Dating Capstone — Appendix A

Whether a post-event moisture engine can build the observed polar ice, and at what efficiency it would have to run to do it.

Appendix A — Polar Ice Volume Budget, Moisture Supply, and Front-Loaded Accumulation

A.0 What this appendix claims, and what it does not

Because the polar ice sheets are, on this model, post-event products — the pre-event world had no perennial ice, for reasons given in A.1 — the entire ice record must be produced within the post-event window. That is a heavier burden than the other layer-counting records carry, none of which can lean on a pre-event bulk. This appendix meets the burden at the level the evidence supports: it shows the volume target is defined, the moisture supply is adequate by one to two orders of magnitude, and the required delivery efficiency is comparable to the value the present Earth already achieves. It does not claim a working climate model. The conversion of adequate supply into the observed ice, at the observed rate and no more, is the invited modeling; what is shown here is that the forced inputs are consistent with that outcome rather than requiring adjustment to reach it.

A.1 Why the ice is entirely post-event

The pre-event configuration precludes perennial ice. The landmass was concentrated in lower, warmer latitudes; the atmosphere was warmer and more uniform, without the steep equator-to-pole gradient that modern glaciation requires; relief was confined to the sutures rather than distributed as continent-wide ranges; and the poles, while cool, were not cold enough for long enough to retain snow through the year. There was neither the location, the climate, nor the terrain for ice sheets. Ice becomes possible only after the event: continents rifted to high latitudes, the thermal gradient steepened by the bimodal ocean, and the moisture engine switched on. The ice age is therefore not an awkward fact the model must absorb — it is an entailment of the event. And it follows that the whole of the ice-core record, not merely a recent cap, falls within the post-event interval and must be accounted for there.

A.2 Volume target

Modern grounded ice volumes (standard published values):

Greenland Ice Sheet: V_Gr ≈ 2.96 × 10⁶ km³
Antarctic Ice Sheet: V_Ant ≈ 26.4 × 10⁶ km³ (grounded)

giving V_modern ≈ 29.5 × 10⁶ km³.

Ice-age excess — the additional ice present at the model's ice-age maximum and later lost — is taken at the lower end of conventional Last Glacial Maximum excess estimates, ≈ 45 × 10⁶ km³, predominantly Northern-Hemisphere continental ice. Taking the low end is deliberate: it makes the volume the model must produce smaller only modestly while keeping the figure defensible.

V_total = V_modern + V_excess ≈ 75 × 10⁶ km³.

Subtraction of the modern-rate tail. Once the basins cool to near-modern surface temperatures, accumulation proceeds at ordinary modern rates and is common to any chronology; that late portion is removed from the differential budget. At a modern combined polar accumulation rate R_m ≈ 2.25 × 10³ km³ yr⁻¹ over the ~3.7 × 10³ yr remaining to the present — the elapsed interval since the event, less the elevated-accumulation interval derived in A.4 — the tail is V_tail ≈ 8.3 × 10⁶ km³. The elevated volume that must be produced while the basins are still hot is therefore

V_e = V_total − V_tail ≈ 67 × 10⁶ km³.

The tail interval and V_e are mutually dependent — the tail is what remains after the elevated interval, and the elevated interval is V_e divided by the required rate — so both are solved together rather than assumed. The solution is stable: V_e ≈ 66.7 × 10⁶ km³ over an elevated interval of ≈ 2,080 yr, leaving ≈ 3,700 yr of modern-rate tail.

A.3 Moisture supply

The supply term is set by the area of hot new-basin water and its evaporation rate, both of which the model already owns from the tectonic and thermal architecture rather than assuming for the ice.

