctorate in chemistry, and this page is valuable for three things rather than for original measurement.  First, it preserves two verbatim passages from Davis A. Young's chapter in Portraits of Creation (1990, pp. 68-69) on the Muav-Redwall erosion surface and on the karst developed on top of the Redwall — the primary old-earth geologist statements about those contacts, quoted at length by someone who had the book.  Second, it gives the correct answer to the Mount St. Helens analogy, which is hardness: the Colorado runs year-round and still cuts the Vishnu schist almost imperceptibly, while the St. Helens canyons were cut in unconsolidated ash.  Buchanan's illustration is a jet of water against a dirt hillside versus against a concrete wall.

Third, and most useful for this article, he concedes the premise of the young-earth argument outright and locates the answer in plate tectonics: "If it weren't for active plate tectonics, all the continents would erode down to nothing and be covered by shallow seas."  That is the same concession Duff makes and the same one Neyman makes, arrived at independently.  Note the URL: the slug is hyphenated, and the underscore form 404s.

Too Many Events, Too Little Time

Nelstead, Kevin.  The GeoChristian, 11 November 2025.

https://geochristian.com/2025/11/11/too_many_events_too_little_time/

topic: method camp: old-earth-christian type: blog access: blog timescale: one flood year against ~500 Myr of section direction: undercuts

Key figures, all recorded as unverified — see the caution below: Castile Formation 500 m thick, about 200,000 anhydrite-calcite couplets and so about 400,000 laminae of 1-2 mm each, laterally continuous for up to 110 km; a young-earth budget of roughly 36 days for the Castile gives about 8 seconds per lamina; the Castile is about 10% of a 5,000 m local section; compressing Cambrian through Pliocene into a flood year is a ratio near 500,000,000:1; the largest Yellowstone caldera eruption 2,500 km³, with at least ten large caldera eruptions along the hotspot track; Columbia River and Siberian flood basalts from hundreds of thousands to more than a million km³.

The structural mirror of the erosion argument, run in the other direction.  Nelstead is an ex-young-earth creationist with a geology background who wrote the Novare middle-school Earth science textbook.  Rather than dividing a stock by a flux, he counts discrete events preserved in the record — laminae, lava flows, caldera eruptions, glaciations — and divides by the time a young-earth chronology allows.  His companion post on post-Flood residual catastrophism (30 June 2025) attacks the same escape hatch from the other side, since "residual catastrophism" is exactly the device used to argue that present erosion rates are unrepresentative.

Two things a reader should know. These figures were retrieved through a summarising fetch rather than raw page text, so no wording from this post should be printed as a quotation until the live page is re-read.  And a sweep of the GeoChristian archive under six separate search terms found no dedicated post by Nelstead on the continental erosion-rate argument at all.  He mentions erosion rates once, in a list of 1970s young-earth arguments he was raised on and that the movement has largely dropped, and otherwise attacks flood geology through unconformities, paleosols, paleokarst and event counting.  The absence is worth stating: the erosion clock is not what the old-earth Christian geologists spend their time on.

The Geologic Column and its Implications for the Flood

Morton, Glenn R.  Answers In Science archive, undated; mirrored after the author's death.

http://answersinscience.org/science/Geology/GeologicColumn.htm

topic: hiatus camp: old-earth-christian type: review access: OA timescale: n/a direction: undercuts

Key figures: no rates.  The document is a count, not a clock: intraformational erosion surfaces, subaerial exposure surfaces, karst and cavern systems in the Madison Limestone, paleosols, root casts and erosional channels, tallied through the column.

The most useful move against flood geology in this whole group, and it inverts the erosion question.  Instead of asking how long it takes to erode a continent, Morton asks how many erosion surfaces are preserved inside the sedimentary pile.  Each one requires the surface to have been exposed to air, weathered, and then re-submerged.  A single year has no budget for repeated subaerial exposure, and the Madison karst requires dissolution by rainwater in strata that flood geology deposits in weeks.  Counts are harder to argue with than rates, because there is no equivalent of "the rate was different then."

Morton is the same person who wrote "Geologic Challenges to a Young Earth" for the First ICC in 1986 and called continental denudation the best geological argument for a young earth.  He was a young-earth creationist and an oil-industry geophysicist then, and he left young-earth creationism in 1994.  His account of why is "The Transformation of a Young-Earth Creationist," Perspectives on Science and Christian Faith 52(2) (June 2000), and the erosion item in it is worth chasing: he reports seeing buried mountains in seismic data that had lost more than ten thousand feet to erosion before burial.  That is an erosion observation from the subsurface rather than a surface rate, and it is much harder to dismiss as an unrepresentative modern measurement.  Both documents were read only through summarising fetches or partial extraction, so no sentence from either should be printed as a quotation until the raw text is re-read.

