ESTIMATION OF UNCERTAINTY FOR EARTHQUAKE SOURCE INVERSIONS

Seismologists utilize seismic waveforms in order to estimate the moment tensor (MT) on a routine basis; however, it is an inverse problem, and the result has limited reliability. Therefore, a realistic estimate of the uncertainty of MT is of crucial importance in evaluating solution quality. In general, two major sources of uncertainty can be recognized: observational and modeling uncertainties. The first one is associated with the data error, which may be owing to instrumental or ambient noise effects. Among the second source of uncertainty, we focus here on the uncertainty of Green’s functions (GFs) –the most dominant source of modeling error– owing to the inaccuracy of the considered crustal model. Here we briefly describe the linear inverse problem of finding the MT, methods to estimate the aforementioned two sources of uncertainty, and finally assessing the uncertainty of earthquake focal mechanisms. In the linear MT problem in a given earthquake origin time and hypocenter, the observed data d and model parameters m are related by:d=Gm

Keywords:

Moment tensor inversion , Focal mechanisms , Uncertainty estimation , Data error , Green's function