Basin area from the spreading history. The separation history is the velocity profile established in the foundational standalone, v(t) = 12 exp(−t/417.5) km yr⁻¹, taken as the full opening rate (both flanks — its decay constant follows from the observed ~5,000 km separation and the peak the force balance produces, so no additional factor of two is applied). Integrating gives basin width w(t) = 12 × 417.5 × (1 − e^(−t/417.5)) km, with an asymptote near 5,010 km consistent with the observed Atlantic width. Over a combined Atlantic-plus-Indian rift length L ≈ 20,000 km (order-of-magnitude working length for the two new basins; not a surveyed ridge inventory), the open-water area grows as the basins widen. Because the basins are widening while they boil, the supply comparison uses the time-averaged open-water area over the boiling phase, not the final area. The boiling-phase duration is taken from Trigger Appendix F (water-side flux cap, ~310–478 yr across the 10–20 kW/m² cases). Across that band the time-averaged area is ≈ 25–42 × 10⁶ km², for which the representative order-of-magnitude value is

A_avg ≈ 30 × 10⁶ km².

(The final, fully-open combined area is several times larger; it is deliberately not used for the boil-phase supply.)

Flux. At the model's rift-basin evaporation rate of 100–1,000 mm day⁻¹ — against a modern open-ocean baseline near 3 mm day⁻¹, a factor of 30–300× per unit area — the boil-phase moisture flux is

Flux = A_avg × evaporation rate ≈ 1 × 10⁶ to 1 × 10⁷ km³ yr⁻¹.

Against a modern global ocean evaporation of ≈ 5 × 10⁵ km³ yr⁻¹, the event boil-phase flux is therefore roughly 2 to 20 times modern global ocean evaporation — from two rift basins, not the whole ocean.

A.4 Required accumulation rate, and why the "peak" is an average

The elevated volume divided by the elevated-accumulation interval (≈ 67 × 10⁶ km³ over ≈ 2,080 yr) gives a required average rate over that interval of ≈ 32,000 km³ yr⁻¹ — about 14× the modern polar rate. This is a genuine confession: it is what the fixed volume and the interval demand, not a tuned value.

A note on the accumulation shape, to prevent a misreading. The real production history is not a plateau. Production climbs as the basins widen (more hot surface area) while they hold at 100 °C, peaks near the end of the boiling phase, and then decays as the ocean cools and widening slows — a rise-to-peak-then-decay curve. That curve is a supply and accumulation-rate history. It is not a count of seasonal bands. For the supply question the shape does not matter: the volume target is fixed, and all that is asked is whether supply meets it. A rectangle-plus-decay is used only as a legible proxy for the integral. Consequently any single rate quoted for the boiling phase is an average over that phase, not an instantaneous peak. Solving the proxy integral for that boiling-phase average gives roughly 30× the modern rate (order-of-magnitude "tens of times modern"); its exact value shifts with the assumed boil duration and cooling timescale, which is precisely why it is reported as an order-of-magnitude average and not a headline figure. The robust results are the fixed volume integral and the strongly front-loaded character — not any particular rate.

A.5 The efficiency the model requires — a benchmark

Supply and demand meet through an efficiency: the fraction of gross basin evaporation that survives transport to high latitudes and is retained as grounded ice. Rather than assume that efficiency, it can be solved for as a benchmark — requirement divided by gross supply — which is forced arithmetic with no free parameter.

Required end-to-end efficiency = 32,000 km³ yr⁻¹ ÷ (1 × 10⁶ to 1 × 10⁷ km³ yr⁻¹) ≈ 0.3% to 3%.

The model requires that only three-tenths of one percent to three percent of gross basin evaporation end up as retained polar ice. For reference, the modern ratio of polar accumulation to global ocean evaporation is ≈ 0.45%. The required event efficiency thus brackets the value the present Earth already achieves — below it at the high-supply end, and at most several times it at the low-supply end. The ice budget closes, in other words, without positing any enhancement of atmospheric efficiency over the modern value; it needs only the far larger source the thermal model already supplies, processed at an efficiency in the neighborhood of the one the planet runs today.

This comparison is an analog, not an identity — the modern ratio is a whole-planet steady-state quantity, while the required figure applies under a different basin geometry and a steeper gradient — so it is a plausibility anchor, not a proof. But it converts the delivery hand-off from a bare gap into a bounded one: whatever the coupled model returns, the efficiency it must find is modest and has a real-world precedent.

