Abstract
We present a new focal mechanism stress inversion technique to produce
regional-scale models of stress orientation containing the minimum
complexity necessary to fit the data. Current practice is to divide
a region into small subareas and to independently fit a stress tensor
to the focal mechanisms of each subarea. This procedure may lead
to apparent spatial variability that is actually an artifact of overfitting
noisy data or nonuniquely fitting data that does not completely constrain
the stress tensor. To remove these artifacts while retaining any
stress variations that are strongly required by the data, we devise
a damped inversion method to simultaneously invert for stress in
all subareas while minimizing the difference in stress between adjacent
subareas. This method is conceptually similar to other geophysical
inverse techniques that incorporate damping, such as seismic tomography.
In checkerboard tests, the damped inversion removes the stress rotation
artifacts exhibited by an undamped inversion, while resolving sharper
true stress rotations than a simple smoothed model or a moving-window
inversion. We show an example of a spatially damped stress field
for southern California. The methodology can also be used to study
temporal stress changes, and an example for the Coalinga, California,
aftershock sequence is shown. We recommend use of the damped inversion
technique for any study examining spatial or temporal variations
in the stress field.
Users
Please
log in to take part in the discussion (add own reviews or comments).