Research on the forecasting strategy of early aftershocks in North China

Abstract

To systematically investigate short-term aftershock forecasts following moderate-to-strong earthquakes of North China so as to develop "operational" aftershock forecasting strategies tailored to regional seismic characteristics, we adopt the widely used Reseanberg-Jones (R-J) model and the Omi-R-J model to explore aftershock forecasting strategies of 24 earthquake sequences of North China, and use the N-test and T-test to evaluate the forecasting effectiveness. Early forecast results after mainshock show that the R-J model and the Omi-R-J model have an average effectiveness rate of 77.0% and 87.9% for the selected sequences, respectively. The R-J model has a lower ratio of forecasting "too low" numbers of earthquakes while the Omi-R-J has a rather low overall "failure rate". With the rapid development of monitoring network after 2008, the efficacy of earthquake sequence forecasting has improved significantly, with monitoring capability being an important factor constraining forecasting effectiveness. The possible scientific strategies for the aftershocks forecasting in North China include strengthening the construction of seismic networks and applying better algorithms for detecting or simulating aftershocks, giving priority to using the Omi-R-J model within a short period of time (within 2 hours) after the mainshock, and weigh the range of the actual number of future aftershocks by appling the R-J model and the Omi-R-J model simultaneously.