Stromatolites and Banded Iron Formations

"The evidence seems to favor the interpretation that most of the Chuar Group is pre-Flood (most representing, perhaps, the first 1,656 years after Creation), with only the uppermost strata representing redeposition by the initial upheaval beginning the Flood." — Steven Austin, Grand Canyon: Monument to Catastrophe, 1994¹

"Long-term vertical growth rates of <0.4 mm/year were determined by radiocarbon dating and supported by uranium-series methods." — Allan Chivas and colleagues on the living stromatolites of Shark Bay, 1990²

"There are, as yet, no data to support that BIF precipitation was linked directly to microbial activity." — Cornelis Klein, presidential address to the Mineralogical Society of America, 2005³

Summary

A stromatolite is a layered mound of carbonate rock.  Most geologists read it as the fossil of a microbial mat that trapped sediment or made minerals grow, one thin lamina at a time.  A banded iron formation, or BIF, is a rock made of alternating iron-oxide and silica layers, with almost no mud in it.  Both are found mostly in Precambrian rock, the rock beneath the lowest fossil-bearing strata that Flood geology assigns to the Flood.  This article asks three things.  Where do young-earth models put these rocks?  What does each placement require them to have done, and how fast?  And what do the rocks say about the ocean and air they formed in?

The verdict is that no published young-earth placement survives its own numbers.  The decisive case uses no mainstream chronology at all.  It takes a creationist geologist's field description of one stromatolite bed, his own date for it, and one measured modern growth rate.

Eight conclusions, stated first.

1. The young-earth literature has no single placement for these rocks. It has five, spanning its whole timeline. Steven Austin's 1994 Grand Canyon monograph puts the Grand Canyon Supergroup in Creation Week Day Three and the pre-Flood era, "perhaps, the first 1,656 years after Creation."¹ Andrew Snelling's 2009 textbook puts the first stromatolites on Day Three and the great BIFs at the Day Three to Day Four transition, about 48 hours.⁴⁴ Kurt Wise in 2002 makes the stromatolite reefs pre-Flood and then "the first Flood deposits."⁵ Harry Dickens in 2017 puts the big BIFs on Days One and Two.⁶ Timothy Clarey in 2026 moves the Flood's onset below the Supergroup, so the same rocks fall inside the Flood year.⁷ Two of these five authors are the sitting research directors of ICR and AiG, and they are the two furthest apart. There is no one model to test. Each owes its own arithmetic.
2. Austin's own stromatolite bed cannot grow in Austin's own window. He describes at the base of the Awatubi Member "a massive limestone stromatolite layer 12 feet thick," 3.66 metres, with domes and branching columns.¹ The living stromatolites of Shark Bay, radiocarbon-dated by Chivas and colleagues in 1990, grow at 0.18 to 0.34 mm a year.² Dividing 3,658 mm by those rates gives 10,800 to 20,300 years for that one bed. Austin allows the whole Chuar Group 1,656 years. The arithmetic is this article's own, and it contains no radiometric date.
3. The rest of the Chuar makes it worse. Austin gives the group "a measured maximum thickness of 6,610 feet," 2,015 metres.¹ Spread across 1,656 years, that is 1.22 metres of sediment a year, every year, in rock the Park Service describes as shallow-water carbonate and mudstone with exposure surfaces.⁸
4. The Hamersley iron formation has a thickness clock, and it needs no astronomical tuning. The Hamersley Group is 2.5 km thick and dated between 2.63 and 2.45 billion years.⁹ Those two numbers alone give 13.9 metres per million years. Lantink and colleagues in 2022 measured 5 to 15 metres per million years inside one member of it by two independent methods.¹⁰ A pre-Flood placement needs 110,000 times that rate; a Flood-year placement, 180 million times; a Creation-Week placement, 33 billion times.
5. The one paper the creationist "rapid BIF" claim rests on says 30 to 1,000 metres per million years. Barley, Pickard and Sylvester 1997 is the citation behind Snelling 2009, Dickens and Snelling 2008 and Dickens 2017. Its fastest rate, 1,000 metres per million years, is one millimetre a year.¹¹ Nothing in it supports hours, days, or a year.
6. The rocks require an ocean and an atmosphere unlike the present. The largest BIFs hold "over 10^14 tons of iron."⁹ Iron dissolves in seawater only when there is no oxygen, and the field abandoned rivers as the source because it would take "rivers the size of the modern Amazon" carrying orders of magnitude more iron than they do.⁹ Sulfur isotopes in rocks older than 2.45 billion years carry a mass-independent signal that forms only when the air has essentially no oxygen and no ozone, "probably less than 0.001% of the present atmospheric level."¹² Every young-earth placement puts that atmosphere either in Creation Week, days before God fills the sea and sky with air-breathing life, or over Adam's head for 1,650 years, or over Noah's ark.
7. The young-earth side has real points, and they are granted. Whether the oldest stromatolites are biological is contested inside the field. Grotzinger and Knoll say biogenicity "cannot be easily demonstrated" at the outcrop.¹³ A 3.7-billion-year stromatolite claim was overturned in 2020 by a re-mapping of the outcrop, though its authors have not withdrawn it.¹⁴ Klein says there are "no data" tying BIF precipitation to microbes.³ Counting laminae as years is unreliable.¹⁵ Austin made the biogenicity caution honestly in 1994.¹ None of it shortens the clock. The mainstream alternative to biology is quiet in-place chemical precipitation on a still sea floor, which is harder to fit in a Flood, and the mainstream dispute about Archean oxygen is a dispute about the air at 3.5 billion years, with both sides 3.5 billion years from Genesis 1.
8. The 1994 Catastrophic Plate Tectonics model and Clarey's 2026 move each trade one impossible requirement for another. The CPT authors, Snelling and Wise among them, need a pre-Flood ocean full of fine detrital carbonate to make the Flood limestones.¹⁶ The Precambrian carbonates and BIFs record a clean ocean: BIF aluminium oxide runs 0.09 to 1.8 percent by weight, "devoid of detrital input."³ Clarey's move into the Flood year removes the pre-Flood growth problem and replaces it with 6.7 metres of iron formation a day.