A.6 Oversupply is required, not convenient

That the required efficiency is a small percentage is not a weakness to be explained away; it is a necessary feature. Most delivered moisture cannot become retained ice. A large fraction falls on open ocean or on sea ice that later melts; more ablates on warm ice-sheet margins or runs off. Net grounded-ice retention is therefore inevitably a small fraction of gross evaporation, and gross transport must exceed net accumulation by a wide margin. The one-to-two order-of-magnitude gross margin established in A.3 is exactly what pays for both losses — transport and retention — and still closes the net budget. Oversupply at the source is the model working as it must, not a factor smuggled in to cover a shortfall.

A.7 Enough, but not too much — the absence of a snowball

A supply this large invites the opposite worry: if delivery is efficient, does the model bury the planet in ice — a snowball rather than an ice age? The physical answer is that the same heat that drives the moisture engine also opposes runaway glaciation, and it opposes it several ways at once. Snowball glaciation is an ice-albedo runaway on a cold planet; this world is hot, its oceans radiating latent and sensible heat globally, so the albedo feedback cannot gain traction while the equatorial ocean is a heat source. The ice-making phase is bounded by the same declining thermal forcing that ends the boil — as the basins cool, the engine winds down, so accumulation cannot run unbounded. And a hot world with warm margins and vigorous circulation supplies abundant energy for ablation, so high delivery is met by high melting and the net stays bounded. The expected outcome is therefore a glaciation that advances to mid-latitudes and later recedes — an ice age — not a snowball.

This lands the model in a narrow window: enough ice to glaciate, not so much as to snowball, receding on the thermal-relaxation timescale. That window is hit by numbers fixed elsewhere — observed ice volumes, the velocity-curve basin area, the thermal-model evaporation rate, an efficiency near the modern value — none tuned to produce it. A snowball outcome would falsify the model; a barely-glaciated one would falsify it; the forced inputs land between. The fit is therefore a result, not an assumption: the model was exposed to coming out wrong in both directions and, at order-of-magnitude, came out consistent with the observed ice age. What it is not is a proof — the coupled balance that would confirm the window is threaded remains the invited modeling.

A.8 What the cores should carry — and what is not claimed

The supply curve in A.4 is the claim: accumulation is front-loaded, rising while the rifts widen and still boil, then decaying as the basins cool. That is a rate history. It is not a statement that the ice is built of annual bands.

During the high-supply interval the expected deposit is thick, wet, and chemically mixed — dust, salt, and volcanic load arriving with almost continuous snowfall rather than as a neat seasonal couplet. After the engine relaxes, ordinary seasonal banding can appear. The change with depth is therefore a change in character of the ice (rate, chemistry, how continuous the fall was), not a published count of “layers per year” at the base.

Reading that character out of a real core is the ice-sheet modeler’s job. This appendix does not treat a boil-phase fabric as a finished discriminator. What it does claim is the volume integral, the supply margin, and a front-loaded rate curve any coupled model has to roughly satisfy.

A.9 Summary

The volume target (~67 × 10⁶ km³ elevated) is set by observed ice volumes at the low end of the excess range. The moisture supply (~10⁶–10⁷ km³ yr⁻¹, some 2–20× modern global evaporation from two basins) exceeds the required average accumulation by one to two orders of magnitude. The end-to-end efficiency the model requires (~0.3–3%) brackets the value the modern Earth already achieves, so the budget closes without enhanced atmospheric efficiency. Oversupply is structurally required, since retention is a small fraction of gross supply, and the large margin pays for it. The same heat that drives the supply plausibly self-limits the result to an ice age rather than a snowball. What is claimed is a defined target, a verified supply margin, a benchmarked efficiency, a front-loaded accumulation-rate curve, and a forced-not-tuned consistency with the observed outcome. What is not claimed is the coupled atmosphere–ice-sheet model that would convert this adequacy into the observed ice in detail, or a year-count reconstructed from the deep ice — those remain the development targets these results define.

← Return to paper · Appendix B →

AI Collaboration Disclosure: Drafting and calculations by Claude (Anthropic), with adversarial review by Grok (xAI), under the direction of D. L. White. Neither AI system endorses all conclusions as settled.