首页 | 本学科首页   官方微博 | 高级检索  
     

鄂尔多斯地块北缘b值的时空特征及其地震预测效能分析
引用本文:张帆,韩晓明,陈立峰,李娟,李拴虎,胡博. 鄂尔多斯地块北缘b值的时空特征及其地震预测效能分析[J]. 地震学报, 2018, 40(6): 785-796. DOI: 10.11939/jass.20180020
作者姓名:张帆  韩晓明  陈立峰  李娟  李拴虎  胡博
作者单位:中国呼和浩特 010051 内蒙古自治区地震局
基金项目:内蒙古自治区地震局局长基金(2018ZD02)、内蒙古自治区2016年度科技重大专项“重点地区地震预测预警技术研究开发与推广示范”和中国地震局监测预报司2017年度震情跟踪定向跟踪任务面上课题(2017020102)共同资助
摘    要:使用中国地震台网1970—2017年地震目录,在评估目录完整性和删除余震的基础上,使用最大似然法对鄂尔多斯地块北缘进行b值时空扫描. 扫描结果显示,鄂尔多斯地块北缘的低b值区域主要分布于阿拉善和晋冀蒙地区,低b值区域的空间分布随时间呈现一定变化. 低b值区域与中强地震存在一定的对应关系,对14次震例的回溯性检验显示,其中10次地震发生在低b值区域及其边缘. 使用Molchan图表法和R值评分法对b值时空扫描结果进行预测效能检验,以低于阈值的区域为预测区域,计算不同阈值的漏报率、报准率和时空占有率。Molchan图表法检验结果显示,本文所使用的b值时空扫描方法的预测效果优于泊松分布,最优结果的概率增益达到2,显著性水平达到0.25. R值曲线显示,当阈值为0.8时,达到最高R值0.28,并达到97.5%置信水平. 

关 键 词:b   时空扫描   预测效能   Molchan图表法   鄂尔多斯地块北缘
收稿时间:2018-02-04

Spatio-temporal characteristics of b value in the northern margin of Ordos block and its earthquake prediction efficiency
Affiliation:Earthquake Agency of Inner Mongolia Autonomous Region,Hohhot 010051,China
Abstract:Using earthquake catalogue of China Earthquake Networks Center from 1970 to 2017, the b values of the north edge of Ordos block were scanned by the maximum likelihood method after evaluating completeness and deleting aftershocks. The results show that the low b value region of the north margin of Ordos block is mainly distributed in the Alxa area and the Shanxi-Hebei-Inner Mongolia area, and the spatial distribution of low b value region changes with time to a certain extent. Meanwhile, there is a certain corresponding between the low b value region and the location of moderate-strong earthquakes. The retrospective test of 14 earthquake cases shows that ten earthquakes occurred in the low b value regions and their edges, and four earthquakes occurred beyond the low b value region. And then Molchan chart method and R value scoring method were used to test the prediction efficiency of b value spatio-temporal scanning. Taking the area below the threshold as the predicted area, the missing rate, the hit rate and the spatio-temporal occupation rate of different thresholds were calculated. The result of Molchan chart test shows that the prediction efficiency of spatio-temporal scanning is better than that of Poisson distribution, the probability gain of the optimal result reached 2, and the significance level reached 0.25. The R value curve shows that the highest R value is 0.28 at the threshold of 0.8 with 97.5% confidence level. 
Keywords:
点击此处可从《地震学报》浏览原始摘要信息
点击此处可从《地震学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号