The Salty Sea Argument Is Still Wrong — And Still Being Taught

Duff, R. Joel.  Naturalis Historia, 3 April 2026.

https://thenaturalhistorian.com/2026/04/03/the-salty-sea-argument-is-still-wrong-and-still-being-taught/

topic: method camp: old-earth-christian type: blog access: blog timescale: 10^8^ yr of ocean chemistry direction: background

Key figures: Austin and Humphreys (1990) put the sodium input at 457-458 Mt/yr and catalogue sinks of about 122 Mt/yr, or 27% of input, leaving a residual of 335 Mt/yr; Joly's 1899 estimate 80-90 Myr; Becker's 1910 estimate 50-70 Myr; the Messinian Salinity Crisis (5.97-5.33 Ma) removed 821,000-927,000 km³ of salt, "roughly 4% of the total salt dissolved in the world's oceans, removed in a single geological episode."

Background rather than erosion, and included because it is the template.  The salt clock and the erosion clock are the same machine with different parts, and Duff says so explicitly.  Both divide a stock by a present-day flux.  Both assume the input rate has held constant.  Both need a known starting condition.  And both, in Duff's account, fail on the removal term: "most fatally, it ignores the processes that remove salt from the system."  Anything this article concedes about the erosion clock, a skeptic will immediately test against the salt clock, so the analogy has to be met head-on rather than avoided.

One citation in this post could not be verified: Duff gives "Biebie et al. (2021), Science of the Total Environment, 889, 164138" for an 86% anthropogenic increase in global river sodium flux.  Volume 889 of that journal is 2023, and the author name appears garbled.  Treat the 86% figure as unsourced until the paper is found.

Niagara Falls and the Limits of the Design Inference

Duff, R. Joel.  Naturalis Historia, 25 July 2026.

https://thenaturalhistorian.com/2026/07/25/niagara-falls-and-the-limits-of-the-design-inference/

topic: method camp: old-earth-christian type: blog access: blog timescale: 12 kyr direction: undercuts

Key figures: Laurentide retreat about twelve thousand years ago; Lockport dolostone caprock about 430 Myr old over softer Rochester shale; the gorge has retreated 11 km in 12 millennia, which is 0.92 m/yr; hydroelectric diversion has since "slowed the retreat to a crawl."

The cleanest old-earth Christian statement of the Niagara numbers, and it matters because Niagara is the oldest erosion clock in the young-earth catalogue.  The falls undercut a hard caprock over soft shale, the caprock collapses, and the lip walks upstream.  The classic figure of about a metre a year over twelve thousand years follows the ice, not the earth, so the clock dates the falls and nothing else.  Note also that the modern rate is suppressed by human water diversion, which cuts against any young-earth calculation that starts from a present-day measurement.

A live disagreement inside this group, recorded rather than smoothed over. Duff says 11 km in twelve millennia.  Greg Neyman's argument index says the falls are "about 7,000 years" old.  TalkOrigins CD610 also says about 7,000 years, following G.K. Gilbert.  Neither side gives a citation at the point of use, and the two figures differ by five thousand years.  Nothing on this page should treat either number as established.

Dead Sea Chronicles Part VIII — Mt. Sodom

Duff, R. Joel.  Naturalis Historia, 28 January 2025.

https://thenaturalhistorian.com/2025/01/28/dead-sea-chronicles-part-viii-mt-sodom-a-colossal-salt-pillar-challenging-young-earth-chronology/

topic: uplift camp: old-earth-christian type: blog access: blog timescale: years of InSAR and levelling direction: mixed

Key figures: Mount Sodom rises 226 m (742 ft) above the Dead Sea shoreline, measures 8 by 3 miles, and holds about a billion tons of halite; the root salt layer sits under nearly 10,000 ft of sediment; present uplift 3.5 mm/yr (Yechieli et al. 2016), locally up to 5.5 mm/yr by InSAR (Zucker et al. 2018); Duff's conclusion, "it would take over 60,000 years for Mount Sodom to reach its current height," reproduces exactly — 226 m ÷ 0.0035 m/yr = 64,600 yr.

