Do Planets Form By Themselves?

Meter-size drift and bouncing are real, and no model gives a complete path from dust to planets. Plausible routes exist, so the gaps do not show natural formation is impossible.

Draft:  August 2026

"For s ∼ 100 cm, τc < 100 yr at r = 1 AU." — Stuart Weidenschilling on how fast a metre-sized boulder falls into the Sun, 19771

"Are we seeing any planetary systems in formation today?  I think it is yet too early to say." — John Hartnett, February 20162

"Unless disk dust masses are heavily underestimated, our results imply that planet formation has already occurred in younger systems…" — Carlo Manara, Alessandro Morbidelli and Tristan Guillot, 20183

Summary

This is the harder of the two questions, and it gets a more mixed answer than its companion article on star formation.

Creationist writers say that the standard account of planet formation is broken in the middle.  On the specific mechanical objection, they are largely right, and their numbers check out.  Where they go wrong is in the conclusion they draw from it.

The metre-size barrier is real and the number quoted is correct.  Gas in a disk orbits slightly slower than a solid body would, so solids feel a permanent headwind and spiral inward.  The drift is worst for boulders about a metre across.  Stuart Weidenschilling worked it out in 1977 and his own paper says a body of that size at Earth's distance falls into the Sun in under a hundred years.1 John Hartnett quotes this accurately.2

Growth stalls long before a metre, and this got worse after 2010, not better.  Laboratory collision experiments fed into a Monte Carlo growth model showed that aggregates stop growing because they bounce, not because they shatter, and that the maximum mass reached at Earth's distance is about one gram.4 Then in 2019 the standard escape route — that water ice is much stickier than rock — was undercut by measurements showing the surface energy of ice falls by nearly two orders of magnitude between 273 K and 180 K.5

The disks may not have enough material.  Comparing the measured dust masses of one-to-three-million-year-old disks with the masses of known exoplanet systems gives the wrong sign: "exoplanetary systems masses are comparable or higher than the most massive disks."3

But three things have changed since the creationist articles were written, and all three cut the other way.

  1. A planetary system has now been caught forming. PDS 70 has two directly imaged planets inside the gap of its disk.6 7 One of them has a resolved disk of its own — a circumplanetary disk, imaged by ALMA in 2021 at 2.3 astronomical units' resolution.8 Hartnett's "too early to say," written in February 2016, was overtaken within two years.
  2. The jump past the barrier has a candidate mechanism with independent evidence. The streaming instability concentrates pebbles aerodynamically until a clump collapses straight to a body about 100 km across, skipping the dangerous sizes entirely. Three separate lines of solar-system evidence support it: asteroids must have been "born big" at about 100 km to reproduce the observed size distribution;9 Kuiper belt binaries are overwhelmingly prograde and equal-sized, which is what collapsing clumps produce and what hierarchical capture does not;10 and Arrokoth is a contact binary that merged at a few metres per second, "a gentle, low-speed merger."11
  3. The slow-growth problem for the outer planets has a candidate fix. Pebble accretion shortens the core growth timescale "by a factor 30–1,000 at 5 AU and by a factor 100–10,000 at 50 AU."12

The angular momentum argument does not work.  Hartnett says the Sun should hold most of the system's angular momentum and holds about 1 %.  The correct figure, computed here, is that the Sun holds 99.87 % of the mass and about 0.6 % of the angular momentum, with Jupiter alone holding 61 %.  But that split is mostly geometry — orbital angular momentum grows with the square root of orbital distance — and the deeper problem is different and larger: a dense core has about ten million times more angular momentum per unit mass than the Sun does.13 Nearly all of it has to leave the system entirely, and that removal is now observed in protostellar jets and winds.  Also worth noting: the solar system's specific angular momentum, 1017 cm²/s, sits squarely inside the range measured for known exoplanet systems.13 Ours is not the odd one out.

Where each argument lands

Claim Is it accurate? Does the conclusion follow?
Grains stop growing at millimetres to centimetres Yes — bouncing barrier, ~1 g maximum No
Metre boulders fall into the star in ~100 years Yes — Weidenschilling's own figure No
Ice stickiness was the escape route and it is contested Yes, since 2019 No
Disks look too light to build the observed planets Yes — an active, unresolved problem No
Uranus and Neptune are hard to build Yes, though less so since pebble accretion No
No planetary system has been seen forming Was defensible in 2016; false since 2018 No
The angular momentum split is unexplained Overstated — and ours is typical of exoplanet systems No
Simulations assume their starting conditions Partly No
Planet formation needs dark matter No — off by a factor of ~700 inside a cloud No
"It is all just storytelling" No — the streaming instability makes checkable predictions that were checked No

What this article does

It states the objection at full strength, gives the primary source for every number, and then reports what has happened since — including the parts that make the problem worse.

Two honest framings run through it.  First, "unsolved" and "impossible" are different claims, and the creationist articles slide between them.  Second, a mechanism can be unknown while the product is observed forming.  Six thousand three hundred and thirty-six planets are now confirmed around 4,749 stars.14 They are found in disks, in gaps in disks, and in one case still glowing with the heat of their own accretion.

Every source, with what it measured and which way it cuts, is catalogued in a companion page: the literature.

Part 1 — What creationists actually claim

The most detailed young-earth treatment is John Hartnett's Planetary system formation: exposing naturalistic storytelling, published by Creation Ministries International in February 2016.2 It is a careful article.  Hartnett reads the standard sources, quotes them accurately, and identifies the genuine weak points rather than inventing them.

His case has five parts.

One: dust does not stick well enough.

"Growth via collisions has problems.  For example, as the grain size increases so does the mean collision velocity, which is likely to stall the growth process or actually reverse it, as grains bounce off one another and also break one another into smaller sizes.  So, some other mechanism needs to be found."2

Two: the metre-size barrier.

"Furthermore, simulations show that once the putative grain size has reached about 1 metre at a distance of 1 a.u. from the protostar the grains will spiral into the protostar in only 100 years.  This is known as the 'metre-sized barrier' to planetesimal formation."2

Three: the textbook papers over the gap.  He quotes a Swinburne University course note — "While the mechanism is not yet fully understood, the grains will eventually become kilometre-sized planetesimals" — and comments: "More hopeful storytelling."2

Four: angular momentum.

"Consider our sun with a mass representing 99.86 % of the mass of the solar system but only 1 % of its angular momentum. … Basic physics tells us that the collapsing star must 'spin up' and hence contain most of the angular momentum of the system.  Yet, after some time nearly all the angular momentum is allegedly transferred to the planets somehow.  This is a huge problem for the evolutionary theory."2

Five: nothing has been seen.

"Are we seeing any planetary systems in formation today?  I think it is yet too early to say."2

He ends: "I accept that God made extrasolar planetary systems on Day 4 of Creation Week about 6,000 years ago."2

Points one, two and three are substantially right.  Point four is confused.  Point five was reasonable when written and is now false.

