TESTING AN ALGORITHM TO RAPIDLY DETERMINE EARTHQUAKE FAULT PARAMETERS FROM AFTERSHOCK LOCATIONS

Numerous methods have been developed to determine the size and rupture extent of large earthquakes. Most, such as full seismic waveform or geodetic inversions and field mapping, are relatively time consuming and therefore not suitable to estimate the impact of the event within the first hour. However, aftershock locations may be helpful in a faster analysis of earthquake rupture. The spatial distribution of aftershocks for an earthquake sequence has a close empirical relation to its mainshock rupture extent. As such, the epicentral locations of aftershocks can be used to delineate the entire mainshock rupture zone and to determine first – order mainshock rupture parameters such as earthquake rupture length and orientation. An important goal of this research project is to determine how early in an aftershock sequence we can reliably and accurately determine these mainshock rupture parameters. To this end we test an algorithm we developed for rapid estimation of mainshock rupture parameters in a near real-time environment. This algorithm will serve to quickly assess the impact (damage and fatalities) of large global earthquakes when incorporated in a global earthquake analysis system, such as implemented by the US Geological Survey National Earthquake Information Center. Required input parameters to this algorithm are limited to an earthquake catalog starting with the mainshock time and other fixed input parameters. The algorithm automatically removes outlier events such as background seismicity and triggered events by spatial binning. Then true aftershock events are projected onto straight lines centered on the mainshock epicenter with a fixed incremental azimuth. Subsequently, a best fitting strike and rupture length for the mainshock are determined on the basis of this aftershock distribution. The results of the algorithm agree well with the rupture parameters for earthquakes from the Next Generation Attenuation database, as determined by other reliable but time consuming methods, for strike slip and large dip-slip events. We will also show results for recent earthquakes that caused more than one thousand fatalities. Our preliminary results indicate reliable estimates of fault rupture parameters may be obtained as early as one hour or less after the mainshock occurrence.