How Long Does it Take to Form a Slot Canyon?

Print Article
Posted on November 26, 2010

Some of the most striking features of the southwest are the slot canyons – the narrow, winding defiles in sandstone that can be well over a hundred feet deep and go for miles (photo).  A whole culture of slot canyoneering takes on the challenge of hiking through them, and the amazing patterns of reflected light that develop in them make them major targets for photographers (photo).  Each year, flash floods scour the walls and floors of these canyons, changing them slightly, but hikers often notice little change in depth from year to year.  One might think that it would take many eons for these canyons to form.  A team of geologists decided to test erosion rates by watching an actual incipient canyon starting during floods in the Henry Mountains, Utah.  They monitored the site over three years and reported their findings in the Geological Society of America Bulletin.1
    The researchers studied a stepped slope created by humans that provided “a well-constrained initial geometry of a steep, unchannelized bedrock slope” and found a remarkable amount of cutting – half a meter – in just 23 days of flooding from seasonal snowmelt.  If that were the averaged annual rate, a slot canyon 100 meters deep could be cut in just 200 years.  Most likely other factors could slow down the rate significantly, but even if it took a thousand years, or 5,000, that presents a problem: why is there any sandstone left after millions of years?  They seemed to recognize the issue while adhering to the geological time scale:

Rates of fluvial bedrock incision mimic rates of external landscape forcing (e.g., tectonic uplift and eustacy [sic]) when averaged over geological time scales, but local rates of channel downcutting into bedrock can be fast during the individual floods that actually drive bedrock incision: we measured up to 1/2 m of local vertical incision into bedrock over 23 days of snowmelt runoff (Fig.9).  Local channel morphology and high but not overwhelming rates of sediment transport enabled such a high local erosion rate.  The local thalweg2 slope was high (~20%, Fig. 8), and the cross-sectional morphology of an inner channel focused flow and sediment transport over a narrow zone where almost all erosion occurred.  While poorly constrained, field measurements demonstrated high rates of coarse-sediment transport.  Additionally, preexisting inner-channel alluvium was entrained during this snowmelt runoff event, and so alluvial cover was not consistently present to mantle the inner-channel bed and inhibit bedrock erosion.  Field observations also suggest that thresholds of detachment for abrading the local sandstone are negligible (Fig. 6).

The bulk of erosion occurred during snowmelt in April and May, and then again during flash floods from thunderstorms in the summer and fall.  These events “show rapid increases followed by approximately exponential reductions in flow depth, with durations measured in hours, not days.”  One single flood on October 5, 2006, carried nine cubic meters of material per second, and left five inches of coarse sediment in the channel, accounting for 90% of the total flash-flood flow volume.  This was in a year with no snowmelt.
    Surprisingly, they found that small floods can cause more erosion than large floods.  This means that uniformitarian rates are sufficient to cause rapid erosion in these canyons.  They gathered a lot of sediment in a trap during just normal flow periods.  It’s not so much the size of the flood, but the size of the particles, cobbles and boulders carried along, that cause the most erosion.  As they said, boulders and particles detach easily, and become hammers against the rock downstream.
    Their eyewitness evidence contradicts other dating methods, they said with some confusion: “Nonetheless, the rapid short-term erosion rate suggests an interesting and converse question: why are long-term fluvial bedrock incision rates so much lower?  For example, Cook et al. (2009) reported long-term incision rates of ~0.4 mm/yr based on cosmogenic dating of alluvial terraces along a well-adjusted channel in the Henry Mountains, consistent with regional measurements of long-term incision (Garvin et al., 2005).”  They did not try to reconcile the numbers.  Erosion in Taiwan has been estimated at 10mm/year – but this team witnessed 500mm in just 23 days.  Maybe these things are not well understood because “bedrock-eroding floods only occur rarely.”
    They claim that their research site is similar to that of well-known slot canyons of the Escalante River:

Erosion formed a narrow inner channel with rough sidewalls.  This transient bedrock channel morphology is consistent with other natural slot canyons, in particular the Coyote Gulch narrows, Peek-a-boo slot, and Spooky slot canyons that are tributaries to the Escalante River in southern Utah.  These canyons also incised from an initial condition of flow over steep, unchannelized bedrock slopes following channel diversions by sand dunes.  The erosional topography in all of these cases is consistent with feedbacks between flow, sediment transport, and erosion observed in flume experiments (Finnegan et al., 2007; Johnson and Whipple, 2007).