Part 2 — The barriers, stated at full strength

Planet formation has to cross thirteen orders of magnitude in size, from a micron of dust to a ten-thousand-kilometre planet.  Different physics governs each stretch, and one stretch in the middle is genuinely not understood.

Stage one works: dust to pebbles

Micron grains in a disk drift slowly and collide gently, at millimetres per second.  Van der Waals forces are enough to make them stick.  Fluffy, fractal aggregates grow quickly — centimetre sizes within about a thousand years.  This part is measured in the laboratory and seen in disks, whose millimetre-wavelength brightness requires grains that have already grown far beyond interstellar sizes.

Stage two fails: pebbles to boulders

As aggregates get bigger they couple less tightly to the gas, so they move relative to it and relative to each other.  Collision speeds rise.  Above roughly one metre per second for rock, colliding aggregates stop sticking.

The failure mode turned out not to be the one originally expected.  Andras Zsom and colleagues built a collision model directly from laboratory experiments and ran it through a Monte Carlo growth code at Earth's distance from the star, at three different gas densities.  Their conclusion:

"The particle growth is stopped by bouncing rather than fragmentation in these models.  The final Stokes number of the aggregates is rather insensitive to the variations of the gas density and the strength of turbulence.  The maximum mass of the particles is limited to ≈ 1 g (chondrule sized particles)."4

One gram.  A pebble the size of a marble.  That is where growth by sticking stops in the inner disk, and the result is robust against the two parameters that were expected to matter.

This is the bouncing barrier, and it is worse than the fragmentation barrier it replaced, because bouncing produces no fragments to recycle.

Stage three is a race the solids lose: radial drift

Gas in a disk is partly supported by its own pressure gradient, so it orbits a little slower than Keplerian.  A solid body feels this as a permanent headwind, loses angular momentum, and spirals inward.

The effect is strongest when the aerodynamic coupling time matches the orbital time — around a metre in size at Earth's distance.  Weidenschilling computed the drift speeds in 1977 and found them "of the order of 104 cm/s for metre-sized objects,"1 which is 100 metres per second.  His characteristic loss time is stated directly:

"For s ∼ 100 cm, τc < 100 yr at r = 1 AU."1

Under a century, from a nebula that lasts a few million years.

Put the two together and the problem is sharp.  Growth by sticking stops at about a gram.  Anything that does reach a metre is removed in a hundred years.  There is a window of sizes that solids cannot grow through and cannot survive in.

Hartnett's summary of this is accurate.  The Berean Archive's own research notes quote a 2015 explainer saying that thirty-eight years after Weidenschilling's paper "there is still no commonly accepted solution to circumvent the barrier."  That was fair in 2015.

And the escape route got narrower in 2019

For years the standard hope was that ice changes the arithmetic.  Water ice has a higher surface energy than silicate, so icy grains should stick at higher speeds, and beyond the snow line — where ice exists — growth might get past the barrier.

Bastian Gundlach and Jürgen Blum measured it: micron water-ice particles stick below 9.6 metres per second, about ten times the silicate threshold, with erosion setting in above 15.3 metres per second.15 That looked like the answer.

Then Grzegorz Musiolik and Gerhard Wurm repeated the measurement at the temperatures that actually apply in a disk.  Ice is sticky at 273 K.  At 180 K it is not.  The surface energy falls by nearly two orders of magnitude across that range, and at disk temperatures water ice is not measurably stickier than silicate or organic material.5 Some later work puts the ice fragmentation velocity as low as 1 metre per second — no better than rock.

This is contested, and other groups disagree.16 But the honest statement is that the most-cited escape route from the growth barrier was reopened as a question in 2019, not closed.  A creationist writing today would be entitled to make more of this than Hartnett could in 2016.

Part 3 — The proposed way across, and the evidence for it

The leading answer is that solids never grow through the dangerous sizes at all.  They jump.

The streaming instability

Andrew Youdin and Jeremy Goodman showed in 2005 that the drag between solids and gas is not a one-way street.  Solids drag on the gas too.  In a disk with enough solids, that coupling is linearly unstable: small clumps of pebbles slow the gas around them, which lets more pebbles catch up, which strengthens the clump.17

In 2007 Anders Johansen and colleagues ran the instability in three dimensions with self-gravity.  Clumps reached densities high enough to collapse under their own weight, forming bound objects "with masses comparable to dwarf planets," and — critically — "gravitational collapse happens much faster than radial drift."18 Jacob Simon and colleagues later measured the mass and size distribution the process produces.19

The output is a body roughly 100 kilometres across, made in one step.  No metre-sized boulders are ever produced, so the barrier is never encountered.

That is the claim.  It could be wrong.  Here is why it is probably not.

Test 1 — asteroids were born big

If planetesimals grew gradually from kilometre-sized seeds, the asteroid belt should have a smooth size distribution.  It does not.  There is a distinct bump at about 100 kilometres in diameter.

Alessandro Morbidelli and colleagues treated the initial size distribution as a free parameter and asked which starting distribution ends up matching the belt.  Their answer:

"…reproducing the bump observed at diameter D ∼ 100 km in the current SFD of the asteroids requires that the minimal size of the initial planetesimals was also ∼100 km.  This supports the idea that planetesimals formed big, namely that the size of solids in the proto-planetary disk 'jumped' from sub-meter scale to multi-kilometer scale, without passing through intermediate values."9

That paper was published in 2009.  It was not written to defend the streaming instability; it was a constraint derived from the asteroid belt, and it happens to demand precisely the jump that the instability produces.

Test 2 — Kuiper belt binaries

Cold classical Kuiper belt objects have an extraordinary property: a large fraction are binaries whose two members are nearly the same size, nearly the same colour, and orbiting each other in the prograde sense — the same direction the pair orbits the Sun.

Gravitational collapse of a rotating pebble clump makes exactly this.  A clump inherits prograde rotation from the disk's shear, and it usually has too much angular momentum to collapse into a single body, so it splits into two similar pieces of identical composition.

Capture models make the opposite prediction.  David Nesvorný and colleagues analysed streaming-instability simulations and compared the resulting binary orbits to the observed ones:

"Here we report evidence that Kuiper belt planetesimals formed by the streaming instability, a process in which aerodynamically concentrated clumps of pebbles gravitationally collapse into ∼100-km-class bodies.  …which imply predominantly retrograde binary orbits can be ruled out."10

A prediction that could have failed, tested against data collected for other reasons, and it held.

Test 3 — Arrokoth

New Horizons flew past the Kuiper belt object Arrokoth on 1 January 2019 and photographed a contact binary: two flattened lobes touching.  If the two lobes had collided fast, the contact surfaces would show it.  They do not.

William McKinnon and colleagues simulated the merger.  Impacts at 10 metres per second shear or slice off sections of the bodies rather than merging them; even 5 metres per second produces distortion incompatible with what is seen.  Their conclusion is that Arrokoth "is the product of a gentle, low-speed merger in the early Solar System,"11 at or below the pair's mutual escape speed of a few metres per second.11

They connect it directly to the mechanism: such a merger is what happens when a cloud of pebbles produced by the streaming instability collapses and its two densest concentrations spiral together under gas drag and dynamical friction, on timescales of a thousand years or less.11

A photograph of an object 6.6 billion kilometres away is not proof of a theory.  But it is exactly what the theory says the oldest surviving planetesimals should look like, and it is not what a violent accretion history would produce.