The thing that is actually surprising

The clearest refutation in this article comes from a creationist book, not a mainstream one.  Austin described the bed, measured the group, and dated the whole thing to 1,656 years.  He did not divide.  Nobody on the young-earth side has published an accretion rate for a stromatolite, an iron mass balance for a BIF, or an oxidant budget for either, in thirty years of writing about these rocks.  The numbers were always available.  This article runs them.

Part 1: What the rocks are, and where young-earth geologists put them

Grotzinger and Knoll define a stromatolite without mentioning life: "Stromatolites are attached, lithified sedimentary growth structures, accretionary away from a point or limited surface of initiation."¹³  The definition is neutral on purpose, because what built them is still argued.  The laminae of Archean and early Proterozoic stromatolites mostly precipitated in place from seawater.  Later Proterozoic ones mostly grew as microbial mats trapped loose carbonate grains.¹³  Either way, they dominate the early carbonate record: "Rocks deposited during the first 85 percent of Earth history tell a different story, one in which stromatolites are the principal features of platform and shelf carbonates."¹³  Stromatolitic units "extend for hundreds of meters vertically and hundreds of kilometers laterally."¹³

An iron formation is a chemical sediment with at least 15 percent iron, usually layered with chert.  Cornelis Klein's 2005 presidential address to the Mineralogical Society of America gives the record in one sentence: iron formations begin at 3.8 billion years at Isua, reach their largest volume at about 2.5 billion years in the Hamersley Basin, "disappear from the geologic record at about 1.8 Ga, only to reappear between 0.8 and 0.6 Ga."³  The Hamersley Group of Western Australia is the type case.  Konhauser and twelve co-authors describe it as "the 2.5-km-thick Hamersley Group (dated between 2.63 and 2.45 Ga)."⁹  Its Brockman Iron Formation holds the largest iron formations by mass on Earth.⁹  The largest Superior-type formations contain "over 10^14 tons of iron"⁹ over areas "estimated in some cases to be over 100,000 km2."⁹  The Marra Mamba Formation alone "has an average outcrop thickness of about 210 m."⁹  Harry Dickens, the young-earth side's BIF specialist, gives the same mass in his own units: "Giant (100,000 billion tonnes or more) BIFs are located in South Africa, Australia, Brazil, Russia, and Canada."⁶

None of these rocks holds a shelly fossil.  In the Grand Canyon they sit in the tilted Supergroup below the Great Unconformity, where stromatolites are known from four formations.⁸  That puts them under the level where most Flood geologists start the Flood.  Each model has to decide what to do with them, and the decisions do not agree.

Five placements

Two cautions about this table.  Austin's 1994 book is about the Grand Canyon, which has no banded iron formation.  He is cited here for the stromatolites of the Chuar Group, and nothing he wrote is a young-earth account of the Hamersley.  Clarey's 2026 piece is a proposal, hedged with "possibly" and "may have," not an ICR position statement.⁷  It is included because it is the most recent word from ICR's director of research, and because it moves the same rocks Austin's ICR monograph called pre-Flood into the Flood year.

The spread matters more than any single entry.  Snelling, who directs research at Answers in Genesis, gives the Hamersley two days of Creation Week.  Clarey, who directs research at ICR, gives the Proterozoic a Flood year.  Austin and Wise give the stromatolites a pre-Flood era.  Snelling and Wise also signed the 1994 plate-tectonics paper that called the Archean sediments pre-Flood, a position Snelling's 2009 book abandons.  Nothing in any of the five computes how fast the rock would have to form.  So the article does it for each.

Part 2: What the young-earth side gets right

Five of the creationist claims about these rocks hold up.  They are stated here first, with their mainstream support, because a reader who meets them later as "objections" will suspect they were hidden.

Biogenicity is contested inside the field.  Grotzinger and Knoll quote Hoffman's line from 1973 as still true in 1999: "Something that haunts geologists working on ancient stromatolites is the thought that they might not be biogenic at all."¹³  Their own rule is strict: "Biogenicity cannot be easily demonstrated by relationships observed at the outcrop scale; it is essential to examine lamination textures petrographically and demonstrate the presence of textures uniquely attributable to the presence of microbial mats or biofilms."¹³  Bosak, Knoll and Petroff said in 2013 that the problem is unsolved, naming "the problematic distinction between biogenic and abiotic structures whose laminae accrete exclusively by mineral precipitation."¹⁷  Allwood's team, who made the best case for the 3.43-billion-year Strelley Pool stromatolites, wrote: "For simplicity we adopt the term 'stromatolite' (a term implying biogenesis) but acknowledge the as yet unproven biogenicity."¹⁸  In 2020 a team that included Allwood re-examined the 3.7-billion-year Isua "stromatolites" announced in 2016 and found "a tectonic fabric that preserves no fine-scale primary sedimentary structures of what were probably never stromatolites."¹⁴  For iron formations, Klein is blunter: "There are, as yet, no data to support that BIF precipitation was linked directly to microbial activity."³

