多方法组合预测模型在CSEP计划中的应用和预测效能检验

(1.中国地震台网中心,北京 100045; 2.中国地震局地震预测研究所,北京 100036)

多方法组合预测模型; CSEP计划; 加卸载响应比; R值评分方法; 预测效能检验

Application of Multi-Method Earthquake Prediction in CSEP and Testing of Prediction Efficiency
YU Chen1,HAN Ying1,ZHANG Yongxian2,ZHANG Xiaotao1,YU Huaizhong1

(1.China Earthquake Networks Center,Beijing 100045,China)(2.Institute of Earthquake Forecasting,China Earthquake Administration,Beijing 100036,China)

Multi-Method Earthquake Prediction(MMEP); CSEP; Load/Unload Response Ratio; R-value test; prediction test

DOI: 10.20015/j.cnki.ISSN1000-0666.2025.0021

备注

为了更好地评估CSEP中国检验中心的实用性,对基于多方法组合预测模型MMEP的2018—2023年全国5级以上地震的年度预测结果进行了系统梳理,并采用CSEP中国检验中心的R值评分模块对年度预测结果进行了评估。结果表明:①近几年MMEP年度预测结果的R值评分均值在0.3~0.4,年度预测效能波动较小; ②中国大陆西部地区的预测结果优于东部地区,其中2019年四川长宁6.0级、2020年新疆伽师6.4级和于田6.4级地震的年度预测以及四川泸定6.8级地震的中短期预测的“时、空、强”三要素基本正确。
In recent years,dependent on the national key research and development program,the Collaboratory for the Study of Earthquake Predictability(CSEP)initiated by the Southern California Earthquake Center has been transplanted to China.The CSEP-China center has been established,and prediction modules such as Pattern Informatics(PI),Load/Unload Response Ratio(LURR),and State Vector(SV)have been developed.In order to better evaluate the practicability of CSEP-China center,the annual prediction results of the Multi-Method Earthquake Prediction(MMEP)based on the above three prediction methods are summarized.Furthermore,the R-value scoring module of CSEP-China center is used to evaluate the annual prediction results.The results show that:(1)The R-value score of MMEP annual prediction results is in the range of 0.3-0.4,and the annual prediction efficiency has little fluctuation;(2)The prediction results in the western region of Chinese mainland are better than those in the eastern region.Among them,the annual prediction of the Changning 6.0 earthquake in Sichuan,the Jiashi 6.4 earthquake and the Yutian 6.4 earthquake in Xinjiang,and the medium-short term prediction of the Luding 6.8 earthquake in Sichuan are correct.(3)For different regions,how to give the appropriate scale of time prediction results needs to be improved.