Finite slip inversions from far-field seismic data: How much can we hope to resolve?
Abstract
Earthquake finite-slip inversions have large uncertainties in practice due to unevenly distributed observations, poorly known velocity structures, and generally ill-conditioned inversion problems. Results from different groups for the same earthquake using similar data often exhibit large variations. To quantify the theoretical resolution limits in the source inversion problem, we study an idealized synthetic model of a fault embedded in an uniform whole space in which the Green's functions are exactly known for a distant sphere surrounding the source. For slip models of varying complexity, we examine the effects of limited versus complete station coverage, the frequency band of the observations, and various imposed regularizations. We plan to address the following questions: (1) For a given frequency band, how many stations and what station distribution is required to learn as much as possible? (2) What is the relationship between frequency band and the time/space resolution of slip? (3) Which regularizations work best for given slip models? In general, we hope to determine the limits to what can realistically be resolved in finite fault models from seismic data alone, even with good data coverage and perfectly known Green's functions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.S21B2454F
- Keywords:
-
- 7209 SEISMOLOGY / Earthquake dynamics;
- 7215 SEISMOLOGY / Earthquake source observations