The creationist literature has said this all along, and said it fairly.  Austin's 1994 Grand Canyon book lists the abiotic look-alikes: "numerous laminated structures are known to form without biological activity.  Many precipitated materials (travertine, tufa, caliche, onyx, etc.) have laminated structures, some of which closely mimic stromatolites."¹  He quotes Nitecki's 1971 paper against the Grand Canyon's own oldest fossils in the Bass Limestone: "There are a number of distinct structures in the Bass Limestone, but none of these can be considered unquestionably organic."¹  Dickens and Snelling in 2008 wrote that "Archean stromatolite-like structures may lack evidence of the microorganisms which are supposed to have constructed them" and that "Abiotic models for the origin of Archean stromatolite-like structures have been put forward."¹⁹  All of that is correct.

Laminae are not years.  Reid and eleven co-authors watched the Bahamian stromatolites at Highborne Cay for three years and found that a lamina records a microbial succession, not a calendar: "growth of modern marine stromatolites represents a dynamic balance between sedimentation and intermittent lithification of cyanobacterial mats.  Periods of rapid sediment accretion, during which stromatolite surfaces are dominated by pioneer communities of gliding filamentous cyanobacteria, alternate with hiatal intervals."¹⁵  In the fast phase "accretion rates of one grain-layer per day produce mats with Type 1 fabrics."¹⁵  Klein reports that the Brockman microbands have "been interpreted as chemical varves, or annual layers of sedimentation,"³ and Lantink's team shows the two published readings of the Weeli Wolli microbands, 23 per bundle as annual layers or 28 to 30 as lunar fortnights, cannot both be right: "at least one of these interpretations must be incorrect, illustrating the large uncertainty involved in the Weeli Wolli reconstructions lacking independent control on sedimentation rate."¹⁰  Anyone who counts Precambrian laminae as years and reports the total as an age is doing something the field itself distrusts.

The rocks are not uniformitarian.  Dickens writes: "BIFs are clear-cut examples of non-uniformitarianism in the earth's history; modern analogues are unknown and BIFs are restricted in time to the Archean, Paleoproterozoic, and Neoproterozoic."⁶  Klein's timeline says the same.³  Konhauser reports Isley and Abbott's estimate that "perhaps as much as 60% of the global volume of preserved IF was deposited" in one 100-million-year window.⁹  Barley, Pickard and Sylvester tied the 2.45-billion-year Hamersley BIFs to a large igneous province, and Snelling was right to cite them for that link.¹¹  The Precambrian ocean was not the modern ocean.  No mainstream author says it was.

Archean oxygen is disputed by real geochemists.  Michael Oard's 2015 article cites Watanabe, Farquhar and Ohmoto's 2009 laboratory work, in which anomalous sulfur fractionation "can be caused by reactions between powders of amino acids and sulfate at temperatures between 150 and 200°C.  So, the deduction from sulfur isotopes is now equivocal."²⁰  James Farquhar, a co-author of that paper, is the man who discovered the sulfur signal in 2000.  Hiroshi Ohmoto has argued for an oxygenated Archean for decades.  This is not a fringe.  Oard also notes that the iron in a BIF is oxidised iron, which needs an oxidant.²⁰  And Konhauser reports that in Kabuno Bay, Congo, iron-oxidising phototrophs work at rates that "appear to exceed those required for IF deposition."⁹  A creationist can say, correctly, that oxidising the iron fast was never the hard part.

Modern stromatolites are small, few, and slow.  Hamelin Pool at Shark Bay is "the largest modern stromatolite system in the world," about 1,400 km² with 135 km of shoreline, and it has been growing "for the past 2000 years."²¹²¹  Its columns are about 350 mm tall.²  Precambrian stromatolitic platforms run hundreds of metres thick and hundreds of kilometres across.¹³  Wise's point that the Precambrian world must have differed from ours is the field's own conclusion.

None of these five points shortens the clock.  Biogenicity is not what the clock measures.  Lamina-counting is not the method used below.  Non-uniform in time is not the same as fast.  The oxygen dispute is over what the air was like 3.5 billion years ago, not over whether it was 3.5 billion years ago.  And the gap between a modern column and a Precambrian platform is the size of the problem, not its solution.

Part 3: The stromatolite clock

The only measured rate

Only one kind of number can turn a stromatolite's thickness into a time: a measured accretion rate on a stromatolite that is actually lithifying.  Chivas, Torgersen and Polach supplied it in 1990.  They radiocarbon-dated five columns from Hamelin Pool and reported: "Long-term vertical growth rates of <0.4 mm/year were determined by radiocarbon dating and supported by uranium-series methods.  These rates are up to 250 times slower than those growth rates estimated from the laminae patterns of some Precambrian stromatolites.  Thus, daily bands are not recognized and individual subtidal stromatolite columns from Shark Bay take up to 1000 years to reach their present heights of ~ 350 mm."²  The five rates were 0.34, 0.28, 0.29, 0.19 and 0.18 mm a year.²  Their summary: "Modern stromatolitic accretion is struggling against erosion, with net vertical growth rates of <0.4 mm/year."²

Reid's Bahamian stromatolites, in open ocean water of normal salinity, tell the same story from a different site.  They "form ridges and columnar heads up to half a metre high,"¹⁵ and their lithified crusts "form at intervals averaging 1-2 mm."¹⁵  Fast burial happens, one grain-layer a day, but a crust forms only in the pauses, when a boring climax community fuses the grains.  Take away the pauses and the product is loose sand, not a laminated column.  Bosak, Knoll and Petroff give the field's own summary: "Macroscopic stromatolites take a long time to grow under realistic conditions (years to hundreds or thousands of years), outlasting those who study these structures."¹⁷

Chivas gives the young-earth side its best number in the same abstract.  Rates "estimated from the laminae patterns of some Precambrian stromatolites" run up to 250 times faster than what he measured.  That is about 100 mm a year.  Nobody has measured a lithifying stromatolite growing anywhere near it, and the estimate comes from the lamina-counting the last section showed to be unreliable.  It is used below anyway, as a ceiling, because an argument that grants the opponent his best number is the one that holds.