Included for candour rather than for force, because it is the same move Duff criticises elsewhere.  He takes a present-day rate, divides a present-day height by it, and reads off an age.  The calculation ignores the fact that the salt mountain is dissolving while it rises, so 64,600 years is a minimum and not an estimate — and Duff does not say so.  A young-earth reader is entitled to point out that a one-directional rate extrapolation is exactly what the salt-clock post condemns.  It is also worth recording that his reference list dates Frumkin, Ford and Schwarcz to 2011 for a Quaternary Research volume that is 1999.  None of this rescues a young earth; all of it belongs on the page, because this article holds both sides to one standard.

Eroding Continents? (review of Tas Walker, "Eroding Ages")

Neyman, Greg.  Old Earth Ministries, 1 February 2003; reviewing Creation 22(2), March 2000.

http://www.oldearth.org/rebuttal/magazines/Creation/2000/erode_cont.htm

topic: denudation-global camp: old-earth-christian type: blog access: OA timescale: n/a direction: mixed

Key figures: none.  Not one erosion rate, in any unit, appears anywhere on the page.  The only quantities are that ocean water would cover a flattened earth to 1.7 miles, that Himalayan marine fossils sit above 20,000 ft, and that catastrophic plate tectonics would therefore need Everest to rise 3.8 miles inside a 370-day flood.

Tagged mixed because of one sentence, and it is the strongest concession available to a young-earth writer anywhere in this group: "He is right, if you accept today's erosion rate as constant, or even decrease it by some order of magnitude, you would not have the continents."  Neyman then adds that tectonic, volcanic and depositional gains do not add up to the amount being eroded, and declines to argue the way a geologist would: "I'm not going to argue for changing these rates through time."  His actual answer is supernatural deposition during the creation week, on the strength of Genesis 1:9.

Both halves of that have to be said together.  The concession is real, it comes from someone with every motive not to concede, and it survives a tenfold reduction in the assumed rate.  But Neyman is an apologist without geological training who explicitly refuses to engage the geomorphological literature, and his answer will persuade nobody this article is written for.  It is considerably worse than the ordinary answer that Duff, Buchanan, Hill and Moshier give.  Quoting the concession without that caveat would be cherry-picking, and a skeptical reader would be right to say so.

YEC Argument Index — the erosion family (G61, G71, G72, G313, G331, G332, G333, G334)

Neyman, Greg.  Old Earth Ministries / Answers in Creation, undated argument index.

http://www.oldearth.org/argument/G61_creation_science.htm

topic: denudation-global, sediment-budget camp: old-earth-christian type: review access: OA timescale: n/a direction: undercuts

Key figures: almost none.  G331 gives a Colorado River delta of more than 3,000 square miles, up to 3.5 miles deep, holding over 10,000 mi³ = 41,700 km³ of sediment, with no source cited.  G332 gives a Mississippi delta seven miles thick.  G333 answers Henry Morris's 30 Myr ocean-sediment claim with the observation that thickness tracks seafloor age and reaches 150 Myr worth at the margins.  G334 answers Humphreys' 25 Gt/yr in against 1 Gt/yr subducted with a list of other sinks and no numbers at all.  G72 answers the topsoil clock.  G313 concedes that the Sahara is only a few thousand years old and says so.

Valuable as a map rather than as an argument.  This index is the clearest available inventory of which erosion-family claims young-earth writers actually make, with the original sources named: Morris's Scientific Creationism (1974), Humphreys' "Evidence for a young world" (1999), John D. Morris's ICR Impact #359 on Niagara (2003), Grand Canyon: Monument to Catastrophe, and Kent Hovind.  Anyone auditing the argument should start here for the bibliography and then go elsewhere for the answers.

The answers themselves are thin, and one of them fails on its own numbers.  G331 offers a Colorado delta of 41,700 km³ against Snelling and Vail's claim that about 400,000 km³ was stripped off the Colorado Plateau.  That accounts for roughly a tenth of the volume the young-earth side claims, so the rebuttal does not close the gap on its opponent's own figures — though the figure it is answering is itself sourced to a popular creationist book rather than to a measurement.  G61 answers "after millions of years there would be no mountains" with two sentences about mountain building and no uplift rate.  Its one substantive move is that mountain ranges look like different ages, with the Appalachians far more eroded than the Rockies, which is a real observation that deserves real numbers from somewhere else.

The Grand Canyon, Monument to an Ancient Earth: Can Noah's Flood Explain the Grand Canyon?

