Spatiotemporal variation characteristics for b-value of induced seismicity in Southeast of Sichuan Basin, China

Abstract:

To investigate the mechanism beneath the induced seismicity and estimate seismic hazards, it’s necessary to obtain enough information from both experiments and observations. Among kinds of information, the statistical analysis could tell us some useful clues. In the Southeast of Sichuan Basin, China (SBC), the most local seismicity with magnitude up to M_S6.0 since 2015 are induced by or related to hydraulic fracturing operation (Meng et al, 2019; Lei et al, 2019), which includes the 28 January 2017 M_S4.9, the 4 May 2017 M_S4.9, the 16 December 2018 M_S5.7, the 3 January 2019 M_S5.3, the 25 February 2019 M_S4.9, the 17 June 2019 M_S6.0, the 8 September 2019 M_S5.4 and the 18 December 2019 M_S5.2 earthquakes. Thus, by using the catalogs in the Southeast of SBC, we explore the spatiotemporal characteristics of b-value in G-R relation to gain insight into the possible dynamic processes beneath the induced seismicity and assess the seismic hazards. First, we divide the Southeast of SBC into three regions: Changning region (CNR), Junlian-Xingwen region (JXR) and Rongxian-Weiyuan region (RWR). For each region, we obtain the background b-value by using maximum likelihood method and 10-year catalogs before 2015. Then we calculate the spatiotemporal b-value of three regions with the same method. Our current results show that the occurrence of moderate events changes the spatiotemporal b-value in each region. In CNR, the average b-value decrease from 1.17 to 0.75 when M_S6.0 earthquake occurred. For JXR, the b-value shows a rapidly decreasing and increasing trend before and after the occurrence of moderate earthquakes. In RWR, the b-values before and after the 25 February 2019 M_S4.9, the M_S5.4 and the M_S5.2 events aren’t same, but the temporal variations for b-values don’t show a common pattern. Furthermore, current average b-value in CNR is smaller than the background b-value and the one in RWR and JXR is larger, which indicates the CNR is at risk. As for spatial b-value, an area with low b-value appears around the epicenters before M_S5.7 and M_S4.9 on 25 February 2019. In addition, the b-values of each region during active period are all close to 1, which indicates a fault reactivation process (Chen et al, 2018). In a word, the mechanism beneath the induced seismicity is complex, but during fault reactivation the induced events change the b-value.