Radon Anomaly and Bayesian Probabilistic Modeling for Seismic Precursor Study in the Sichuan-Yunnan Region, China

: The variation of radon concentrations can effectively reflect the changes of crustal stress and strain. Therefore, radon concentrations monitoring is an effective means of earthquake prediction. However, how to use observed water radon or gas radon to predict earthquakes still need to be researched. In this paper, 20 observation stations of water radon and gas radon in the Sichuan-Yunnan region has been studied. Firstly, the threshold method has been used to anomaly detection after the observation data were denoise by the residual method, and R -value test was used to determine the best threshold line and the sliding window. Then, the method of Hurst index has been used to test the randomness of the observed radon data. Finally, based on the background earthquake recurrence probability, the Bayesian theorem was used to calculate the earthquake risk probability of radon anomalies. The results showed that it has better prediction results in Yunnan and Sichuan-Yunnan border areas, due to the relatively dense distribution of observation stations in these regions. While, other regions have relatively poor prediction results, due to fewer or uneven distribution of stations. In addition, the study has shown that the precursor anomaly patterns of radon are closely related to the state of the regional stress field. Both the occurrence of high-value anomalies and low-value turning anomaly may have certain short-term imminent prediction significance. The research results indicated that the method of probability prediction of earthquake risk based on radon anomalies was feasibility, and this study also provided a way of exploring earthquake prediction methods.