Hill, Carol; Davidson, Gregg; Helble, Tim; and Ranney, Wayne (eds.), with eleven Christian geologists and paleontologists.  Kregel Publications, 240 pp.  (2016).

https://ncse.ngo/files/pub/evolution/excerpt--grand.pdf

topic: sediment-budget, hiatus camp: old-earth-christian type: book access: book timescale: n/a direction: undercuts

Key figures: not obtained.  Only a publisher-authorised excerpt was read — Ranney's foreword, the full table of contents, and part of chapter 7.  The erosion and canyon-carving chapters are chapter 16 (Helble and Hill, "Carving of the Grand Canyon: A Lot of Time and a Little Water, a Lot of Water and a Little Time (or Something Else?)", p. 163) and chapter 17 (Davidson, Hill and Ranney, "How Old Is the Grand Canyon?", p. 173).  Neither is in the excerpt.  Second-hand from Kevin Nelstead's review: the book reports Surprise Canyon Formation stream channels cut up to 400 ft into the Redwall Limestone, and argues the young-earth breached-dam hypothesis supplies nowhere near enough water.

The flagship old-earth Christian treatment, written by a team of believing geologists and paleontologists and endorsed by Wayne Grudem and C. John Collins.  Its two carving chapters are the most important unobtained text bearing on this article, and nothing here should characterise their contents beyond what the excerpt and the reviews support.

What can be reported is how the young-earth side answered it.  The Answers Research Journal review states plainly that it analyses only the foreword, the first two chapters, two pages of chapter 7 and the concluding chapter 20, and leaves the geological evidence "to creation geologists."  That is an admission on the record that the flagship young-earth journal did not engage the canyon-carving chapters.  Nelstead reports that Answers in Genesis, ICR and CMI otherwise met the book with silence.  Whether that is still true should be checked before the main article says so.

The Bible, Rocks and Time: Geological Evidence for the Age of the Earth

Young, Davis A. and Stearley, Ralph F.  IVP Academic, 510 pp.  (2008).  ISBN 9780830828760.

topic: method camp: old-earth-christian type: book access: book timescale: n/a direction: undercuts

Key figures: the book was not obtained — not in the local library, no lending copy on archive.org, and the Google Books API returned nothing.  The one figure available is second-hand, from Steven Newton's review in Reports of the National Center for Science Education 29(5) (9 February 2010): Young and Stearley calculate that flood geology requires sediment to accumulate at 36,000 feet per year, which is 10,973 m/yr, about 30 m/day or 1.25 m/hr.  Newton quotes their reaction as "Do Flood geologists really expect anyone to believe that?" (p. 378).

Listed with its text unobtained because it is the most important gap in this section, not because it can be used.  Davis Young is the old-earth Christian geologist whose 1988 Christianity and the Age of the Earth contains the most developed engagement with Nevins's erosion argument in print, and this 2008 book with Stearley is its successor.  Nothing about their treatment of continental denudation is known here.

One number is worth flagging even so.  Young and Stearley's 1.25 m/hr required deposition rate is four times Scott Buchanan's independently derived "foot an hour" for North America, presumably because they use a thicker reference section.  The two figures should not be blended, and neither should be quoted without saying which section it refers to.

The secular rebuttal corpus

An Index to Creationist Claims — CD501, Mountain erosion

Isaak, Mark.  TalkOrigins Archive  (created 22 April 2003, revised 14 September 2004).

https://www.talkorigins.org/indexcc/CD/CD501.html

topic: denudation-global camp: counter-apologetics type: review access: OA timescale: n/a direction: undercuts

Key figures: none.  The claim quotes a 15 Myr figure and attributes it to Henry Morris, Scientific Creationism (1974), p. 155.  The response contains no quantity of any kind.

Included because it is the most-cited skeptical answer to this argument, and because it is a cautionary example.  It answers a quantitative claim with three assertions.  Mountains are still being built, and the Himalayas are still rising — true, but no uplift rate is given.  "Present rates of erosion are particularly high due to more mountain building and higher mountains than usual in earth's history" — a factual claim about earth history, offered with no citation and no number, and one that happens to be well supported by evidence Isaak does not cite.  And the claim contradicts a different creationist claim about volcanism — which is about creationist consistency, not about the earth.

The third point does not defend an old earth at all.  It only says two young-earth arguments conflict.  A creationist reader who has met only this rebuttal has good reason to think the argument survived, and the honest thing for this page to say is that on this topic the canonical skeptical reference contains no data.

