Abstract：
Disaster mitigation necessitates scientific and accurate aftershock forecasting during the critical 2h after an earthquake. However, this action faces immense challenges due to the lack of early postearthquake data and the unreliability of forecasts. To obtain foundational data for sequence parameters of the land–sea adjacent zone and establish a reliable and operational aftershock forecasting framework, we combined the initial sequence parameters extracted from envelope functions and incorporated small-earthquake information into our model to construct a Bayesian algorithm for the early postearthquake stage. We performed parameter fitting and early postearthquake aftershock occurrence rate forecasting and effectiveness evaluation for 36 earthquake sequences with M ≥ 4.0 in the Bohai Rim region since 2010. According to the results, during the early stage after the mainshock, earthquake sequence parameters exhibited relatively drastic fluctuations with significant errors. The integration of prior information can mitigate the intensity of these changes and reduce errors. The initial and stable sequence parameters generally display advantageous distribution characteristics, with each parameter’s distribution being relatively concentrated and showing good symmetry and remarkable consistency. The sequence parameter p-values were relatively small, which indicates the comparatively slow attenuation of significant earthquake events in the Bohai Rim region. A certain positive correlation was observed between earthquake sequence parameters b and p. However, sequence parameters are unrelated to the mainshock magnitude, which implies that their statistical characteristics and trends are universal. The Bayesian algorithm revealed a good forecasting capability for aftershocks in the early postearthquake period (2 h) in the Bohai Rim region, with an overall forecasting efficacy rate of 76.39%. The proportion of “too low” failures exceeded that of “too high” failures, and the number of forecasting failures for the next three days was greater than that for the next day.
2025, Vol. 22 
|