Austin's bed

Austin describes the base of the Awatubi Member of the Kwagunt Formation, in the Chuar Group of the Grand Canyon Supergroup: "The bottom of the Awatubi Member contains a massive limestone stromatolite layer 12 feet thick, with domes up to 12 feet in height and width, and with branching columns possessing thin laminae."¹  Twelve feet is 3,658 mm.  On the same page he dates the group: "most of the Chuar Group is pre-Flood (most representing, perhaps, the first 1,656 years after Creation)."¹

The arithmetic, which is this article's and not Austin's:

At any measured rate the bed needs six to twelve times the whole window Austin allows the group.  No radiometric date enters.  No old-earth premise enters.  The thickness, the bed's description, and the 1,656 years are all Austin's.  The rate is a radiocarbon measurement on living carbonate.

The Chuar is Neoproterozoic, deposited after 782 million years ago in conventional terms.⁸  That matters for the choice of rate.  Grotzinger and Knoll's secular trend puts Neoproterozoic stromatolites in the trapping-and-binding class, the very process Chivas and Reid measured.¹³  The abiotic-precipitation escape, which is real for Archean crystal-fan stromatolites, is not available for Austin's bed.  His own description, "branching columns possessing thin laminae," is a trapped-and-bound column.

The rest of the group

The bed is one layer.  Austin gives the whole Chuar Group "a measured maximum thickness of 6,610 feet,"¹ which is 2,015 m.  Spread across 1,656 years, that is 1.22 m of sediment a year, every year, in a unit the Park Service's stratigraphic inventory describes as laid down in "shallow subtidal to intertidal settings, with more frequent episodes of subaerial exposure" than the formation below it.⁸  Exposure surfaces are pauses.  Each one is time subtracted from the 1,656 years and added to the rate required of the rest.  And the Chuar is only the top of the Supergroup: the same inventory records stromatolites in four formations, the Bass and Dox of the Unkar Group below the Chuar as well as the Chuar's own Galeros and Kwagunt, the Bass being "the oldest fossils known from the park … between 1.25 and 1.2 billion years old."⁸  On Austin's chronology the Bass belongs to Day Three itself.  Its stromatolites, if biological, grew in the hours between the making of dry land and the making of the sun.

Wise's reef and Snelling's day

Wise gives his reefs at most the pre-Flood era, about 1,650 years.  At 0.4 mm a year that is 660 mm of stromatolite, about two Shark Bay columns, against platforms "hundreds of meters vertically."¹³  His model does make one testable prediction: the reefs grew at hot springs, and "Sites examined by the author tend to have a lot of evidence of hot spring activity and seem to be located near the margins of pre-Flood continents."⁵  The best-studied Archean stromatolite platform tests it.  At Strelley Pool, Allwood's team found that "Onset of stromatolite development coincided with the end of siliciclastic sediment influx, and the end of stromatolite formation coincided with the resumption of siliciclastic sedimentation, hydrothermal activity and volcanism."¹⁸  Hydrothermal activity ended the reef.  It did not build it.

Snelling's Creation Week placement leaves the arithmetic with nothing to divide.  His first stromatolites belong to "the middle or latter part of the third day."⁴  The Strelley Pool platform they include is, in Allwood's description, "a layer of laminated stromatolitic carbonate/chert about 10-20 m thick, subdivided into three beds,"¹⁸ with members 2 to 4 merging southward into a chert succession more than 70 m thick.  The three beds record three regressive cycles, each capped by evaporite crystal pseudomorphs.  Bed 1's cap is 0.5 to 1.8 m thick, and its upper contact carries erosion surfaces, solution cavities and collapse breccia, which Allwood's team labels karst.  Karst is fresh water dissolving exposed rock.  Three stacked cycles of growth, drying, exposure and dissolution in part of one day is not a rate problem.  It is a sequence that a day cannot contain.

Part 4: The iron clock

Three routes to one rate

The Hamersley Group's own description gives a rate before any cycle-counting begins.  Konhauser's 2.5 km of section between 2.63 and 2.45 billion years⁹ is 2,500 m in 180 million years: 13.9 m per million years.  The bracket comes from uranium-lead zircon dates on volcanic layers, not from the iron bands.  Barley's SHRIMP zircon ages on a tuff band and a rhyolite inside the group, 2,470 ± 4 and 2,449 ± 3 million years, are of that kind.¹¹

Lantink, Davies, Ovtcharova and Hilgen measured the rate a second way inside one member of the same group.  They dated four shale horizons in a drill core through the Joffre Member of the Brockman Iron Formation, obtaining ages from 2,469.1 to 2,454.0 million years in stratigraphic order, and those dates "allow the calculation of a mean depositional rate of ∼5 m/My to 15 m/My."¹⁰  Then a third way: the band thicknesses themselves.  The 47 m of Joffre outcrop shows about 8.9 cm cycles bundled by nines into 85 cm couplets, bundled by fours into 3.3 m cycles, ratios that match climatic precession nested in short and long orbital eccentricity.¹⁰  The rate the orbital reading implies is the rate the zircons give.  And "our new depositional rate estimate for the Joffre Member is identical to the previously established 10 m/My mean rate of the lithologically similar and coeval Kuruman IF in South Africa."¹⁰  A whole-group radiometric bracket, a single-member zircon series, and a band-thickness hierarchy, none tuned to the others, agree on metres per million years.