An Index to Creationist Claims — CD220 and CD220.1, ocean sediment and subduction

Isaak, Mark.  TalkOrigins Archive  (CD220.1 created 3 July 2000).

https://www.talkorigins.org/indexcc/CD/CD220.html

topic: sediment-budget, crust-recycling camp: counter-apologetics type: review access: OA timescale: 10^8^ yr of seafloor direction: undercuts

Key figures: the claims, from Morris (1974, pp. 155-156) and D. Russell Humphreys (1999): 30 Myr of ocean sediment at present rates; about 25 Gt/yr deposited against about 1 Gt/yr removed by subduction, with an average sediment thickness of 400 m, implying 12 Myr.  The responses give sediment thickness of zero at the mid-Atlantic Ridge rising to "about 150 million years' worth of sediment at the continental margins," with no citation.

The strongest structural argument in the whole skeptical corpus is on CD220, and it is not a rate argument at all.  Four independent quantities agree about the age of the ocean floor: radiometric ages of the basalt, extrapolation from measured spreading rates, the depth predicted by thermal subsidence as the crust cools, and the thickness of sediment on top.  A conveyor belt with four independent speedometers is much harder to dismiss than a single divided quotient.  The page gives no references for any of it, which is a real defect, but every element is checkable.

CD220.1 is unusually honest and correspondingly weak.  It concedes the imbalance outright — more sediment enters the ocean than subduction removes — and then argues that the imbalance is not a clock, because sediment is also scraped off into accretionary prisms and because the Wilson cycle opens and closes basins rather than running at steady state.  That is a legitimate answer.  It is also entirely qualitative: no mass is assigned to prism scraping, to margin storage or to carbonate dissolution, so a reader who wants to know whether those terms actually close the gap will not find out here.

How Good Are Those Young-Earth Arguments?

Matson, Dave E.  TalkOrigins Archive  (1994).

https://www.talkorigins.org/faqs/hovind/howgood-yea.html

topic: denudation-global, sediment-budget camp: counter-apologetics type: review access: OA timescale: n/a direction: undercuts

Key figures, as quoted by Matson from the claim: 27.5 Gt/yr of sediment lost to the oceans against a continental mass above sea level of 383 × 10¹⁵ t, giving 13.9 Myr; ocean sediment of 820 × 10¹⁵ t against the same flux, giving 29.8 Myr.  Both reproduce exactly.  For Niagara: G.K. Gilbert's 7,000-year retreat estimate, against Wisconsinan deglaciation 11,000-12,000 years ago, plus four independent nineteenth-century post-glacial clocks giving 6,000, 7,500, 8,000 and 10,000 years.

The load-bearing skeptical document, since every TalkOrigins index page on this topic links here.  Its best section is Niagara, where five independent nineteenth-century estimates converge on 6,000-10,000 years and cross-check against the modern deglaciation date, and where Matson makes the sharp point that a waterfall cannot cut a gorge through unconsolidated flood sediment and then encounter hard rock.  His best bibliographic contribution is elsewhere: he traces the creationist residence-time table to Goldberg's chapter in Riley and Skirrow's Chemical Oceanography (1965), which means the "ages" are a copied oceanographic table with the column heading changed.

On continental erosion itself he throws the argument away.  He answers with continental growth by accretion and mantle addition, and then writes: "I could also mention that the current rates of erosion are particularly high and that isostatic rebound would greatly increase the time for a continent to erode flat, but that's just icing on the cake."  Those two are the only quantitative corrections available, and isostatic compensation multiplies planation time by roughly five.  Matson had the right answer and declined to develop it.  On the ocean-sediment half he changes the subject: asked about total sediment mass, he answers about sediment distribution, and the mass-balance question he quoted goes unaddressed.  The rhetoric elsewhere in the document ("only a geological moron") is a liability and should not be quoted.

The Age of the Earth (TalkOrigins FAQ)

Stassen, Chris.  TalkOrigins Archive.

https://www.talkorigins.org/faqs/faq-age-of-earth.html

topic: method camp: counter-apologetics type: review access: OA timescale: ocean residence times, 10^2^-10^8^ yr direction: undercuts

Key figures: the full residence-time list, which is the point.  Aluminium 100 years, iron 140, titanium 160, chromium 350, thorium 350, tungsten 1,000, manganese 1,400, lead 2,000, silicon 8,000, nickel 9,000, mercury 42,000, copper 50,000, tin 100,000, zinc 180,000, gold 560,000, silver 2,100,000, potassium 11,000,000, strontium 19,000,000, lithium 20,000,000, magnesium 45,000,000, sodium 260,000,000.

