1976年以来,四川理塘温泉水温观测积累了大量的观测资料及异常变化信息。通过对理塘温泉周边7级以上地震前理塘水温原始观测值出现的显著变化进行总结,发现理塘温泉水温在多次7级以上地震前均出现了大幅度破年变下降,下降变化幅度超过5℃,持续低值状态时间超过半年。同时,利用Molchan图表法计算理塘水温的原始观测值、一阶差分原始值及绝对值、去趋势差分值、滑动值,并对经验模式和Dobrovolsky地震能量模型所选取的2个地震目录分别进行了检验和分析,计算时间占有率、预测效能和概率增益等参数。结果显示,由于经验模式条目少于Dobrovolsky地震能量模型所筛选的条目,计算出的预报效能整体高于Dobrovolsky地震能量模型,而其中一阶差分原始值及绝对值检验结果较好,其差分绝对值在阈值为4.89℃时的预报效能为0.6,证明理塘水温短时间内温度大幅变化与周边地震存在对应关系,该阈值也较为接近经验预报的相关阈值。

Based on the temperature monitoring data of Litang hot spring since 1976,we have found continuous low value for the range of more than 5℃ and for the time of more than 6 months before M_S7.0 earthquakes. In this paper,we apply Molchan Error Diagram method to analyze the monitoring data with two kinds of earthquakes according to the traditional model and the Dobrovolsky model. The results show that the traditional model produces better results than the Dobrovolsky model because of the decrease of earthquakes. After we have calculated "time occupancy","forecast efficiency" and "probability gain" of the monitoring data,sliding mean value,difference value we found that the original value and absolute value of difference have a good significance of the traditional model earthquakes. The prediction efficiency of the absolute difference value is about 0.6,under the same threshold value with experience forecasting method for about 4.89℃.

P315

中国地震局地震科技星火计划攻关项目(XH19055)、国家自然科学基金(U2039202)、震情跟踪定向工作任务(2022010307)和中国地震台网中心青年基金(QNJJ-2021-07)共同资助