Project Description

The goal of this work is the development of an approach to modeling moderate-to-small size earthquakes throughout central Asia. The idea is to use the joint inversion of seismograms and surface-wave amplitude spectra in a joint grid search for strike, dip, rake, depth, and seismic moment. The grid search is performed using an L1 misfit norm. We are in the process of using Harvard-CMT studied events to calibrate the method and to examine the performance of different weighting schemes.

Station Performance • Spectral Misfit Compilations

For each station and event we can compute the misfit as a function of period. These summaries can be used to assess the value of the station for our particular application and identify stations that are difficult to fit. Many factors can affect the misfit at a station, including the earthquake faulting parameters, the assumed earth structure used in the modeling, the geology along the path from source to station, the ambient background noise, or errors in the assumed instrument. The point of these summaries is to help us sort out these various sources of error. Geology is the most likely cause of many problems, and we plan to eventually use this information to construct tomographic correction maps for this type of modeling. For now, these are crude summaries of the inversion/station perfomance. The data are not weighted in these summaries - in the inversion, each observation is weighted by seismic background noise estimates, distance, azimuthal distribution.

Harvard Faulting Geometry (optimized moment and depth)

Complete Grid Search for mechanism, moment, and depth

Office: (814) 865-2310
FAX: (814) 863-7823
FAX: (814) 863-8724
E-Mail: cammon@geosc.psu.edu