Deposition Series

How Deep Did the Waters Cut?

Part Three of the Deposition Series. The dust settled. The rocks remembered. Now the ice melts and the water cuts. Meltwater discharge, episodic flooding, and the canyon that should not have taken millions of years.

River carving through post-catastrophe sedimentary layers in a deep canyon gorge

How Deep Did the Waters Cut?

Part Three of the Deposition Series

Paper 7 put a decaying wind-and-surge engine on a stripped continental surface. Paper 8 asked whether two published stacks remember that engine. This paper asks what happens when the water that built the column is later forced through a single channel.

The Grand Canyon is the test piece because the column it cuts is the one just scored, and because a spillover origin is already on the conventional table (Lake Bidahochi and its relatives). The question here is not whether a canyon can form. It is whether the same cork-pop decay that sorted the Tonto can then supply the water and the shaking to cut it.

1. What has to be true before a canyon exists

A sheet-flow sea does not cut a gorge. It builds or reworks a pile. A canyon starts when drainage concentrates: a sill fails, a barrier goes, or a broad wet surface finds one outlet and keeps it.

That is a change of state, not a new physics. Paper 8 already has the basin leaving “fill” and entering “drain / expose.” Carving belongs to that later state. It does not compete with Tapeats dune migration. It follows it.

Two water sources sit on the same weather engine:

Rain and runoff on a large catchment while the rift contrast is still strong, then declining with that contrast.
Local ice on the Rockies and the Plateau rim, built by the same moisture export. No pre-event ice sheet. Ice is a reservoir the engine filled; melt is that reservoir emptying as the engine relaxes.

Neither source needs a global flood still standing. The standing water is gone. What remains is a high country that was soaked and, in places, armored with ice, draining off a soft, young stack.

2. Continuous cut and pulse cut

Base flow does the daily work: abrasion, plucking, sapping. Young cement, a high water table, and a stack that has not had millions of years to lithify all move the same direction — faster than a mature canyon in hard, dry rock. Groundwater sapping undercuts; walls fail; talus becomes the next abrasive. Most of the volume that leaves a widening gorge leaves as gravity, not as grains in the thalweg. That split (a little river, a lot of wall) is ordinary canyon accounting. It is not a catastrophe special.

Pulses do the days that matter. A landslide dam in a narrowing, then a failure, puts a short discharge on the bed that base flow never sees. Costa’s 1985 dam-break scaling is the published relation used here for order of magnitude:

peak discharge (m³/s) ≈ 10.5 × (dam height in meters)^1.87

A 100 m barrier then yields a pulse of order 10,000 to 100,000 cubic meters per second, lasting hours. How often such barriers form is a function of shaking and of steep, wet walls. Shaking tracks plate speed. Plate speed follows the Trigger profile — start at 12 km/yr, time constant 417.5 years — so the pulse rate is front-loaded: more dams early, fewer late. This paper does not publish a flood census or a quake census.

Stream power rises with discharge. A pulse at tens of thousands of cubic meters per second in a few-hundred-meter slot is a different machine than a pre-dam Colorado at a few hundred. The first can take meters off a poorly cemented bed in a day. The second cannot.

3. The stack is cut in the order it was laid

Soft, barely cemented beds go first and fast. Carbonate and well-cemented sandstone go slower. Basement goes last and sets the inner gorge.

That order is the Tonto-plus-above column standing on Vishnu. Carving begins after the pile exists and while the rock is still young. What follows:

Early cut is through the upper, softer part of the Plateau stack. Depth accumulates fastest there.
The Redwall–Muav couplet is the first serious hard layer. Alcoves and undercuts are the expected form, not a smooth slot.
Tapeats is a thin, tougher floor above basement. It slows the vertical, not the lateral.
Vishnu is the last problem. Narrowing concentrates power. Plucking and fracture do more of the work than polishing. The inner gorge is that remainder.

Published canyon relief in the central Park is hundreds to about two thousand meters depending on rim and river. The model’s claim is that a young, pulsed, well-watered cut on this stack can reach that relief inside the post-event window. It is not a claim that a particular meter of the South Rim was cut in a particular year.

4. Bidahochi and the first outlet

A large interior water body finding a southwest outlet and keeping it is compatible with this paper and with a family of conventional spillover readings. This paper does not need to own Lake Bidahochi. It needs a concentrated drain. If the first outlet was a sill failure, the first pulse is a dam-break by another name. If it was progressive notch-cutting, the first pulse is smaller and the later landslide dams do the episodics. Either way the channel, once locked, is the tool.

5. What would break this

A column that was already mature, dry, and fully lithified before any canyon existed — the rates above assume a young stack.
No plausible catchment and no local ice or high-country rain in the window — then there is no water.
A requirement that every meter of the inner gorge be cut by modern-scale discharge over millions of years as the only working history. That is the conventional default. It is not a field measurement that this paper contradicts on day one; it is the competing integration. Dated lava dams and terrace sequences are the measurements that have to be read against both integrations, and they are not re-litigated here.

6. What this paper does not claim

It does not date the canyon to an integer year.
It does not publish a flood census or a quake census. Front-loaded pulses follow from (v(t)).
Costa 1985 is used for peak discharge given a barrier height, not as a formation-rate census.
It does not route a global ice sheet through the Colorado. Local highland ice and catchment rain are the water.
It does not claim the walls fell on a timetable. Gravity is the volume term; the river is the slot.
It does not treat a Bidahochi spillover as proof of the cork-pop event. It is one way a drain can start.

7. Close

The waters that sorted the Tonto were a sheet. The waters that cut the Canyon are a river, then a pulse, then a river again, on a pile that had not had ages to harden. Same engine, later state. How the last of that water left the high country, and what the conventional scorecard looks like against this one, is the next paper.

← What Did the Rocks Remember? Deposition Series The Convention or the Clock? →

AI Collaboration Disclosure: D. L. White directed the inquiry and made all substantive determinations. Claude (Anthropic) and Grok (xAI) assisted with drafting and review. Neither AI system endorses all conclusions as settled.