Comparative Analysis of Catalog based Magnitude of Completeness Estimation methods

Abstract

Gutenberg-Richter (GR) law is among one of the three empirically established laws of earthquake occurrence which describes the power law distribution of earthquake magnitude. However, the inherent inability of contemporary seismic network to detect earthquakes of all sizes leads to a divergence of observed frequency-magnitude distribution from GR law at lower magnitudes. Therefore, Magnitude of completeness (Mc) defined as the magnitude above which all the earthquakes are detected, has a direct impact on evaluation of G-R law parameters which are the basis of seismic hazard studies and earthquake forecast models. In this study, we present a rigorous comparison and ranking of various Mc estimation methods, such as MAXC, MBS, MBASS, EMR, GFT and Clauset method. Our findings will contribute to identification of optimal methods for specific data and can be used to improve estimates of Mc and thus accuracy of earthquake forecasts in the region. We performed completeness analysis on several constrained and unconstrained earthquake catalogs and observed a significant decrease in Mc on removing the events located far away from country border. Furthermore, we observed a systematic decrease in Mc around 1970’s indicating the rapid improvement in sensitivity and coverage of seismic networks during this period. From Mc analysis of original and constrained catalogs for all catalogs available to us, we conclude that the Clauset method yield best estimates of Mc followed by GFT, which is followed by EMR, MBS, MBASS, MAXC method each of which gives a similar estimates for Mc.