A performance evaluation of a single-station earthquake location algorithm (LESSLA) implemented at WDD Seismic station, University of Malta

Abstract

The standard method of analyzing earthquake locations uses large seismic networks consisting of many seismic stations. As more seismic stations are added the distribution of the networks are improved, resulting in better solutions. However there is poor network coverage around the Sicily Channel because established international networks lack seismic stations in North Africa. Therefore, an automated single-station earthquake location algorithm (later called LESSLA) was designed by Agius (2007) to locate local and regional earthquakes to address this problem. It was required to investigate the performance of LESSLA with regard to several parameters. In this dissertation, the database for 2006-2008 has been used. During this period, LESSLA detected and located 135 earthquakes, mostly south of Malta, that were not found in international bulletins. For the whole dataset, an analysis of the automated solution was carried out, particularly the weighting scheme used to classify events, and the P and S-picking algorithm. The results indicate that the weighted schemes and phase picking times functioned very well. Less than 10% of earthquakes (40 from 462) were missed and adjustments made to the P- and S-pick times were usually less than a second from the automated pick. A comparison of the manually reviewed LESSLA epicentre solutions was made with published locations from EMSC/INGV bulletins when the latter was available. Both the origin time and final epicentre location of each earthquake were compared. The error in the origin time was found to be normally distributed about 0.358s with a standard deviation of 2.45s. The results of the final epicentre location was also very good with 4 7% of LESSLA epicentres being within 100 km of the EMSC/INGV locations. By contrast, only 19% of LESSLA epicentres were more than 300km away from the published international locations. The primary source of error in the location was because of the instability of the azimuth. There were less errors in the distance calculation, although recalibration of the formula used was required for long-range earthquakes. Using seismic data from WDD seismic station, several features of the Central Mediterranean can be determined. Although Italy is one of the most seismically active regions in the Mediterranean, very few earthquakes have been detected by WDD. The seismic signals are being attenuated before they reach Malta, probably due to subduction zones. In contrast, there is no attenuation of seismic signals from earthquakes originating in Greece. This means that LESSLA has a longer detectable range to the east from Greece than to the north from Italy.