What is still missing

The streaming instability needs the local solid-to-gas ratio raised above the solar value before it will run.  Where that enhancement comes from is not settled — proposals include drift pile-ups, pressure bumps, ice lines and photoevaporation.  So the mechanism is not self-starting from generic conditions.

This is the honest state: a specific, quantitative mechanism exists; it produces the right sizes; three independent solar-system observations match its distinctive predictions; and its trigger condition is not yet derived from first principles.  That is much stronger than "storytelling" and much weaker than a solved problem.

Part 4 — From planetesimals to planets

The old timescale problem

Once bodies are 100 kilometres across, gravity does the work.  Runaway growth is fast in the inner disk.  It is not fast far out, where orbits are slow and material is thin.

Building Uranus and Neptune has been the standard illustration.  The classical estimate for Neptune's growth time exceeded the age of the solar system.  Ravit Helled and Peter Bodenheimer's 2014 study concluded that "in order to form Uranus-like and Neptune-like planets in terms of final mass and solid-to-gas ratio, very specific conditions are required."20

Their result is more interesting than the simple version, and it cuts both ways.  With the higher solid accretion rates proposed more recently, "the challenge in forming Uranus and Neptune at large radial distances is no longer the formation timescale, but is rather finding agreement with the final mass and composition of these planets.  In fact, these conditions are more likely to lead to gas-giant planets."20

So the problem changed shape.  It used to be that the ice giants could not be built fast enough.  Now it is that they should have grown into gas giants and did not.  Fine-tuning of the disk conditions is still required.  That is a real weakness and it should be stated.

Pebble accretion

The mechanism that changed the timescale is pebble accretion.  A growing core does not only sweep up what it physically hits.  Centimetre pebbles are aerodynamically coupled to the gas, so when they enter the core's Hill sphere, gas drag bleeds their energy and they spiral in.  The effective capture cross-section becomes enormously larger than the core's geometric size.

Michiel Lambrechts and Anders Johansen quantified it:

"We conclude that pebble accretion can resolve the long-standing core accretion time-scale conflict. … The core growth time-scale is shortened by a factor 30–1,000 at 5 AU and by a factor 100–10,000 at 50 AU, compared to the gravitationally focused accretion of, respectively, low-scale-height planetesimal fragments or standard km-sized planetesimals."12

They also state the conditions it needs: a near-unity dust-to-gas ratio in the midplane, growth of dust to millimetre and centimetre sizes, and either massive planetesimals or low radial pressure support.12 Those are assumptions, and they are the same assumptions the streaming instability needs.

Migration, and the test that could have failed

Planets embedded in a gas disk exchange angular momentum with it and move.  For decades this was treated as an embarrassment: the calculated migration rates were fast enough to drive growing planets into the star before they finished forming.

The reason it is no longer only an embarrassment is that migration makes a distinctive, checkable prediction.  If planets migrate inward through a disk at different rates, they will catch each other and become trapped in mean-motion resonances — orbital periods in exact small-integer ratios.  Nothing else naturally produces a long chain of such ratios.

Two systems settle it.

Kepler-223 has four transiting planets in a chain of resonances: successive period ratios near 4:3, 3:2 and 4:3, with all three resonant angles librating.21 The authors state that this configuration is a natural outcome of disk migration and is very difficult to reach otherwise.

TRAPPIST-1 is more extreme: seven planets, every adjacent pair near a two- or three-body resonance, forming a chain from the innermost to the outermost.22

A resonant chain is fragile.  It has to be built by something that moves planets slowly and smoothly and then stops.  Gas-disk migration does that.  A system created in place would need its period ratios to have been set to those values from the start.

Part 5 — The angular momentum question

Hartnett's fourth objection deserves careful handling, because the popular version and the real version are different problems.

The Sun holds almost all the mass and almost none of the spin.  Computing this directly from the planets' masses and orbits, and from the Sun's helioseismically determined moment of inertia:

Body Angular momentum (kg m²/s) Share
Jupiter 1.93 × 1043 61.1 %
Saturn 7.82 × 1042 24.8 %
Neptune 2.51 × 1042 7.9 %
Uranus 1.69 × 1042 5.4 %
Earth 2.66 × 1040 0.08 %
Venus 1.84 × 1040 0.06 %
Mars 3.51 × 1039 0.01 %
Mercury 8.96 × 1038 0.003 %
Sun (spin) 1.9 × 1041 0.6 %

The Sun holds 99.87 % of the mass and about 0.6 % of the angular momentum.  Hartnett's "1 %" is close enough.

But this split is not by itself mysterious.  Orbital angular momentum is M √(GM☉ a) — it grows with the square root of orbital distance.  Jupiter is 1,100 times farther from the Sun's centre than the Sun's own surface is.  Even a tiny mass placed that far out carries an enormous amount of angular momentum.  Move Jupiter's mass onto the Sun's surface and its contribution would drop by a factor of about 33.

The real version, which is bigger

The serious problem is not how the angular momentum is divided.  It is how much of it disappeared.

Arnaud Belloche's review tabulates the measured specific angular momentum — angular momentum per unit mass — at each stage:

Object J/M (cm² s−1)
Dense cores in molecular clouds 1021–1022
Protoplanetary disks 1019–1021
Pre-main-sequence stars 1016–1017
Extrasolar planetary systems (star + planets) 1016–1018
Solar system (Sun + planets) 1017
Sun alone 1015

13

"It has long been noticed that parcels of gas in molecular clouds need to reduce their specific angular momentum by 6 to 7 orders of magnitude to participate in the building of a typical star like the Sun."13

Six to seven orders of magnitude.  Counting the planets does not rescue it: the whole solar system's specific angular momentum is 1017, still ten thousand times below the core it came from.  The angular momentum was not redistributed within the system.  Most of it left.

And that removal is observed

Belloche identifies the stages where it happens: magnetic braking during collapse, jets and outflows during the protostellar phase, star-disk interaction and magnetised winds afterward.13

Each of those is now measured rather than assumed.  Jet rotation has been detected in DG Tau and in HH 212.  Core-to-disk angular momentum profiles have been mapped in three protostars, and in July 2026 Chin-Fei Lee reported that all three "are broadly consistent with an inside-out collapse scenario with approximate conservation of specific angular momentum in the collapsing region," with the accreted masses matching the central masses derived from kinematics once ejection by jets and winds is allowed for.23

There is also a control that is easy to miss: the solar system is not unusual.  Its specific angular momentum, 1017 cm²/s, falls inside the 1016–1018 range measured for known extrasolar planetary systems.13 If the Sun's slow rotation were an anomaly demanding a special explanation, other stars with planets should not share it.  They do.