The paper the creationist claim rests on

Barley, Pickard and Sylvester 1997 is cited for rapid BIF deposition by Dickens and Snelling in 2008, by Snelling's 2009 textbook, and by Dickens in 2017.  Here is what it says, on p. 58: "more than 500 m of BIF and ~100 m of shale of the Brockman and Woongarra supersequences must have been deposited in the ~20-Myr period between 2,470 ± 4-Myr Brockman Supersequence tuff band and the 2,449 ± 3-Myr Woongarra Rhyolite."¹¹  From that: "As BIFs are compacted and lithified sedimentary rocks, depositional rates for uncompacted Hamersley Province BIFs must have been faster than 30 m Myr⁻¹.  This can be compared to the rate of 40 m Myr⁻¹ estimated for modern uncompacted pelagic sediments."  And at the outside, "the Hamersley Province BIFs may have been deposited at even greater rates (between 100 and 1,000 m Myr⁻¹), as required by models which infer that paired chert–iron oxide laminae are seasonal deposits or varves."¹¹  One thousand metres per million years is one millimetre a year.  Barley's "rapid" is rapid compared with an earlier estimate of 3 to 4 m per million years.  It is slower than modern deep-sea ooze.

Dickens and Snelling report this accurately and leave the word to do the work: "The case has been put for the 'rapid' (similar to or faster than the rate of deposition of modern pelagic sediments) rather than gradual (millions of years) origin of BIFs."¹⁹  A reader who does not open Barley hears "rapid" as catastrophic.  Snelling's 2009 book goes further and cites the same paper for ocean waters that "couldn't hold these in solution and thus rapidly deposited them,"⁴ which Barley does not say.  Dickens's 2017 claim that deposition "may have only lasted a few hours to days!"⁶ does not rest on Barley at all.  It rests on reference 9, Lascelles 2013, "Plate tectonics caused the demise of banded iron formations."  Only that paper's abstract could be obtained, and the abstract does not say hours or days.²²  It argues that hydrothermal plumes precipitate iron on quenching and are then "subsequently resedimented by turbidity and density currents across the ocean floor," and that the temporal distribution of BIFs reflects which sea-floor deposits survived subduction over the Archaean and Proterozoic eras.  Dickens's own sentence paraphrases Lascelles's precipitation clause almost word for word.²²  The hours-to-days figure appears to be Dickens's extrapolation from the resedimentation step, not a Lascelles result.  Whether the paper's body puts a number on it is still open, because the full text is closed access.

The required rate for each model

Take the 2.5 km Hamersley Group and the measured 13.9 m per million years.  Working: 2,500 m ÷ 180 Myr = 1.39 × 10⁻⁵ m/yr; 2,500 m ÷ 1,656 yr = 1.51 m/yr; 2,500 m ÷ (371/365.25) yr = 2,461 m/yr; 2,500 m ÷ (2/365.25) yr = 4.57 × 10⁵ m/yr.

The 2.5 km includes shale and volcanics as well as iron formation, so this is a whole-group average.  Lantink's 5 to 15 m per million years inside one BIF member is the cleaner number, and it is the same number.

Where the iron comes from

The rate is only half the problem.  The largest Superior-type formations hold "over 10^14 tons of iron."⁹  That is 10¹⁷ kg, or 1.79 × 10¹⁸ moles of iron.  Iron stays dissolved in seawater only as ferrous iron, and only without oxygen.  Konhauser's review gives the estimated Archean seawater ferrous-iron range as 0.03 to 0.5 millimolar, against 0.05 to 2 nanomolar today.⁹  At the most generous end, 0.5 mM, delivering 1.79 × 10¹⁸ moles means stripping every iron atom from 3.6 × 10¹⁸ m³ of seawater, 2.7 times the 1.335 × 10¹⁸ m³ of the modern ocean.  At 0.03 mM it is 45 ocean volumes.  Konhauser names five such formations.⁹

Rivers cannot do it.  Trendall and Blockley's iron flux for the Hamersley was of order 10¹³ g a year, "requiring rivers the size of the modern Amazon to transport orders of magnitude more iron than they do today.  Furthermore, IFs typically lack significant detrital input, which argued against a riverine source of the iron."⁹  So the field moved to a hydrothermal source feeding a whole anoxic ocean.  Klein reads the rare-earth signature the same way: "the source of Fe and Si was the result of deep ocean hydrothermal activity admixed with sea water."³

Now the oxidant.  Bekker and co-authors give the largest flux any published mechanism supplies, ultraviolet photo-oxidation: "this process alone could have accounted for deposition of as much as 1.8 × 10^14 mol Fe(III) annually (François, 1986).  Other estimates place the total amount of Fe(II) photo-oxidized annually at 2.3 × 10^13 mol (Braterman and Cairns-Smith, 1986).  These rates are much greater than annual rates inferred during deposition of the largest Archean and Paleoproterozoic BIFs."²³  Against 1.79 × 10¹⁸ moles, that ceiling gives: 1,656 pre-Flood years, 1.09 × 10¹⁵ mol/yr, six times the ceiling; one Flood year, 1.79 × 10¹⁸ mol/yr, about 9,800 times; 48 hours of Creation Week, 3.3 × 10²⁰ mol/yr, about 1.8 million times.  Barley's 20 million years needs 8.9 × 10¹⁰ mol/yr, one twentieth of one percent of the ceiling.

