Can we obtain reliable seismic b -values for real-time catalogs?

Linxuan Li,Gang Luo

DOI: https://doi.org/10.1093/gji/ggae124

IF: 3.352

2024-03-30

Geophysical Journal International

Abstract:Summary The seismic b-value in Gutenberg–Richter law is an important parameter in earthquake science research and earthquake risk assessment. People have tried to use b-values for short-term earthquake forecasts, and this requires the premise of estimating reliable b-values for real-time seismic catalogs. However, estimating b-values for real-time catalogs, which are usually of poor qualities, is usually faced with many difficulties and problems. In this study, through a series of numerical tests, we investigate the performance of three methods, including the commonly used maximum likelihood estimation method and two relatively new b-value estimation methods, namely the b-positive and K − M slope methods, on calculating b-values for real-time seismic catalogs. We also apply these three methods to both observed seismic catalogs (the seismic sequence in the Costa Marchigiana, Italy, and a high-resolution early aftershock sequence of the 2023 two Mw ∼7.8 earthquakes in Türkiye) and synthetic real-time seismic catalogs. The results in this study show that it seems difficult to obtain accurate b-values for real-time earthquake catalogs by each of these three methods, but the combination of these methods may give a better judgment—if all three methods suggest that the change in b-value is significant, the probability of making a correct decision is very high. Facing the uncertainty of b-value estimation that always exists, we advocate exploring the effectiveness of standard b-value estimation strategies in practical applications.

geochemistry & geophysics