The Sun's 7-degree tilt

A related argument appears in creationist writing: the Sun's equator is tilted about 7 degrees to the plane of the planets, which a simple collapsing-disk picture does not predict.

The observation is right, and the explanation is unsettled.  Elizabeth Bailey, Konstantin Batygin and Michael Brown state the problem cleanly: the obliquity "suggests that either an asymmetry was present in the solar system's formation environment, or an external torque has misaligned the angular momentum vectors of the sun and the planets.  However, the exact origin of this obliquity remains an open question."24

Their own proposal — that a distant undiscovered planet torqued the plane over billions of years — depends on Planet Nine existing, which is not established.  Other proposals invoke a warp in the original disk or a passing star.  None is confirmed.

This is a genuine open question about our own system.  It is not evidence that planets cannot form, because whatever tilted the Sun by 7 degrees left the planets in a plane flat to a fraction of a degree.

Part 6 — The problem nobody has solved: the disks are too light

Here is a live difficulty that creationist writers have not picked up, and it is sharper than most of what they do use.

Take every protoplanetary disk whose dust mass has been measured in a one-to-three-million-year-old star-forming region.  Take every confirmed exoplanet system and work out how much heavy material it contains — full masses for super-Earths, core masses for giants.  Compare.

Carlo Manara, Alessandro Morbidelli and Tristan Guillot did this with distances corrected using Gaia parallaxes and got the wrong sign:

"While expecting that the mass in exoplanetary systems is much lower than the measured disk masses, we instead find that exoplanetary systems masses are comparable or higher than the most massive disks."3

There are three ways out and none is comfortable.  The dust masses might be systematically underestimated, because millimetre observations are blind to material already locked into large bodies.  Disks might be continuously resupplied from their surroundings.  Or planet formation might start much earlier — before the Class II stage where the disks were measured.

The third option now has support.  Łukasz Tychoniec and colleagues measured dust masses in the youngest disks and found median values of 158 Earth masses for Class 0 and 52 for Class I, "larger by at least a factor of 10 and 3, respectively, compared with dust masses inferred for Class II disks."25 Their conclusion is that the sums close only if planet formation begins in the Class 0 phase, at about 15 % efficiency.25

That in turn is supported by the rings.  HL Tauri, at most one to two million years old, has a ringed disk.  IRS 63, a Class I protostar less than 500,000 years old, has four annular structures at 5-astronomical-unit resolution.26

The honest summary: the material budget does not obviously balance, the fix is that planet formation starts far earlier than anyone assumed a decade ago, and the observations have moved in that direction.  It is an open problem being actively worked, and it could still go badly.

Part 7 — A system caught in the act

The single strongest response to "no planetary system has been seen forming" has a name and a date.

PDS 70 is a young star 112.4 parsecs away — about 367 light-years — with a wide gap in its disk, roughly 54 astronomical units across.  In 2018 Miriam Keppler and colleagues used the SPHERE instrument on the Very Large Telescope to image a point source sitting inside that gap, at a projected separation of about 22 astronomical units.  It was confirmed at five epochs, in three filter bands, with three different instruments, and its motion showed it to be gravitationally bound.  Its colours and brightness put it in the planetary mass range.6 In 2019 Sebastiaan Haffert and colleagues confirmed it in the hydrogen-alpha line — emitted by gas crashing onto a growing body — and found a second Hα source in the same gap, PDS 70 c.7 Both planets are still accreting.

Then in 2021 Myriam Benisty and colleagues resolved, with ALMA at about 2.3 astronomical units' resolution, a compact dust disk around PDS 70 c itself.8 A disk around a planet inside a gap inside a disk around a star.

This chain of objects is what the theory says a forming planetary system should look like.  Each element was predicted before it was seen: the planet clears a gap; the planet accretes; the accreting planet keeps a disk of its own from which moons could form.

Hartnett's February 2016 judgment — "I think it is yet too early to say" — was a fair reading of the evidence available to him.  It is not a fair reading of the evidence available now.

Part 8 — Checking the quotations and their dates

The creationist articles on this subject quote real sources accurately.  The problem is not misquotation.  It is age.

Hartnett's primary technical source is a Swinburne University course module and the second edition of Carroll and Ostlie's textbook.  The Berean Archive's own research notes on this topic cite Blum and Wurm's 2008 review, a 2007 Nature news piece, a 2009 NASA press release, a 2006 workshop announcement, a 2000 Astronomy magazine article and a 1989 book.

Every one of those predates:

  • the streaming-instability planetesimal simulations with self-gravity (Simon and colleagues, 2016)
  • the asteroid size-distribution constraint being tied to it
  • the Kuiper belt binary orientation test (2019)
  • the Arrokoth flyby (2019)
  • pebble accretion (2012)
  • ALMA's resolved disks (2014 onward)
  • PDS 70 (2018–2021)
  • the disk-mass problem (2018)
  • the reopening of the ice-stickiness question (2019)

Some of those changes help the standard account and some hurt it.  A current audit has to include both.  Quoting a 2009 statement that a process is "unknown" as though it were a statement about 2026 is the same error, whichever direction it points.

One more specific correction.  Hartnett carries the dark-matter argument over from his star formation articles: "Alternatively they invoke hypothetical dark matter, which has conveniently anomalous properties."2 Dark matter plays no role in present-day star or planet formation.  Its local density is about 6.8 × 10−25 grams per cubic centimetre; a molecular cloud is about 700 times denser and a protoplanetary disk is denser still by many orders of magnitude.  The measured upper limit on dark matter inside Saturn's orbit is 7.1 × 10−11 solar masses.  This is treated in full in the companion article.

Common objections

"You have shown mechanisms, not proof.  The gap is still there."  Correct on the first clause.  The pebbles-to-planetesimals step has a leading candidate with three independent lines of solar-system support, not a demonstration.  The honest position is that it is the best-supported unfinished explanation in planetary science, and that if it fails there is currently nothing to replace it.

"Everything you cite is a simulation."  Not everything.  Weidenschilling's drift, Zsom's collision model and the streaming instability are calculations.  The asteroid size bump, the Kuiper belt binary orientations, Arrokoth's shape, the disk rings, the resonant chains, the disk mass measurements and the PDS 70 images are observations.  The strongest arguments in this article are the places where a calculation made a distinctive prediction and an observation collected for another purpose matched it.

"Planet formation must produce habitable planets, and that is much harder."  A separate question, treated in the habitability article.  Nothing here bears on it.

"God could have made systems that look like they are at different stages."  He could.  Hartnett says exactly this: the systems "might be incorrectly labelled by evolutionists as being at various 'evolutionary stages' of formation."2 This is not refutable and it is not testable, which is the difficulty with it.  It also gives up any evidential use of astronomical appearances, in either direction.  If a star with a gap in its disk, with two accreting planets in the gap, one carrying its own disk, is a created snapshot rather than a process, then no astronomical appearance can be evidence of anything.