The Kabuno Bay observation answers none of this.  Photoferrotrophs there oxidise iron faster than a BIF requires,⁹ but the bottleneck was never the oxidation step.  It is the supply of dissolved ferrous iron, which is set by ocean volume and solubility, and no biology changes either.

Bands from a beaker

Dickens's mechanism for the banding is self-organisation in a colloid: "Laboratory studies show that colloidal solutions rapidly precipitate into regular and ordered bands."⁶  Nobody disputes that Liesegang bands form fast without life.  But a diffusion cell gives a spacing law, not a hierarchy.  What Lantink measured is 8.9 cm bands, 8.9 to a bundle, four bundles to a cycle, in a 47 m section whose top and bottom are independently dated 15 million years apart.  A Liesegang model that reproduces those ratios with published parameters has never been run.  It is a computation a creationist could do.

Part 5: The ocean and the air the rocks record

No mud

Across two billion years of iron formation, Klein reports, "Al2O3 contents are very low, ranging from 0.09 to 1.8 wt%.  These chemical values show that they are clean chemical sediments devoid of detrital input."³  Konhauser's compilation for deep-water iron formation: "low concentrations of Al2O3 (<1 wt.%) and trace elements that are enriched in crustal rocks (Ti, Zr, Th, Hf, and Sc <20 ppm)."⁹  A Flood competent to strip continents to their basement and bury them in kilometres of sand and mud cannot, in the same year, lay down hundreds of metres of chemical sediment over 100,000 km² with no continental detritus in it.  Dickens's own mechanism, turbidity and density currents off a rifting crust, is a mud-carrying mechanism.  He quotes Klein's "lack of detrital components" approvingly, as proof of chemical precipitation, without noticing it is a constraint on his model.

The 1994 Catastrophic Plate Tectonics paper has the opposite requirement.  Its six authors, Snelling and Wise among them, "believe that substantial quantities of very fine detrital carbonate sediment existed in the pre-Flood oceans.  This is deduced primarily from the fact that not enough bicarbonate can have been dissolved in the pre-Flood ocean (and/or provided by outgassing during the Flood—see below) to have produced the Flood carbonates."¹⁶  The model needs a muddy carbonate ocean before the Flood.  The Precambrian carbonates and iron formations, which the same paper calls pre-Flood, record a clean one.  Grotzinger and Knoll's Archean laminae precipitated in place on a still sea floor, and Klein's BIFs took in no detritus.  The same authors need the same ocean to be a detrital reservoir and a quiet chemical precipitator at once.  This series' chalk and limestone article covers the Flood-carbonate budget itself.

No oxygen

Klein: "All of the most 'primary' mineral assemblages appear to be the result of chemical precipitation under anoxic conditions."³  Barley's model has the largest BIFs forming "as these waters mixed with the more oxygen-rich shallow hydrosphere,"¹¹ which requires a deep ocean holding its iron in solution because it had none.  The atmosphere above it is recorded separately, by sulfur.

Farquhar, Bao and Thiemens measured four sulfur isotopes in Precambrian sulfides and sulfates in 2000.  Rocks younger than 2,090 million years show only the ordinary mass-dependent spread.  Rocks older than 2,450 million years show a mass-independent anomaly ten times larger than the intermediate interval.²⁴  Their conclusion: "Before 2450 Ma, the cycle was influenced by gas-phase atmospheric reactions.  These atmospheric reactions also played a role in determining the oxidation state of sulfur, implying that atmospheric oxygen partial pressures were low and that the roles of oxidative weathering and of microbial oxidation and reduction of sulfur were minimal."²⁴  Biology is excluded by the isotope geometry: "These arrays are inconsistent with biological fractionation processes."²⁴  Lyons, Reinhard and Planavsky put the requirement in a number: "For the signals to be generated and then preserved in the rock record requires extremely low atmospheric oxygen levels, probably less than 0.001% of the present atmospheric level (PAL)."¹²  With no oxygen there is no ozone, and Konhauser's review treats that as given: before the rise of oxygen "and the development of a protective ozone layer, Earth's surface would have been subjected to high levels of ultraviolet radiation."⁹  Konhauser lists five independent markers of the change at 2.4 billion years, and the sulfur signal is only one of them: detrital uraninite, pyrite and siderite vanishing from river sediments; iron leached from older soils; red beds appearing; the sulfur anomaly ending; chromium and uranium rising in iron formations.⁹

That atmosphere had no ozone and nothing in it could breathe.  Each placement has to put it somewhere.

• Snelling and Dickens put it in Creation Week, on Days One to Four, a day or two before God fills the sea and sky with air-breathing animals. Neither mentions the sulfur record. Snelling's volume 2 contains no discussion of mass-independent fractionation, the Great Oxidation Event, or an oxygen budget.
• Wise and the 1994 plate-tectonics paper put the Archean before the Flood, which puts the anoxic, ozone-free air over Adam for up to 1,650 years. The pre-Flood world of young-earth literature is usually described as oxygen-rich, not oxygen-free.
• Clarey puts the Proterozoic in the Flood year, so the sulfur-anomaly rocks are Flood rocks and the atmosphere is Noah's. His article mentions neither BIFs nor oxygen.
• Austin's Grand Canyon Supergroup is conventionally 1.25 to 0.73 billion years old, above the change. His model owes no answer here. It also offers none for the Hamersley.

