Abstract: The synthetic probability gain model for earthquake prediction is first introduced into the prediction of reservoir induced earthquake.Its basic principle is described.The possible factors affecting the maximum magnitude of reservoir induced earthquake are approached.According to the basic reservoir data and the cases of reservoir induced earthquakes in China,the forecast efficiency R-value and the empirical probability gain K-value of partial factors or predicting indices for maximum magnitude forecast of reservoir induced earthquake,such as comprehensive effect parameter E-value,the reservoir capacity and the water depth of reservoir,are statistically calculated and appraised.The results demonstrate that the synthetic probability gain model can make a quantitative analysis of the efficiency of the various predicting indices.Then some of reservoirs at home and abroad are selected for retrospective forecast test of maximum magnitude of reservoir induced earthquake using the above indices.The results indicate that the probability gain method is practicable as a new method forecasting the maximum magnitude of reservoir induced earthquake.

Key words: reservoir induced earthquake, synthetic probability gain prediction, maximum magnitude, comprehensive effect parameter E-value

摘要： 在水库地震预测中首次引进了地震概率增益综合预测模型,叙述了该方法的基本原理;探讨了影响水库地震最大震级发生的因素;结合中国水库及其水库地震震例资料,对水库地震综合影响因素E值、库容、库水深度因素或指标预测水库地震最大震级的效能R值和经验概率增益K值进行了统计和评价,表明概率增益综合预测能够对各种预测方法预测水库地震的效果进行定量分析。在此基础上,选择国内外若干水库进行了水库地震最大震级回溯性检验,表明该方法作为一种新的水库地震最大震级预测方法是可行的。

关键词: 水库地震, 概率增益综合预测, 预测效能, 水库地震综合影响因素E值