"The metre-size barrier alone should sink the theory."  It sinks one route — steady growth by sticking.  It is not an argument against a route that skips those sizes.  And the observations that support the skip — a 100 km bump in the asteroids, prograde equal-size Kuiper belt binaries, a gently merged contact binary — are exactly the fingerprints a skip would leave.

Verdict

Mixed, and the split is clean.

On mechanism, the young-earth critique is substantially correct and better sourced than most creationist arguments this site has audited.  Growth by sticking really does stall at about one gram.  Metre boulders really do fall into the star in under a century.  The escape route through sticky ice was reopened as a question in 2019.  The disks really do look too light to build the planets we see.  Uranus and Neptune really do require special conditions.  Hartnett quotes his sources accurately and identifies the right weak points.  Anyone claiming that planet formation is a solved problem is overstating.

On the conclusion, the critique fails, for four reasons.

First, the gap has a candidate mechanism that produces the right answer in one step, and three independent solar-system observations carry its distinctive fingerprints.  "Unsolved" is not "impossible."

Second, the angular momentum argument is confused.  The mass-versus-spin split is largely geometry.  The real problem is six to seven orders of magnitude of angular momentum that had to leave the system, and the routes by which it leaves are now measured in individual protostars.  The solar system's specific angular momentum is ordinary compared with other planetary systems.

Third, the dark matter claim carried over from the star formation articles is wrong by a large factor and should be dropped.

Fourth, and decisively, a planetary system has now been imaged in the act of forming.  PDS 70 has two accreting planets inside a cleared gap and one of them has its own disk.  Whatever the details of how solids cross the metre barrier, planets are visibly being built around other stars right now.

The honest position is the uncomfortable one: planet formation is observed to happen and is not fully understood.  That is not a paradox.  It is the normal condition of a science that has good telescopes and incomplete theory.  What it will not support is the inference that because a step is unexplained, the whole process never occurred and the systems were made six thousand years ago.

Sources

  1. 1.

    Weidenschilling, S. J.  “Aerodynamics of solid bodies in the solar nebula.”  Monthly Notices of the Royal Astronomical Society 180:57–70.  1977.  A theoretical paper, with no new measurements, that works out how gas drag makes solid bodies in the solar nebula (the disk of gas and dust that formed the Sun and planets) drift inward.  It gives analytic and numerical results for all the applicable drag laws and for several model nebulae, and finds that the drift speed depends strongly on size, reaching about 104 cm/s (100 metres per second) for metre-sized objects.  The article relies on it for the claim that the metre-size barrier is real.

    1. 1a.

      “Radial velocities depend strongly on particle size, reaching values of the order of 104 cm/s for metre-sized objects.” ↩

    2. 1b.

      “A particle would spiral into the sun in a characteristic lifetime of τc = r/|dr/dt|. … For s ∼ 100 cm, τc < 100 yr at r = 1 AU.” ↩ ↩ ↩

  2. 2.

    Hartnett, J. G.  “Planetary system formation: exposing naturalistic storytelling.”  Creation Ministries International.  Feb 17, 2016.  A young-earth creationist article arguing that naturalistic models do not yet explain how planets form.  Hartnett's article was updated on 14 April 2016.  The article calls it the most detailed young-earth treatment of the question, and says Hartnett reads the standard sources, quotes them accurately, and identifies genuine weak points. ↩

    1. 2a.

      “Are we seeing any planetary systems in formation today?  I think it is yet too early to say.  But does the naturalistic model have any substantive basis to claim an understanding of planet formation using only known physics?  Definitely not, especially when dark matter and unlikely supernovae events have to be invoked.” ↩ ↩

    2. 2b.

      “Growth via collisions has problems.  For example, as the grain size increases so does the mean collision velocity, which is likely to stall the growth process or actually reverse it, as grains bounce off one another and also break one another into smaller sizes.  So, some other mechanism needs to be found.” ↩

    3. 2c.

      “Furthermore, simulations show that once the putative grain size has reached about 1 metre at a distance of 1 a.u. from the protostar the grains will spiral into the protostar in only 100 years.  This is known as the 'metre-sized barrier' to planetesimal formation.” ↩ ↩

    4. 2d.

      Quoting a Swinburne University of Technology course module:  “While the mechanism is not yet fully understood, the grains will eventually become kilometre-sized planetesimals.”  Hartnett's comment:  “More hopeful storytelling.” ↩

    5. 2e.

      “Consider our sun with a mass representing 99.86 % of the mass of the solar system but only 1 % of its angular momentum. … Most of the angular momentum of any solar system must reside initially in the central core of the collapsing cloud, which eventually becomes the star, according to the theory.  Basic physics tells us that the collapsing star must 'spin up' and hence contain most of the angular momentum of the system.  Yet, after some time nearly all the angular momentum is allegedly transferred to the planets somehow.  This is a huge problem for the evolutionary theory.” ↩

    6. 2f.

      “I accept that God made extrasolar planetary systems on Day 4 of Creation Week about 6,000 years ago.” ↩

    7. 2g.

      “Alternatively they invoke hypothetical dark matter, which has conveniently anomalous properties, such that it is not subject to normal thermodynamics and hence provides gravitational force, tending to collapse the cloud, yet no thermal energy to support the cloud against gravitational collapse.” ↩

    8. 2h.

      “According to the biblical model there is no reason to expect that the Creator did not create many different forms of planetary systems, which might be incorrectly labelled by evolutionists as being at various 'evolutionary stages' of formation.” ↩

  3. 3.

    Manara, C. F. et al.  “Why do protoplanetary disks appear not massive enough to form the known exoplanet population?”  Astronomy & Astrophysics 618:L3.  2018.  A short letter that compares the dust masses of disks around young stars, measured with the ALMA radio telescope array in star-forming regions 1 to 3 million years old, with the masses of known exoplanet systems.  The authors expected the planets to weigh less than their disks, but found they weigh as much as or more than the most massive disks.  The article uses it for the objection that the disks may not hold enough material to make the known planets.

    1. 3a.

      “Unless disk dust masses are heavily underestimated, our results imply that planet formation has already occurred in younger systems, or that disks are continuously replenished with fresh material from the environment.” ↩

    2. 3b.

      “While expecting that the mass in exoplanetary systems is much lower than the measured disk masses, we instead find that exoplanetary systems masses are comparable or higher than the most massive disks.”  Disk masses were recalculated using Gaia DR2 parallaxes; the comparison is repeated in heavy-element content, using core masses for giants and full masses for super-Earth systems. ↩ ↩

  4. 4.

    Zsom, A. et al.  “The outcome of protoplanetary dust growth: pebbles, boulders, or planetesimals? II. Introducing the bouncing barrier.”  Astronomy & Astrophysics 513:A57.  2010.  A computer study that models how dust grains in a disk stick and grow, using collision rules taken from laboratory experiments on dust aggregates (loose clumps of grains) in a random-sampling growth model.  It runs the model at 1 au (the Earth's distance from the Sun) at three gas densities, and finds that growth stops because the aggregates bounce apart rather than shatter, with a maximum mass of about one gram.  The authors say the collision model rests on necessary assumptions about the shape of the aggregates.  The article uses it for its claim that growth stalls long before a metre.