The most methodologically rigorous treatment on the skeptical side, and the one a young-earth writer should read first.  Stassen makes three distinct arguments and only the first is commonly quoted.  The scatter over five orders of magnitude from a single ocean means the method has no internal check, and "[m]ost creationist works do not produce all of the numbers, only the ones whose values are 'convenient.'"  A genuine upper limit requires taking both influx and efflux at the extremes of their uncertainty bands, not at their central values, and if equilibrium is reachable within those extremes then no upper limit exists.  And any nonzero efflux turns the problem from a division into a differential equation.

The second of those is the one that transfers directly to the erosion clock, and it is almost never answered.  A stock divided by a flux is an upper limit only if the removal term is zero.  Every one of the erosion clocks in this catalogue has a nonzero removal term — uplift, isostatic compensation, recycling — and none of them solves the resulting equation.  Note also that Stassen falls into the same hole as Isaak and Matson on one point: he asserts that today's erosion rates are higher than typical past rates, with no citation.

How Old is the Earth: A Reply to "Scientific" Creationism

Dalrymple, G. Brent.  Paper for the Pacific Division, American Association for the Advancement of Science  (1984).  TalkOrigins reprint.

https://www.talkorigins.org/faqs/dalrymple/creationist_age_earth.html

topic: method, history camp: mainstream type: review access: OA timescale: n/a direction: undercuts

Key figures: Dalrymple's Table 10 reproduces Morris and Parker's own list of 49 "ages of the Earth" without alteration.  The erosion-relevant rows, with the ages the creationist authors themselves assign: influx of sediment to the ocean via rivers 30,000,000 yr; erosion of sediment from the continents 14,000,000 yr; leaching of sodium from the continents 32,000,000 yr; leaching of chlorine 1,000,000 yr; leaching of calcium 12,000,000 yr; influx of carbonate 100,000 yr; influx of sulfate 10,000,000 yr; influx of chlorine 164,000,000 yr; influx of sodium via rivers 260,000,000 yr; influx of silicon via rivers 8,000 yr; influx of juvenile water 340,000,000 yr; influx of magma to form the crust 500,000,000 yr.

The most economical refutation of the whole hourglass family, and it needs no commentary because the table is the creationists' own.  Dalrymple, then of the U.S. Geological Survey, simply reprints Morris and Parker's own tabulation (1982, Table I, pp. 254-259).  The same method applied to the same ocean yields 8,000 years from silicon and 260,000,000 years from sodium — a spread of more than four orders of magnitude.  A method whose answers disagree with each other by that much cannot be used to bound the age of anything.

This is also the origin of most of Matson's material and of the residence-time correction that TalkOrigins repeats: a residence time is how long an atom stays in seawater before removal, not how long it took to arrive.  Dalrymple documented that misuse in 1984, and it was still being reprinted decades later.

The Erosion of Continents as a Creationist Clock, and Ocean Floor Sediment as a Creationist Clock

Fischer, Peter; and Tentinger, Jason.  Honors seminar UHON 390 essays, University of South Dakota, Fall 2005 (instructor Timothy H. Heaton).  The live site is dead; both pages survive only in the Internet Archive.

http://web.archive.org/web/20210730185509/http://apps.usd.edu/esci/creation/age/content/creationist_clocks/continents.html

topic: denudation-global, sediment-budget camp: counter-apologetics type: blog access: OA timescale: annual rates extrapolated direction: mixed

Key figures: Fischer uses a continental erosion rate of "approximately .005 millimeters per year," which is 5 m/Myr, and a mean continental elevation of about 800 m, both sourced to Hugh Ross, A Matter of Days (NavPress, 2004), pp. 186-187; the claim answered is 28 billion tons eroded per year giving 15 Myr; Himalayan uplift more than one centimetre per year; Kilauea had produced 2.7 km³ of lava covering 117 km² as of January 2005.  Tentinger reports that only 1-10% of skeletal debris survives to become sediment, that terrigenous sediment near margins accumulates at up to a few metres per thousand years while siliceous ooze accumulates at a few metres per million years — a ratio near 1,000:1 — and that the oldest seafloor is about 200 Myr.

Fischer's essay fails its own arithmetic and should be read as a warning rather than as evidence.  800 m divided by 0.005 mm/yr is 160 Myr, not the 15-16 Myr he says those figures produce.  The two numbers he supplies are inconsistent with the conclusion he attributes to them by a factor of ten, and the discrepancy goes unremarked.  To reach 15 Myr from 800 m the rate would have to be about 0.05 mm/yr.  Whether the error originates with Hugh Ross's book or with the student could not be checked.  Compounding the problem, the target he chose is a San Diego church handout rather than Nevins 1973 or Morris 1974 — answering the weakest available statement of a claim is a methodological flaw even when it runs in this article's favour.

