Press "Enter" to skip to content

Big Flash Floods can come from Small Catchments

Article by Paul Martz

Geology – September 2002 – Colorado Central Magazine

With one eye on the storm and running late, I was driving way too fast for conditions. But I needed to catch a shift of drillers before they left a remote drillsite an the southeast flank of the Pilot Mountains east of Mina, Nevada. The drillers had been leaving the site by cutting cross-country, making a new two-track in the process, instead of following the permitted route back to the county road on which I was driving.

Because I had not encountered them, I assumed the drillers were still somewhere between me and the thunderstorm. Then I got to a steep ravine that marked the edge of the prospect area, and it was immediately obvious that if they had left the site by the “correct” route, they were no longer with us.

What had been a boulder-laden, talus-strewn gulch with mature desert vegetation was now a flat, smooth palette from canyon wall to canyon wall. This particular drainage was over 100 yards wide, indicating that a monster wall of water had flashed down it, removing, burying, and ultimately sweeping it clean of everything. I hoped that didn’t include a pickup truck with three drillers.

An indication of the force of such an event is shown by one of the first commercial uses of the portable magnetometer. It was used to locate a Mallie steam engine, the largest ever built, that had been swept more than a quarter mile off the tracks of the Southern Pacific and completely buried along with the crew.

The drillers’ pickup could have been miles downstream, but as it turned out, they had left early by the wrong route because they were worried about lightening. Needless to say — after they viewed the smooth canyon floor and high water marks on the walls — I had a difficult time convincing them to ever use the right road again.

The storm that had flushed out that draw hadn’t been large, nor had it lasted a long time, but it had deposited a significant portion of its water and energy in a localized catchment basin that drained past our prospect, generating a rather typical but nonetheless impressive flash flood. A new small delta of debris down at the base of the range where the canyon emptied onto a playa lake was the only tangible record of the event.

Recently here in Chaffee County three people narrowly escaped death when a series of similar flows occurred upstream and uphill from Cottonwood Hot Springs on July 23rd. Although the event was incorrectly termed a “mud flow” by the press, it resulted from a torrential rainfall in a sequence of small catchments, and it basically reproduced the Nevada flash flood on a smaller scale.

[The word “catchment,” refers to the area drained by a river, or in this case by dry gullies which water flows down only on rare occasions.]

[Figure 1: Alluvial Fan Deposit]

The most impressive feature of the Middle Cottonwood Creek events is the very small size of the catchments which liberated substantial accumulations of sand and rock that were deposited in classic examples of alluvial fans over county road 306 and into the channel of the creek.

Figure 1 shows an alluvial fan deposit emerging from a catchment of less than 3 acres. The sheet of sand and rock flowed toward the viewer for over 150 yards, a vivid indication of the volume of water that was deposited in a brief time. The small size of the catchment can be judged by the apparent size of the telephone pole near the base of the slope.

[Figure 2: Alluvial deposit of debris flow]

Figure 2 shows the publishing team of Colorado Central providing scale for an alluvial deposit that resulted from a catchment only slightly larger than that shown in Figure 1. The flow of debris crossed the road and partially damned Middle Cottonwood Creek.

Figure 3 shows Ed Quillen well “over his head” in alluvial debris from the next catchment downstream from those in Figures 1 and 2. By coincidence each succeeding catchment becomes larger as one proceeds toward the hot springs. Thus there was more water in the catchment area, and more surficial debris was mobilized by the event.

[Figure 3: Alluvial debris]

What differentiates a flash flood alluvial deposit from a true mud flow? If you are in a minivan buried up to your chest in cold, wet sand and rocks, the distinction might escape you. But there really are significant differences.

In the Cottonwood event, the rocks and sand were mobilized by the turbidity flow caused by a storm which deposited a large volume of water onto steep slopes in a very short period of time. The resulting process eroded, or gullied, the weathered surface materials in the catchment area and redeposited it as debris fans at the base of the slope. Glacial ice had previously gouged away the canyon walls leaving sharply over-steepened slopes, which further increased the turbidity of the resulting currents and aided the movement of material downslope.

[Turbidity refers to the suspension of sediment in the water. In an event like a flash flood the water gets dense and murky with suspended silt and rock.]

Visualize a flow, mostly of water, carrying a suspended load of rock and sand as the result of the volume and turbidity of water that’s rushing down a very steep gradient.

Although a true mud flow is also the product of the collision of rocks and water; the resulting process is different, the key word being mud. Imagine a cold lava flow moving only slightly faster than a man can run. The material is mostly solids, rather than water, and therefore it has a much higher specific gravity. In other words, a true mud flow is a lot more dense and much more viscous, and it travels significantly slower than the Nevada and Cottonwood events. A mudflow, in fact, may have a duration of several hours, rather than the scant few minutes a flash flood is in existence.

Moisture accumulation (and not necessarily torrential rainfall) in a catchment, causes fluidization of soil, rock and sand whereby the total mass of material is the primary focus of the downslope pull of gravity. Because of their large mass and basically internal lubrication, mud flows can and do move over surfaces with significantly lower slope angles than those along Middle Cottonwood Creek.

According to reports in The Mountain Mail shortly after the event, the statement was made that the prevailing dry conditions “…contributed backwards. Had a similar torrential rain fallen in a normal May when the ground was already saturated with snow melt, the severity of the event would have been much greater and true mud flows might have resulted.

Although the amount of transported debris was impressive, it was actually limited by the shallow depth of weathering present. That’s because glaciation had scraped off earlier soil and debris producing fresh bedrock only a few inches below the surface. In contrast, tropical soils — where weathering may extend dozens or even hundreds of feet below ground surface — frequently produce true mud flows in seasonal monsoonal conditions.

When a hurricane comes onshore, the soil becomes saturated with water, increasing its mass and causing it to lose physical strength, which can produce truly catastrophic results. Those who remember the El Nino winter of 1997 may recall frequent TV broadcasts of mud flows in Southern California which resulted when the weight of whole hillsides, saturated by the continuing rains finally exceeded the shear strength of the soil, and gravity took over. In many, but not all cases, this was the result of localized over-steepening of slopes caused by construction cutting into the hillsides.

The natural uplift of mountain ranges, and the resulting steep gradients, provide a surface so sloped that mud flows can result from ordinary seasonal rains, or even heavy snow melt — even without the exacerbating effects of glaciation.

[Figure 4: Mud flow deposit, hammer for scale.]

Figure 4 shows a true mud flow deposit in Cottonwood Gulch about a mile east of Tenderfoot Hill behind Salida. A rock hammer with a 16-inch handle provides scale near the bottom of the outcrop. This is only one of several mud flows exposed by the down cutting of Cottonwood Creek as it tries to compensate for rapid uplift that has occurred subsequent to the formation of this flow.

To summarize: What happened on the Middle Cottonwood Creek above the hot springs was a typical flash flood resulting from torrential rains, not a true mud flow. Small catchment basins collected and localized the downpour, which subsequently eroded gullies in and below the catchments, and debris was deposited as alluvial fans over, across, and along county road 306.

One other thing is certain. The three individuals who survived being caught in those torrents should have gone and bought lottery tickets immediately afterwards.