29 January 2018

Rattlesnake Hills rockslide – anticipating future behaviour

Posted by Dave Petley

Rattlesnake hills rockslide – anticipating future behaviour

The Rattlesnake Hills rockslide continues to creep.  Washington State DNR (WADNR) have released a drone video shot last Monday (22nd January 2018) that shows the state of the landslide at that point:-


This shows further development of the graben structures at the crown of the landslide:

Rattlesnake Hills rockslide

The graben forming at the crown of the Rattlesnsake Hills rockslide in Washington State, via a WADNR Youtube video.


Meanwhile the lateral scarp has also developed considerably, showing both lateral displacement and some subsidence as the landslide mass deforms:

 Rattlesnake Hills rockslide

The lateral scarp of the Rattlesnake Hills rockslide, via a WADNR drone video on Youtube.


Phys.org has a decent article from late last week about the movement of the landslide, and the problems of anticipating its future behaviour.  It seems to indicate that at present the landslide is displacing at an approximately constant rate of around 30-40 cm per week.  Opinions differ about whether the slide will accelerate to a rapid (or even catastrophic) failure event.  The consulting engineers, Cornforth Consultants (who have a strong reputation in this field) think that a rapid event is not likely:

The most likely outcome is that the slide will continue to move slowly, with much of the 4 million cubic yards of rock and soil spilling into the quarry pit at the base of the hill, according to Cornforth Consulting, the geotechnical engineering firm hired by quarry operator Columbia Asphalt and Ready-Mix.

Others are less sure – Jeff Moore of the University of Utah cautions against assuming this is the case:

Moore describes that as an “optimistic” scenario and says it’s important to plan for the possibility of a catastrophic failure.  “If the thing fails all at once, it’s well within the range of possibility that it could hit the highway,” he said. If the slide slumps in stages, then the potential for damage is likely to be lower.

Unfortunately this prolonged period of approximately constant velocity movement is a poor guide as to what might happen.  Indeed this type of creeping behaviour is well-established and has been widely observed in large rockslides.  A simple, three phase creep model is usually used to describe the patterns of movement seen as large rockslides evolve:

Rattlesnake Hills rocklide

The three phase creep model of rockslide failure.


This landslide appears to be in the secondary creep phase at present.  In some slopes the transition to a tertiary creep phase does not occur, and the slope deforms slowly through time.  In others the slope accelerates to failure.  This behaviour is controlled primarily by the characteristics of the materials forming the landslide, which (as the images above show) are accumulating damage at present in the case of the Rattlesnake Hills rockslide.  In many cases this damage leads to a weakening, such the resistance to movement declines, allowing the slope to collapse rapidly.  In other cases, the materials are able to maintain their strength as they deform, and no transition to tertiary creep occurs.  Other factors controlling this behaviour will include the geometry of the slope (which in this case is very complex) and the response of the mass should significant rainfall occur.  In many cases the best guide might be the behaviour of other rockslides in the same materials, with evidence being drawn from landslide deposits in the region.  The Phys.org article draws comparisons with the Oso landslide, although in fact the two systems are quite different.  However, in the case of Oso the LIDAR data demonstrated clearly that previous slides had failed catastrophically.

At present the WADNR are doing the right thing in monitoring the slope carefully and ensuring that local residents are not at risk.  Thereafter this is a waiting game, one that is beset with uncertainties.