Tentinger's sibling essay is substantially better and deserves separate credit.  It supplies real accumulation-rate ranges from an Open University textbook, and it makes a falsifiable prediction-contrast rather than an assertion: a single global flood predicts compositionally uniform seafloor sediment, while the observed pattern is provincial — diatom ooze at high latitudes and upwelling zones, radiolarian ooze at low latitudes, terrigenous clay near margins.  That is a testable discriminator and the young-earth side has to answer it.  Its faults are sloppiness rather than error: it misspells Humphreys throughout and misstates his position as a claim that the earth cannot be older than twelve million years, when Humphreys argues for about six thousand and offers twelve million only as a ceiling the mainstream model already exceeds.

Replies to the Pamphlet "Evidence for a Young World"

Redfern, Francis.  PRISM, personal site; the server rejects non-browser clients.

https://prism-redfern.org/topics/youngearth.html

topic: sediment-budget, crust-recycling camp: counter-apologetics type: blog access: blog timescale: annual rates extrapolated direction: mixed

Key figures: average ocean sediment depth "closer to 1500 meters," against Humphreys' 400 m, with no source given; sediment removed by subduction about 5 × 10⁸ m³/yr, derived from 1 cm/yr of subduction times 1,500 m of thickness times a trench length of order the earth's circumference; time to remove that sediment "on the order of one billion years"; an implied continental erosion rate of "about 0.004 millimeter per year," which is 4 m/Myr, described as close to "the accepted value of about 0.005 millimeter per year."

Half right and half circular, and the circular half is the more instructive.  Redfern is an emeritus professor of physics who describes himself as a former geology instructor.  His good argument is negative feedback: if sediment accumulates in the oceans faster than removal, then either erosion must slow as relief is destroyed or removal must speed up, so an imbalance measured today does not license an imbalance through all of geologic time.  That is a real physical constraint the steady-state framing ignores.  His illustration is memorable — treating the imbalance as a clock is like dating a car by the fact that fuel leaves the tank faster than it enters.

The self-consistency check that follows is not a check.  He derives 0.004 mm/yr from his own subduction estimate and then presents the agreement as confirmation of that estimate.  Worse, 0.005 mm/yr is not the accepted continental denudation rate.  Portenga and Bierman's compilation of 1,599 cosmogenic beryllium-10 measurements puts the drainage-basin median at 54 m/Myr and the mean at 218 m/Myr; 5 m/Myr is essentially their bare-outcrop median of 5.4 m/Myr, which describes the slowest-eroding surfaces on earth.  Using an outcrop rate for a whole-continent sediment budget understates denudation by about a factor of ten — and that is exactly the error needed to make his budget balance.  The same 0.005 mm/yr figure appears in the University of South Dakota essay above, from a different source, which means two independent skeptical pages are using the wrong statistic.  On salt Redfern simply declines: "I'll have to defer to the chemists on this one."

101 evidences for a young age of the Earth and the universe — items 32 and 33

RationalWiki, community page; the site blocks plain HTTP clients.

https://rationalwiki.org/wiki/101_evidences_for_a_young_age_of_the_Earth_and_the_universe

topic: isostasy camp: counter-apologetics type: review access: OA timescale: n/a direction: undercuts

Key figures: none in either item.  Item 32 answers Beachy Head losing a metre of coast every six years; item 33 answers Tas Walker's vertical continental erosion argument.  The single citation supporting item 33 is a Britannica entry on Leopold von Buch.

Directionally right, badly supported, and worth cataloguing because it is what a curious reader finds first.  Item 32 is sound: coastline lost at one place is deposited at another, and Rye in the same English county was a port and is now inland.  Item 33 reaches for isostatic compensation, which is the correct first-order answer, and then makes three mistakes in a single paragraph.

It illustrates the mechanism with the post-1810 rise of southern Sweden and the Baltic, which is glacio-isostatic rebound from the melting of the Weichselian ice sheet.  That is the same physics but not the same process, and nobody argues that Scandinavia is rising because it is being eroded.  It gives no multiplier — Holmes (1945), quoted by Schumm (1963), puts it at about four feet of rock eroded for every foot of surface lowering, and that number would settle the argument if anyone quoted it.  And it says the continents are lighter than "the underlying crust (the asthenosphere)," which is simply wrong: the asthenosphere is mantle, not crust.  A young-earth reader will spot that in one sentence and will be entitled to discount everything around it.