The one creationist answer

Oard's reply is the Watanabe experiment: heated amino acids and sulfate produce an anomaly, "So, the deduction from sulfur isotopes is now equivocal."²⁰  Three tests decide whether that holds, and this project could not obtain the Watanabe paper to run them.  First, the laboratory anomaly must reproduce the ratio of the two anomalous isotopes seen in the Archean array, not just an anomaly.  Second, a reaction that runs at 150 to 200 °C in sediment runs in Phanerozoic sediment too, so it must explain why the signal stops at 2.3 billion years and never returns.  Third, it explains nothing about the other four markers.  The disagreement Oard cites is real, but its two sides are Ohmoto and Farquhar arguing about the air at 3.5 billion years.  Both are 3.5 billion years from Genesis 1.  Nobody in that dispute has proposed days.

Oard's second argument, that the oxidised iron in a BIF proves an oxygenated atmosphere, skips the alternatives the field exists to weigh.  Konhauser's review treats three oxidants: ultraviolet photo-oxidation, iron-oxidising phototrophs, and cyanobacterial oxygen.  The first two oxidise iron with no free oxygen at all.  Dickens and Snelling's 2008 claim that "the biosphere contained significant oxygen at the time of mineralisation"¹⁹ cites a 1999 paper on the Mount Tom Price ore body, an ore-enrichment study of an event a billion years younger than the Archean.  That paper was not obtained, and the claim is not characterised further here.

Objections and responses

"They are not biogenic, so the growth-rate argument does not apply."  Granted for the Archean crystal-fan stromatolites, and Part 2 says so.  But the mainstream alternative is not fast.  Grotzinger and Knoll's abiotic laminae precipitated in place from carbonate-saturated seawater on a stable sea floor: "On a sterile Earth, carbonates would still be removed from seawater, and they would likely form precipitated laminates in coastal environments."¹³  In-place precipitation forbids transport and needs a still water column, which is harder to fit in a Flood than a trapped-sediment mat, not easier.  And Austin's bed is Neoproterozoic, in the trapped-and-bound class, with branching laminated columns.  The abiotic escape does not reach it.

"God created the rocks with apparent age."  This ends the discussion rather than winning it.  Austin wrote of "sedimentary and tectonic processes which occurred during and after Day Three,"¹ Snelling of hot springs precipitating iron, Dickens of density currents.  All of them propose processes.  A created-in-place rock has no process, no rate, and no iron budget, and it also records a sulfur anomaly from an atmosphere that never existed.  No creationist geologist cited here takes this route, and the article does not answer a position none of them holds.

"A 12-foot bed is not 12 feet of vertical growth."  This is the strongest objection to Part 3, and Austin's own sentence answers it.  He describes the bed as carrying "domes up to 12 feet in height and width, and with branching columns possessing thin laminae."¹  A dome as tall as the bed is thick is one generation of upward accretion, not a stack of amalgamated thin ones.  If a reader still wants to treat the bed as several laterally shifting generations, the total carbonate accreted is unchanged and the required rate per unit area falls only in proportion to how much of the bed is lateral, which no published measurement supplies.

"Accretion was far faster in a warm, mineral-rich pre-Flood ocean."  Part 3 grants 250 times the measured rate, the ceiling Chivas himself reports, and Austin's bed still needs 37 years at a rate nobody has measured.  The group around it still needs 1.22 m a year for 1,656 years, with exposure surfaces inside it.  Reid's field work shows why speed does not help: fast burial gives loose sand, and a lithified crust forms only during a pause.¹⁵  The young-earth side has published no accretion rate for any stromatolite in thirty years of writing about them.  One measured rate above 1 mm a year on a lithifying stromatolite would change this section.  Travertine and tufa at hot springs are the right place to look for Wise's model.

"BIFs precipitated from Flood-year hydrothermal fluids."  Dickens's mechanism carries mud and the rock has none.  His hours-to-days figure rests on Lascelles 2013, whose abstract never mentions hours or days and whose own timescale is the Archaean and Proterozoic eras, and a bed redeposited by a density current in hours is not a formation deposited in hours.  The iron mass needs 2.7 to 45 ocean volumes of the most iron-rich seawater anyone proposes, oxidised at up to 1.8 million times the highest published flux.  And the band hierarchy Lantink measured matches orbital ratios that no reaction-diffusion cell produces.

"Barley 1997 says the BIFs formed rapidly."  It says 30 to 1,000 m per million years, slower than modern pelagic ooze.  The quotations are in Part 4.

"The BIF record is non-uniformitarian, and that favours catastrophe."  Non-uniform in time is not fast.  Klein's timeline, Konhauser's 60 percent in 100 million years, and Barley's igneous-province trigger are all mainstream, and all are measured in millions of years.

"Mainstream geochemists disagree about Archean oxygen, so the science is unsettled."  The disagreement is whether oxygen appeared at 3.0 or 2.4 billion years ago.  Lyons reports it plainly.  Neither side of it helps a chronology in thousands.

"The 2.5 km Hamersley thickness includes shale and lava, not just iron formation."  Granted.  That is why Lantink's 5 to 15 m per million years inside one iron-formation member is given beside it.  They agree.

