基于地震对应概率谱分析的前兆异常识别研究
|
摘要
针对观测或分析资料曲线异常的定量化识别提取问题,研究提出了地震对应概率谱统计方法和累计滑动平均概率分析方法。通过地震对应谱分析可以得到不同值域数据异常信度属性。在不同考察时长的对应概率谱基础上,将原始数据时间序列转换成概率时间序列,并采用多点累计滑动平均方法得到滑动平均概率时间进程曲线,进而进行地震前兆异常的识别研究。以新疆北天山地区地震学参数η值时间进程序列为原始数据进行算例分析,结果显示,当考察时长为18个月时,利用滑动平均概率时间进程曲线可以较好地识别地震前兆异常,异常对应率为83%,有震报准率为86%,异常具有中期属性。分析认为,不同考察时长的地震对应概率谱和滑动平均概率序列,不仅可以用于单项资料的前兆异常识别研究,还可以为综合分析预报提供定量的单项因子数据。
With the aim to the quantification of anomaly identification and extraction for observed or analyzed records,we present earthquake corresponding probability spectrum statistical method and cumulative sliding mean probability analysis method.By earthquake probability spectrum analysis,we can obtain the abnormal confidence attribute of data in different value ranges.Based on the probability spectrum in different studied time-intervals,we convert the original data time sequence into probability time sequence,and can obtain the time curve for sliding mean probability by using multi-point cumulative sliding mean method,then identify seismic precursor anomaly.Taking the time sequence of η-value in the North Tianshan region as original data,we test the methods.The result shows that when the studied time-interval is 18 months,the precursor anomaly can be identified better from sliding mean probability time curve.The anomaly corresponding rate is 83 per cent,earthquake corresponding rate 86 per cent,and the anomaly is in middle term.The analysis indicates that the earthquake corresponding probability spectrum and sliding mean probability sequence in different time-intervals may not only be used to identify the precursor anomaly of single-item data,but also offer the data of quantitative single-item anomdy for comprehensive earthquake analysis and prediction.
With the aim to the quantification of anomaly identification and extraction for observed or analyzed records,we present earthquake corresponding probability spectrum statistical method and cumulative sliding mean probability analysis method.By earthquake probability spectrum analysis,we can obtain the abnormal confidence attribute of data in different value ranges.Based on the probability spectrum in different studied time-intervals,we convert the original data time sequence into probability time sequence,and can obtain the time curve for sliding mean probability by using multi-point cumulative sliding mean method,then identify seismic precursor anomaly.Taking the time sequence of η-value in the North Tianshan region as original data,we test the methods.The result shows that when the studied time-interval is 18 months,the precursor anomaly can be identified better from sliding mean probability time curve.The anomaly corresponding rate is 83 per cent,earthquake corresponding rate 86 per cent,and the anomaly is in middle term.The analysis indicates that the earthquake corresponding probability spectrum and sliding mean probability sequence in different time-intervals may not only be used to identify the precursor anomaly of single-item data,but also offer the data of quantitative single-item anomdy for comprehensive earthquake analysis and prediction.
引文
薄万举,王彦·1993·单项资料异常变化信息提取的一种新方法[J].大地形变测量,9(3):17-20·
薄万举,吴翼麟·1993·异常信息流的标准化方法及其应用[J].地壳形变与地震,13(2):9-15·
杜学彬,阮爱国,范世宏,等·2001·强震近震中区地电阻率变化速率的各向异性[J].地震学报,23(3):289-297·
王海涛,朱令人,杨马陵,等·1991·地震前兆观测数据的信息熵分析[J].地震,11(5):13-18·
郑熙铭,冯?益·1993·前兆值异常的时空演化与地震活动[J].地壳形变与地震,13(2):80-87·
薄万举,吴翼麟·1993·异常信息流的标准化方法及其应用[J].地壳形变与地震,13(2):9-15·
杜学彬,阮爱国,范世宏,等·2001·强震近震中区地电阻率变化速率的各向异性[J].地震学报,23(3):289-297·
王海涛,朱令人,杨马陵,等·1991·地震前兆观测数据的信息熵分析[J].地震,11(5):13-18·
郑熙铭,冯?益·1993·前兆值异常的时空演化与地震活动[J].地壳形变与地震,13(2):80-87·
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心