The Depths of the Oceans: Incompatible with a Global Flood Model

Meert, Joseph.  Personal site, created 5 April 2002, updated October 2005; the domain is now parked and the page survives only in the Internet Archive.

http://web.archive.org/web/20231119223730/http://gondwanaresearch.com/oceans.htm

topic: crust-recycling camp: mainstream type: model access: blog timescale: 10^8^ yr of seafloor cooling direction: undercuts

Key figures: to open a 4,000 km ocean inside a flood year requires a spreading rate of 4 × 10⁸ cm/yr, which Meert glosses as about 750 cm per minute and which reproduces at 760 cm/min; modern spreading for comparison 5 cm/yr; matching the observed bathymetric profile under catastrophic spreading requires raising thermal diffusivity by a factor of 10⁵, which he calls "only a mathematical trick."

Not an erosion argument, and included because the standard answer to the sediment-budget version of the erosion argument is that the Wilson cycle recycles the sediment — and the Wilson cycle only works at ordinary rates.  Meert, a professional academic geologist, takes the Parsons and Sclater conductive cooling law, under which seafloor depth increases with the square root of distance from the ridge, and runs it at catastrophic plate tectonics spreading rates.  The predicted bathymetry does not match observation under any physically justifiable parameter choice.  Rapid spreading would leave the floor essentially flat, because convective heat transport is fast, and the observed square-root profile records slow conductive cooling over tens of millions of years.

Handle with the caveats stated. This is a model, not a measurement.  It was published on a partisan personal website and never peer reviewed, and it is now reachable only through the Internet Archive.  It is also worth noting that Meert has no essay on continental erosion rates at all; his eleven creationism pages cover paleosols, ocean depth, magnetic fields and radiometric dating, and the erosion clock is simply absent.

Common Creationist Attacks on Geology

Weber, Christopher Gregory.  Creation/Evolution 2:10-25  (1980).  Reprinted by the National Center for Science Education.

https://ncse.ngo/common-creationist-attacks-geology

topic: sediment-budget camp: counter-apologetics type: review access: OA timescale: 10^8^ yr of Gulf Coast deposition direction: undercuts

Key figures: the Gulf Coast sediment pile reaches up to 10,000 m thick beneath the Great Plains, the Gulf Coast and the continental shelf, accumulating from Cretaceous time to the present.

The origin of the delta-avulsion answer that later skeptical pages compress into a sentence, and still the clearest statement of it.  The Mississippi's current delta is a few thousand years old, but the river jumps its banks every so often, reaches the Gulf by a new path and starts a new delta somewhere else — so the age of the present delta is the age of the present channel, not of the river or the earth.  Weber's second argument is facies: sand, clay and carbonate zones interfinger laterally in the pattern a shoreline produces, and volcanic ash beds slice across those boundaries, which is what a long-lived coastal system looks like and not what a single depositional event looks like.

One weakness to record, because it is the assumption a flood geologist denies outright: Weber argues that because the sediment types are so similar from top to bottom, "the rates of deposition could not possibly have been much faster than the rates of today."  That is an inference from lithologic similarity, not a measurement, and it is precisely the point in dispute.

The age of the Earth — how real science happens

Braterman, Paul.  Primate's Progress, 4 April 2013.

https://paulbraterman.wordpress.com/2013/04/04/the-age-of-the-earth-how-real-science-happens/

topic: history camp: counter-apologetics type: blog access: blog timescale: n/a direction: background

Key figures: the Victorian hourglass methods — total ocean salt divided by annual river delivery, and total sediment thickness divided by an estimated deposition rate — converged on about 100 Myr, against a modern value of 4.54 Gyr.

Background, and the right historical frame for the whole subject.  Braterman points out that the salt clock and the sediment clock are not creationist inventions at all.  They are nineteenth-century mainstream geology, used seriously by people trying to date the earth honestly, and they were wrong by a factor of forty to fifty.  Geology abandoned them when radiometric dating produced a number that independent lines of evidence supported and these methods did not — not because anyone first proved the rates variable.

That history cuts in two directions and the article should say both.  It warns that stock-divided-by-flux methods have a track record of failing, and that the failure mode is always the same: an unaccounted removal term or a non-constant input.  It also means the erosion argument cannot be dismissed as fringe reasoning.  It is a Victorian method still in use, and answering it requires showing which term is missing rather than pointing at who is making the argument.

Companion pages

• Erosion Rates and the Age of the Earth — the article this register supports.
• Natural Arches, Rock Bridges, and the Age of the Earth — the same style of argument run on a landform that really is young, with its own literature register.