A skeptic's objection: "The whole stromatolite argument rests on one hypersaline lagoon."  Hamelin Pool is the largest modern system on Earth, and Reid's open-marine Bahamian columns are half a metre high with crusts every millimetre or two.  Bosak's "years to hundreds or thousands of years" is the field's general statement.  No faster lithifying stromatolite has been measured anywhere.  The 250-fold ceiling covers the objection with room to spare.

A skeptic's objection: "Why argue with a 1994 book?  ICR has moved on."  Because Austin's book is the only young-earth source that describes a specific Precambrian stromatolite bed, gives its thickness, and dates it.  Clarey's 2026 move does not retract it; it makes the same rocks Flood rocks, and the required rate goes from 1.5 m a year to 6.7 m a day.

A skeptic's objection: "You lean on radiometric dates you elsewhere reject."  The stromatolite clock uses none.  The Hamersley thickness clock uses uranium-lead brackets, and the band-thickness hierarchy, which uses no dating, gives the same rate.  Discard both, and Barley's 30 to 1,000 m per million years, the creationists' own citation, still says the same thing.

A skeptic's objection: "The claim that nobody on the young-earth side has done the arithmetic is an argument from silence."  It is a report of the published record, checked by reading it.  Snelling's volume 2 was searched for an oxygen budget, a mass balance, and the sulfur signal, and has none.  Dickens 2017 cites Klein, Bekker and Trendall but no sulfur-isotope paper.  Clarey 2026 does not mention iron formations.  The calculation each needs is named above and is arithmetic.

A creationist's objection to this article's verdict: "The Bass Limestone stromatolites are photographed 'not in life position,' so they were transported."  The Park Service caption describes one loose block.⁸  The same chapter reports Bass stromatolites from in-place horizons and figures one just below the Hakatai Shale.  And a transported stromatolite still had to grow somewhere first.

Issues and unanswered questions

• Lascelles 2013 was obtained only as an abstract. It is the sole source for Dickens's hours-to-days claim. The abstract mentions no hours and no days, and argues on era timescales, so the claim looks like Dickens's own extrapolation. The body could still put a number on a single resedimented bed; the full text is closed access and was not read.
• Watanabe, Farquhar and Ohmoto 2009 and Ohmoto's 2006 GSA Memoir chapter were not obtained. Oard's answer to the sulfur signal rests on them. The three tests in Part 5 are stated, not run.
• Klein 2005 was read only at p. 1473. The body of the paper may qualify the abstract's numbers.
• Two creationist locators are incomplete. Wise 2002 is cited to PDF page 233 of the library copy because the file carries no printed folios. Oard 2015 is cited to the web version because its Journal of Creation page range could not be pinned.
• Konhauser 2017 is cited to its 131-page author manuscript, whose pagination differs from the journal's 140 to 177.
• Thicknesses for the Transvaal carbonate platform and the Belt Supergroup were not verified, so the stromatolite clock is run only on the Chuar, where a creationist supplied the numbers.
• Snelling's 2018 paper on Hamersley goethite ages and Clarey's 2021 and 2023 Great Unconformity pieces were not read. The claim that the creationist literature is silent on the Hamersley's chemistry is based on the sources listed here.
• The Barley 1999 ore paper behind Dickens and Snelling's "significant oxygen" claim was not read.
• Five mainstream papers were read through archive.org microfilm with leaf-to-page maps rather than printed folios on the page: Allwood, Farquhar, Chivas, Reid, Barley. Their locators were anchored twice each but deserve a second check.
• This article's own weakest point is that the pre-Flood model's iron budget, six times the published oxidation ceiling over 1,656 years, is the least severe of the multiples. A creationist who finds an oxidant flux ten times François's estimate, and an ocean at 0.5 mM iron across three ocean volumes, has closed the pre-Flood iron gap. The rate gap and the atmosphere would remain.

Verdict

Precambrian stromatolites and banded iron formations do not fit any young-earth placement.  The clearest case needs no mainstream chronology.  Austin's ICR monograph describes a 12-foot columnar stromatolite bed in the Chuar Group and dates the group to the first 1,656 years after Creation.  At the only measured accretion rates for a lithifying stromatolite, 0.18 to 0.34 mm a year, the bed alone needs 10,800 to 20,300 years.  The 2,015 m of Chuar around it needs 1.22 m of shallow-water sediment a year, every year, with exposure surfaces inside it.

The Hamersley Group gives its own rate, 13.9 m per million years from its thickness and zircon bracket, confirmed at 5 to 15 by zircons and band ratios inside one member.  A pre-Flood placement multiplies that by 110,000, a Flood-year placement by 180 million, a Creation-Week placement by 33 billion.  The one paper the creationist "rapid BIF" claim cites says 30 to 1,000 m per million years.  The iron itself needs an anoxic ocean holding several ocean volumes of dissolved iron, and the sulfur in the same rocks needs an atmosphere with less than 0.001 percent of today's oxygen and no ozone.  Every placement puts that air either in Creation Week, or over Adam, or over Noah.

The young-earth side is right that biogenicity is contested, that laminae are not years, that the Precambrian ocean was unlike ours, and that Archean oxygen is argued about.  None of it shortens the clock.  What would: a measured accretion rate above 1 mm a year for a lithifying stromatolite, a Liesegang model that yields Lantink's 1:9 and 1:4 ratios, an iron budget that closes, and a reading of the sulfur record that explains why it stops.  None has been published.  Until one is, these rocks are the pre-Flood era's hardest problem, and no published model has 1,650 years to spare for them.

Sources