Distribution of Precarious Rocks Away from the San Andreas
  Fault: Data for Constraining Strong Motion Attenuation
  Relationships for Great Earthquakes

James N Brune (Seismological Laboratory, University of Nevada-
  Reno, Reno, NV  89554; 702-784-4974; 
  email:  brune@seismo.unr.edu) 

There are no strong motion accelerograph data for great strike slip
earthquakes along the San Andreas fault because modern instruments have
only been developed after the most recent such earthquakes, in 1857 and
1906.  Estimates of earthquake hazard for California are critically
dependent on the assumed relationships for such earthquakes, which have
had to be estimated by extrapolation from the data base for much
smaller earthquakes, and thus are quite uncertain.

Precarious rocks, which have been in place for thousands of years, and
have certainly survived historic and geologically very recent
earthquakes, can  place upper bound constraints on the ground motion
from these greatearthquakes, provided quantitative testing of the
ground motion required to topple these rocks is carried out.  The
methods for doing the testing have already been initiated, but need
further development, and there needs to be more extensive computational
and physical testing of large numbers of rocks to improve the validity
of the statistical results.

An important task in further development of the methodology is to study
regions where there are numerous outcrops of the type that produce
precarious rocks distributed at a range of distances from known large
earthquakes.  Such a region exists around the Mojave section of the San
Andreas fault, which ruptured in the great 1857 earthquake. Excellent
granite outcrops occur at all distances from the fault trace.  Near the
fault there are no precarious or semi-precarious rocks, and the
outcrops demonstrate an appearance suggestive of intense shaking.  As
distance from the fault increases semi-precarious rocks (rough upper
bounds on acceleration about .4-.6g) begin to appear at about 15 km,
and precarious rocks (upper bounds about .2-.4g) at about 25 km.
Although these estimates are only approximate, they clearly suggest
that with further quantitative study important ground motion
constraints could be obtained at a minimal cost compared to long term
operation of strong motion accelerographs, and could  provide critical
earthquake hazard information that can be obtained in no other way
prior to the next great earthquake along the San Andreas fault.
Examples of precarious rocks in this region will be shown,  along with
a location map and rough comparisons with attenuation relationships for
large earthquakes published by various authors.