Spatial-temporal variations of b-values prior to medium to large earthquakes in Taiwan and the feasibility of real-time precursor monitoring

Abstract

Abstract This study investigates the temporal and spatial variations of the b-values before earthquakes greater than magnitude 6.0 in the Taiwan region and assesses the feasibility of monitoring earthquake precursors through b-values. Using the 2018 Hualien earthquake as an example, the b-value in the vicinity of the region one year before the earthquake was in the 12th percentile of all Taiwan, indicating a relatively low spatial value. Moreover, the b-value near the epicenter noticeably decreased within a year before the earthquake, showing a temporally relative low value. Based on this experience, we examine the temporal and spatial variations of b-values before earthquakes greater than magnitude 6.0 in Taiwan from 1999 to 2021. It was found that the epicenters of many large earthquakes were in regions with relatively low b-values across Taiwan a year prior. However, the b-values around various epicenters did not show significant changes at different times before the earthquakes. To further evaluate the feasibility of real-time monitoring of earthquake precursors using b-values, we analyzed the 2022 Yuli earthquake and the Chihshang earthquake sequence as case studies. Immediate post-event analysis using the rapid earthquake report catalog showed significant anomalies in both time and space for b-values, even though the real-time catalog obtains a larger magnitude of completeness. Specifically, the b-value in the vicinity of the region one year before the earthquake was in the 3rd percentile for the entire Taiwan region; furthermore, the b-value consistently decreased within the four years preceding the earthquake. These results validate the feasibility of evaluating earthquake precursors by analyzing real-time earthquake data. To more precisely predict future earthquake activity characteristics, it is suggested to incorporate other reference factors for a comprehensive evaluation, aiming to reduce cognitive uncertainty.