    1. 4a.

      “The particle growth is stopped by bouncing rather than fragmentation in these models.  The final Stokes number of the aggregates is rather insensitive to the variations of the gas density and the strength of turbulence.  The maximum mass of the particles is limited to ≈ 1 g (chondrule sized particles).”  The collision model is built from the laboratory experiments of Güttler et al. (2009) and applied at 1 au in the disk midplane at three gas densities. ↩ ↩

  5. 5.

    Musiolik, G., and G. Wurm.  “Contacts of Water Ice in Protoplanetary Disks — Laboratory Experiments.”  The Astrophysical Journal 873:58.  2019.  A laboratory experiment on how water-ice grains stick to one another at the temperatures found in protoplanetary disks.  The article uses it because its measurements undercut the idea that water ice is much stickier than rock, the usual escape route for dust growth.

    1. 5a.

      The measured surface energy of water ice falls by almost two orders of magnitude between 273 K and 180 K, so at the temperatures prevailing in protoplanetary disks water ice is not more sticky than silicate dust or complex organics.  Subsequent work has quoted fragmentation velocities for ice as low as about 1 m s−1. ↩ ↩

  6. 6.

    Keppler, M. et al.  “Discovery of a planetary-mass companion within the gap of the transition disk around PDS 70.”  Astronomy & Astrophysics 617:A44.  2018.  A 2018 observation paper that analysed new and older near-infrared images of the PDS 70 disk from three telescope instruments.  The authors found a point of light inside the disk's gap, bound to the star, with brightness and colours suggesting a planet-mass companion.  They also note that most earlier planet detections in disks were being challenged at the time.  The article uses it for one of the two directly imaged planets inside the PDS 70 gap.

    1. 6a.

      “We detect a point source within the gap of the disk at about 195 mas (∼22 au) projected separation.  The detection is confirmed at five different epochs, in three filter bands and using different instruments.  The astrometry results in an object of bound nature, with high significance.  The comparison of the measured magnitudes and colours to evolutionary tracks suggests that the detection is a companion of planetary mass.”  The gap itself is about 54 au across.  The distance adopted in the follow-up ALMA work is 112.4 pc. ↩ ↩

  7. 7.

    Haffert, S. Y. et al.  “Two accreting protoplanets around the young star PDS 70.”  Nature Astronomy 3:749–754.  2019.  A 2019 observation paper reporting two accreting protoplanets around the young star PDS 70.  Its evidence is strong hydrogen-alpha emission from two distinct locations inside the planets' gap, measured with high-resolution spectral imaging.  The article uses it for the second planet inside the PDS 70 gap.

    1. 7a.

      High-resolution (R ∼ 2,500) integral-field spectral imaging detects strong hydrogen-alpha emission from two distinct locations inside the PDS 70 gap, confirming the earlier detection of PDS 70 b and identifying a second protoplanet, PDS 70 c.  Line widths as well as intensities are measured, constraining the accretion onto each object. ↩ ↩

  8. 8.

    Benisty, M. et al.  “A Circumplanetary Disk Around PDS 70c.”  The Astrophysical Journal Letters 916:L2.  2021.  A short 2021 report, published as a Letter, of new ALMA radio-telescope observations of the PDS 70 system.  The observations were made at high angular resolution to resolve the disks around the planets, which are disks of gas and dust surrounding a planet, and to measure their dust masses.  The article uses it for the resolved disk around one of the two planets.

    1. 8a.

      ALMA band 7 continuum observations at about 20 milliarcsecond resolution (2.3 au) resolve compact dust emission at the location of PDS 70 c, interpreted as a circumplanetary disk; extended low-surface-brightness continuum emission is also detected within the gap. ↩ ↩

  9. 9.

    Morbidelli, Alessandro et al.  “Asteroids were born big.”  Icarus 204:558–573.  2009.  A computer study that simulates small solid bodies colliding, sticking and growing, to ask how big the first planetesimals (the first solid bodies to form in the disk) had to be.  Starting from km-sized bodies cannot reproduce the bump near 100 km in the size distribution of today's asteroids, but starting from about 100 km can.  The article uses it as the first of three lines of solar-system evidence that planetesimals jumped straight to large sizes.

    1. 9a.

      “We find that, if the initial planetesimals were small (e.g. km-sized), the final SFD fails to fulfill these constraints.  In particular, reproducing the bump observed at diameter D ∼ 100 km in the current SFD of the asteroids requires that the minimal size of the initial planetesimals was also ∼100 km.  This supports the idea that planetesimals formed big, namely that the size of solids in the proto-planetary disk 'jumped' from sub-meter scale to multi-kilometer scale, without passing through intermediate values.” ↩ ↩

  10. 10.

    Nesvorný, David et al.  “Trans-Neptunian Binaries as Evidence for Planetesimal Formation by the Streaming Instability.”  Nature Astronomy 3:808–812.  2019.  A short paper that runs three-dimensional computer simulations of the streaming instability and predicts how the orbits of Kuiper belt binary pairs (two small bodies orbiting each other) should be oriented.  About 80 per cent of the simulated binary orbits are prograde, that is, they move the same way as the planets, which matches the observed binaries, and the paper says models with mostly backward orbits can be ruled out.  The article uses it as evidence that Kuiper belt binaries are mostly prograde and similar in size, which collapsing clumps would produce.

    1. 10a.

      “Here we report evidence that Kuiper belt planetesimals formed by the streaming instability, a process in which aerodynamically concentrated clumps of pebbles gravitationally collapse into ∼100-km-class bodies.  Gravitational collapse was previously suggested to explain the ubiquity of equal-size binaries in the Kuiper belt. … [models] which imply predominantly retrograde binary orbits … can be ruled out.” ↩ ↩

  11. 11.

    McKinnon, W. B. et al.  “The solar nebula origin of (486958) Arrokoth, a primordial contact binary in the Kuiper belt.”  Science 367:eaay6620.  2020.  A paper by the New Horizons science team that studies how Arrokoth, a small body in the Kuiper belt made of two lobes that touch, was assembled.  It combines New Horizons' close-up data on the shape of the lobes with computer simulations of low-speed collisions, and concludes the lobes merged gently in the early Solar System.  The article uses it as evidence that Arrokoth's two lobes merged gently, which fits a pebble cloud collapsing under its own gravity.

    1. 11a.

      “…finding it is the product of a gentle, low-speed merger in the early Solar System.”  The two lobes are.  “little disturbed, with no obvious morphological signs of a violent or energetic merger.” ↩ ↩

    2. 11b.

      Numerical impact simulations.  “Oblique impacts at 10 m/s, much greater than escape speed, do not lead to mergers … but instead shear or slice off sections of one or both bodies.”  Even 5 m s−1, slightly above the 4.3 m s−1 escape speed of the simulated spheres, produces.  “distortion and merging incompatible”.  with the observed shape. ↩

    3. 11c.

