J. David Rogers'
Grand Canyon Research
Surprise Valley Landslides



In 1979 I returned to the Canyon to study the landslides in and around Surprise Valley, which I felt may be connected to the old lava dams a short distance downstream.  This shows an exfoliation A-tent in the Esplanade Sandstone above Surprise Valley.



The unofficial narrowest point on the river occurs near the beginning of 135 Mile rapid.  Only about 40 feet of unobstructed channel exists between some rocks that jut out out of the channel on either side (seen here). In the right background is an enormous buried channel of the Colorado River, with its deepest point about 250 feet above the present channel.


This is a closer view of the buried channel above the right bank at Mile 135, at the entrance to Granite Narrows. The old channel is filled with bedrock landslide debris.  This undifferentiated debris is shown to be contiguous with the much larger Surprise Valley Landslide Complex, which occupies an old side canyon between the Esplanade and Granite Narrows, which used to feed into Tapeats Creek (see geologic map below).


At Mile 134.94 the Colorado River experiences it narrowest point in the entire Grand Canyon, with a width of 76 feet between opposing canyon walls.  This determination was made in June 1965 by the USGS-WRD survey during a flow of 48,500 cfs.


Low level aerial oblique view of the Mile 135 Landslide, which slumped from the southern side of Cogswell Butte, into the Colorado River.  At the time the river channel was about 250 feet higher than present, making this an older event (likely between 150 and 200 Ka).  I took this picture from a low flying Cessna 206 in September 1979, before the Park Service instituted the present flight restrictions in the Canyon.


River view looking up the Granite Narrows below Surprise Valley, between Bonita (Mile 134.3) and Deer Creeks (Mile 136.3). In this reach the river reaches depths of up to 85 feet.


Portion of the 1976 Geologic Map of the Grand Canyon showing the Surprise Valley-Deer Creek-Fishtail Canyon area. The areas underlain by extensive bedrock landslides are shown in bright yellow. I became intrigued with studying these landslides during my research trip in June-July 1978.


Detailed geologic map of the Surprise Valley area that I made during the summer of 1979, working with Marv Pyles and Eric Savage at U.C. Berkeley.


This is the detailed landslide map of the Surprise Valley area we also complied during our reconnaissance work in the summer of 1979.  The areas are mapped according to the types of landslides we observed.


High level aerial oblique view of the eastern escarpment of Surprise Valley and the Thunder River Landslide complex.  We studied the Thunder River slide because it was the best exposed, with easy access to the basal slip surface.


Low level aerial oblique view taken in 1979 of the Thunder River Slide, showing its proximity to Thunder River Spring, seen at right.  The fault-controlled spring likely played a role in fomenting this movement, though softening of the shale through lake inundation or rapid drawdown can also be demonstrated as failure mechanisms.


We were able to access the basal slip surface of the Thunder River slide and take samples of the Bright Angel Shale for laboratory testing back at U. C. Berkeley.  A travertine cemented breccia zone between 30 to 40 feet thick lies above the basal slip surface, shown here.


This shows the curvilinear transition zone of the basal slip surface adjacent to Thunder River Springs, where it cuts through the upper Bright Angel Shale. The cemented character of the slide debris above the basal slip surface is ascribable to subsequent travertine cementation.


Trimmed specimens of the Bright Angel Shale were tested in simple shear, under dry and saturated conditions.  Saturation was achieved by placing prepared specimens in a consolidometer for 6 months under back-pressure and evaluating the pore pressure coefficients.  As far as we know, these were the first direct shear tests on intact rock under both dry and saturated conditions.  Normal pressures ranged between 0 and 1100 psi.  The tests revealed a 2/3 drop in cohesion upon saturation.  Such a significant loss of shear strength could occur if the Bright Angel Shale was inundated by the rising waters of reservoirs caught behind lava dams downstream.  If inundation were the triggering mechanism, the age of the Thunder River Slide is likely something on the order of 750,000 years before present (750 Ka).


Initial assumed condition of the Thunder River Landslide for our back-analyses.  This condition occurred when the Tapeats Creek and the proto-Surprise Valley stream had excavated to an elevation of just under 3000 feet, through the Boxcar/Rampart Cave Member of the Cambrian age Bright Angel Shale.  The failure mass was just under 2000 feet high!


The landslide initiated as a composite wedge failure, with one wedge sliding horizontally and the other dropping vertically, as depicted here.


As the passive wedge continued to translate towards Tapeats Canyon, a brecciated zone developed between the back-rotating block and the passive wedge.


Sliding of a passive toe wedge on a weak bedding plane, brecciation of a middle zone and back rotation of the heel of the slide mass, all appear to have occurred concurrently.  This is how the Thunder River Landslide appears today.  In 1979 we called it a “block glide with rotated graben”. Using current nomenclature this would be termed a “composite landslide”, because it exhibited two or more styles of movement simultaneously.

Questions or comments on this page?
E-mail Dr. J David Rogers at rogersda@umr.edu.