Earlier, they had stated, “We argue that the similarity in channel morphology between the monitored channel and these natural slot canyons suggests that the feedbacks we interpret between sediment transport, bedrock erosion, and channel morphology are commonplace.”  In brief, their experimental site is likely a good indicator of what happens naturally.
    The authors also helped answer another puzzle: the formation of potholes.  Steep, round holes in sandstone and granite are well known to hikers.  “Finally, our monitoring fortuitously captured the erosion of a bedrock pothole,” they reported.  Though it only grew about 20 inches deep, and one of them eroded away the following season, they were glad to catch it in the act, because “surprisingly little is understood about their formation.”  They explained in the conclusion, “We interpret that it formed by impact wear from coarse sediment rather than fine suspended load, although distinctions between bedload and suspended load may be less meaningful because localized incipient suspension of larger clasts is required to keep deposition from occurring inside the pothole.”
    The word “rapid” appeared 13 times in their paper, e.g., “Through field monitoring we demonstrate that (1) short-term bedrock channel incision can be rapid, (2) sustained floods with smaller peak discharges can be more erosive than flash floods with higher peak discharges, due to changes in bed alluviation, and (3) bedrock channel morphology varies with local bed slope and controls the spatial distribution of erosion.”  The words “slow” and “gradual” were found, but not in reference to the rate of channel formation.  Furthermore, the authors made no attempt to incorporate their measured rates into a long-ages context.  The sandstones they studied (04/24/2003) are labeled Jurassic, with estimated ages around 190 million years.  Has erosion this rapid been occurring all that time?


1.  Johnson, Whipple and Sklar, “Contrasting bedrock incision rates from snowmelt and flash floods in the Henry Mountains, Utah,” Geological Society of America Bulletin v. 122 no. 9–10 (Sept, 2010), first published online May 2010, pp. 1600–1615, doi: 10.1130/B30126.1.
2.  Eustasy refers to global sea level independent of local factors.  Thalweg or “valley line” is the deepest continuous line along a river valley. 

For fun, let’s take the lowest rate mentioned (0.4mm/year), divide by 4 to get a tenth of a millimeter per year, and extrapolate it over the assumed age of the Navajo sandstone, 190 million years.  That would give you a canyon 19 km deep!  That’s over 15 times the depth of the Grand Canyon, eroding at the slowest rate claimed from cosmogenic radionuclide dating, not actual field observations (if we took their measured rate of .5m/yr, it would be a canyon 95,000 kilometers deep).  By contrast, if you took the very conservative value measured in Taiwan of 10mm/year, you could get a decently deep canyon 50 meters deep (similar to many observed) in just 5,000 years.
    Real erosion is much more complicated, of course, and this simplistic calculation does not take into account many factors.  It should provide a ball-park estimate of reasonableness, though.  Is it reasonable to conclude that erosion anywhere near this rate has been going on for 190 million years?  Arguably, erosion rates have often been higher in the past.  The Biblical time scale for the formation of these canyons begins to look much more reasonable, given that a worldwide flood in that model would have made erosion rates vastly higher for a considerable amount of time (photo).
    Once again, when actual field observations are undertaken, the evidence cries out against the vast eons of time alleged by the evolutionists and secular geologists (for three other examples, see 11/13/2006 about a volcanic field in Nevada, 03/05/2006 about Grand Falls of the Little Colorado, and 01/12/2007 about the fossil forests of Yellowstone).  It’s interesting that these geologists in Utah dared not explore the implications of their own findings for the geologic column and uniformitarianism.
    Your commentator has been to all three slot canyons mentioned in the article (Peek-a-Boo Gulch, Spooky Gulch, and Coyote Gulch) and can confirm that they are fascinating targets for remote wilderness adventure hikes, for those so inclined.  See the Canyon Color and Rock Art sets in the Creation Safaris photo gallery for views from this area, and go exploring if you can (photo).

No comments have been left

Leave a Reply