      The paper attributes the parent pebble cloud to.  “a collective aerodynamic phenomenon called the streaming instability,”.  with gravitational collapse following on time scales of 103 years or less in the outer solar system. ↩

  12. 12.

    Lambrechts, M., and A. Johansen.  “Rapid growth of gas-giant cores by pebble accretion.”  Astronomy & Astrophysics 544:A32.  2012.  A numerical study of how a planet's solid core grows by catching centimetre-sized pebbles with the help of gas drag, rather than by catching kilometre-sized planetesimals.  The authors measured the catch rate for cores from a large planetesimal up to a fully grown core of ten Earth masses, and compared it with analytic formulas.  The article uses it as the candidate fix for the slow growth of the outer planets' cores.

    1. 12a.

      “We conclude that pebble accretion can resolve the long-standing core accretion time-scale conflict. … The core growth time-scale is shortened by a factor 30–1,000 at 5 AU and by a factor 100–10,000 at 50 AU, compared to the gravitationally focused accretion of, respectively, low-scale-height planetesimal fragments or standard km-sized planetesimals.” ↩ ↩

    2. 12b.

      “This requires a near-unity dust-to-gas ratio in the midplane, particle growth to mm and cm and the formation of massive planetesimals or low radial pressure support.” ↩

  13. 13.

    Belloche, A.  “Observation of rotation in star forming regions: clouds, cores, disks, and jets.”  EAS Publications Series 62:25–66.  2013.  A review of observations of how fast gas clouds, dense cores, disks and jets rotate in star-forming regions, covering about four decades of work up to February 2013.  The author says the review covers low-mass star-forming regions only.  The article uses its table of typical values for angular momentum, meaning how much spin or orbital motion a body carries, and its account of the stages at which that angular momentum is lost.

    1. 13a.

      “It has long been noticed that parcels of gas in molecular clouds need to reduce their specific angular momentum by 6 to 7 orders of magnitude to participate in the building of a typical star like the Sun.” ↩ ↩

    2. 13b.

      Table 1:  Typical specific angular momenta J/M in cm² s−1: dense cores in molecular clouds 1021–22; protoplanetary disks 1019–21; pre-main-sequence binaries 1019–20; pre-main-sequence stars 1016–17; extrasolar planetary systems (exoplanets + star) 1016–18; solar system (planets + Sun) 1017; Sun 1015. ↩ ↩ ↩

    3. 13c.

      “Magnetic braking may also play a significant role in carrying away angular momentum via Alfvén waves during the early phases of star formation. … During this protostellar phase, the ejection of matter in jets and outflows can partly contribute to this loss of angular momentum. … Star-disk interactions and magnetized stellar winds are thought to play a significant role in this respect.” ↩

  14. 14.

    NASA Exoplanet Archive.  pscomppars table.  n.d.  Retrieved Aug 9, 2026.  A NASA table of confirmed planets and the stars they orbit.  The article uses its counts of confirmed planets and host stars for the claim that the planets are now being found forming in disks.

    1. 14a.

      6,336 confirmed planets around 4,749 distinct host stars; 567 of those planets are in systems with four or more known planets. ↩

  15. 15.

    Gundlach, B., and J. Blum.  “The stickiness of micrometer-sized water-ice particles.”  The Astrophysical Journal 798:34.  2015.  A laboratory experiment that grew aggregates of micrometre-sized water-ice grains and measured the speeds at which they stuck together and the speeds at which the grown aggregates eroded, at temperatures from 114 K to 260 K.  The sticking result applies only below about 210 K, the authors say.  The article uses it as a measurement that looked like an answer to the growth problem.

    1. 15a.

      “We show with our experiments that for low temperatures (below ∼ 210 K), µm-sized water-ice particles stick below a threshold velocity of 9.6 m s−1, which is approximately ten times higher than the sticking threshold of µm-sized silica particles.  Furthermore, erosion of the grown ice aggregates is observed for velocities above 15.3 m s−1.” ↩

  16. 16.

    Birnstiel, T.  “Dust growth and evolution in protoplanetary disks.”  Annual Review of Astronomy and Astrophysics 62:1–48.  2024.  A review article on how the micrometre-sized dust in planet-forming disks grows, drifts and collides.  The author names three critical uncertainties: the level of turbulence, the collision speeds at which growth stalls, and how porous the dust grains are.  The article cites it for the unresolved question of how dust gets past the growth barrier, the point where collisions stop grains from sticking.

    1. 16a.

      The review summarises the state of the fragmentation-velocity question, including the disagreement over whether water ice raises the sticking threshold at protoplanetary-disk temperatures. ↩

  17. 17.

    Youdin, Andrew N., and Jeremy Goodman.  “Streaming Instabilities in Protoplanetary Disks.”  The Astrophysical Journal 620:459–469.  2005.  A theoretical paper, with no experiments, that studies a simplified two-fluid model of a disk in which gas and solid particles push on each other through drag.  It finds that this coupling produces a linear instability that grows clumps of particles, which could seed planetesimals without the help of self-gravity.  The model leaves out vertical structure, turbulence and self-gravity, which the authors say could change the outcome.  The article credits it with showing that the drag between solids and gas runs both ways, so that pebbles can grow clumps.

    1. 17a.

      “Interpenetrating streams of solids and gas in a Keplerian disk produce a local, linear instability.  The two components mutually interact via aerodynamic drag, which generates radial drift and triggers unstable modes.  The secular instability does not require self-gravity, yet it generates growing particle density perturbations that could seed planetesimal formation.  Growth rates are slower than dynamical, but faster than radial drift, timescales.” ↩

  18. 18.

    Johansen, A. et al.  “Rapid planetesimal formation in turbulent circumstellar disks.”  Nature 448:1022–1025.  2007.  A computer simulation of boulders moving in turbulent gas, with the gravity of the solid particles themselves included.  The authors model a small local box of the disk, not the whole disk, and use unstratified models of the turbulence.  The article uses it as the 2007 result showing that clumps of boulders can collapse under their own gravity faster than they drift inward.

    1. 18a.

      “Here we report that boulders can undergo efficient gravitational collapse in locally overdense regions in the midplane of the disk. … We find that gravitationally bound clusters form with masses comparable to dwarf planets and containing a distribution of boulder sizes.  Gravitational collapse happens much faster than radial drift, offering a possible path to planetesimal formation in accreting circumstellar disks.”  The most massive cluster in the simulation contains about 3.5 times the mass of Ceres. ↩

  19. 19.

    Simon, Jacob B. et al.  “The Mass and Size Distribution of Planetesimals Formed by the Streaming Instability. I. The Role of Self-Gravity.”  The Astrophysical Journal 822:55.  2016.  A computer study that adds self-gravity (the mutual pull between the solid particles) to the Athena simulation code, then runs three-dimensional simulations of how clumps of pebbles collapse into planetesimals.  The authors say their results agree well with earlier simulations, and they note that the small number of runs limits how strongly the claims can be made.  The article uses it as the source for the size distribution that this collapse produces.

    1. 19a.

      “The gravitational collapse of streaming-initiated clumps leads to an initial planetesimal mass function that is well-represented by a power-law, dN/dMp ∝ Mp−p, with p ≃ 1.6 ± 0.1, which equates to a differential size distribution dN/dRp ∝ Rp−q, with q ≃ 2.8 ± 0.1.” ↩

  20. 20.

    Helled, R., and P. Bodenheimer.  “The Formation of Uranus and Neptune: Challenges and Implications for Intermediate-Mass Exoplanets.”  The Astrophysical Journal 789:69.  2014.  A theoretical study of how Uranus and Neptune could have formed under the core-nucleated accretion model, in which a planet builds a solid core before it gathers its gas.  The authors test formation locations from 12 to 30 AU from the Sun, with different amounts of solid material in the disk and different core growth rates.  The article uses it for the standard puzzle of building Uranus and Neptune, and for the point that the answer depends on very specific conditions.

    1. 20a.

      “It is shown that in order to form Uranus-like and Neptune-like planets in terms of final mass and solid-to-gas ratio, very specific conditions are required.” ↩

    2. 20b.

      “We also show that when recently proposed high solid accretion rates are assumed, along with solid surface densities about 10 times those in the minimum-mass solar nebula, the challenge in forming Uranus and Neptune at large radial distances is no longer the formation timescale, but is rather finding agreement with the final mass and composition of these planets.  In fact, these conditions are more likely to lead to gas-giant planets.” ↩

  21. 21.

    Mills, Sean M. et al.  “A resonant chain of four transiting, sub-Neptune planets.”  Nature 533:509–512.  2016.  A Nature paper on the four planets of Kepler-223, a Sun-like star about 6 billion years old, which measures how each planet's transit (its pass in front of the star) is shifted slightly by the pull of the others, using four years of Kepler data.  The authors' simulations show that the chain is stable and that its pattern is a natural result of migration, with planets drifting inward through a gas disk, rather than of scattering.  The article uses it as a case where the authors say migration is the natural explanation.

    1. 21a.

      Kepler-223's four transiting planets lie in a chain of first-order mean-motion resonances with successive period ratios near 4:3, 3:2 and 4:3, and the associated resonant angles librate — a configuration the authors argue requires smooth, dissipative migration in a gas disk. ↩

  22. 22.

    Luger, Rodrigo et al.  “A seven-planet resonant chain in TRAPPIST-1.”  Nature Astronomy 1:0129.  2017.  A short paper that uses observations from NASA's K2 mission to measure the orbit of TRAPPIST-1h, the outermost of the star's seven planets, which had not been pinned down before.  The measured period of 18.77 days matches the value predicted from Laplace relations, and the paper finds three-body resonances linking every planet in the system.  The article uses it as its example of the most extreme resonant chain it describes.

    1. 22a.

      The seven TRAPPIST-1 planets form a chain in which each adjacent pair lies close to a two-body mean-motion resonance and successive triples satisfy three-body Laplace relations — a configuration the authors attribute to migration within the protoplanetary disk. ↩

  23. 23.

    Lee, Chin-Fei.  “Angular Momentum Evolution from Core to Disk Scales in the Early Phase of Star Formation: Constraints from HH 212, HH 211, and B335.”  arXiv.  Jul 30, 2026.  A preprint that re-analyses published observations of three young protostars (HH 212, HH 211 and B335) rather than making new measurements.  It fits the inside-out collapse model to the measured profiles of specific angular momentum (spin per unit of mass) and finds the match broadly good, though the author says magnetic braking or different initial conditions cannot be excluded as the reason for the small disk in B335.  The article uses it as the latest of three protostars whose profiles have been mapped from core to disk.

    1. 23a.

      “I find that the observed angular momentum profiles in all three sources are broadly consistent with an inside-out collapse scenario with approximate conservation of specific angular momentum in the collapsing region.  The inferred collapse ages and mass infall rates yield total masses accreted onto the center that are broadly consistent with the central masses derived from kinematics, allowing for a fraction of the material to be ejected by jets and winds.” ↩

  24. 24.

    Bailey, E. et al.  “Solar Obliquity Induced by Planet Nine.”  The Astronomical Journal 152:126.  2016.  A theoretical study that tests whether a distant, eccentric planet, which the authors call Planet Nine, could tilt the sun's spin axis away from the plane of the planets' orbits.  The authors use an analytic model of the gravitational pull between that planet and the giant planets, and they find that a planet with similar parameters can produce the observed tilt and the sun's pole position from a nearly aligned start.  The article uses it to show that the explanation for the tilt is still unsettled.

    1. 24a.

      “The six-degree obliquity of the sun suggests that either an asymmetry was present in the solar system's formation environment, or an external torque has misaligned the angular momentum vectors of the sun and the planets.  However, the exact origin of this obliquity remains an open question.”  The measured value of the solar obliquity to the mean planetary plane is about 7.2 degrees; the six-degree figure is relative to the plane the authors adopt. ↩

  25. 25.

    Tychoniec, Łukasz et al.  “Dust masses of young disks: constraining the initial solid reservoir for planet formation.”  Astronomy & Astrophysics 640:A19.  2020.  A paper that measures the dust in the disks of young protostars (less than half a million years old, in the Class 0 and Class I stages) in the Perseus star-forming region, combining ALMA and VLA radio telescope data.  The median dust masses are about 158 Earth masses for Class 0 disks and 52 for Class I, but these depend on the dust opacity assumed, and the ALMA-based values are lower limits.  The article uses it as the main evidence that the solids in the youngest disks could be enough to build the observed giant planets, if planet formation starts in the Class 0 phase.

    1. 25a.

      “The median dust masses of the embedded disks in Perseus are 158 M⊕ for Class 0 and 52 M⊕ for Class I from the VLA fluxes. … The dust masses of young Class 0 and I disks are larger by at least a factor of 10 and 3, respectively, compared with dust masses inferred for Class II disks in Lupus and other regions.” ↩

    2. 25b.

      “The solid content in observed giant exoplanets can be explained if planet formation starts in Class 0 phase with an efficiency of ∼ 15 %.  A higher efficiency of ∼ 30 % is necessary if the planet formation is set to start in Class I disks.” ↩

  26. 26.

    Segura-Cox, Dominique M. et al.  “Four annular structures in a protostellar disk less than 500,000 years old.”  Nature 586:228–231.  2020.  A Nature paper that uses the ALMA radio telescope array at 1.3 mm to image the dust in the disk around IRS 63, a Class I protostar (a young star still wrapped in a dense envelope of gas and dust) less than 500,000 years old.  The image, at 5 au resolution, shows two bright rings and two dark gaps, and the authors say it is ambiguous whether planets are carving the gaps.  The article uses it as its example of rings in a disk that is this young.

    1. 26a.

      “Here we present 1.3 mm dust emission observations with 5 au resolution that show four annular substructures in the disk of the young (<500,000 years) protostar IRS 63.” ↩