Daubechies多尺度小波分析技术及其在漠河盆地断裂带划分中的应用  

Daubechies multi-scale wavelet analysis method and the application in fault zone partition of Mohe basin

作　　者：许家姝[1] 刘立家[1] 刘银萍[1] 

机构地区：[1]吉林大学地球探测科学与技术学院,长春130026

出　　处：《地球物理学进展》2014年第2期595-602,共8页Progress in Geophysics

基　　金：国家科技支撑计划项目<东部铁铜铅锌重要矿集区深部资源勘查技术与示范>(2009BAB43B00);吉林大学基本科研业务费项目--科学前沿与交叉学科创新项目(450060481099)联合资助

摘　　要：在位场数据处理及解释中,断裂的提取至关重要,因此,如何更有效提取边界信息,压制数据中的噪声成为解释人员重点研究目标.本文通过讨论用于识别重力断裂构造模型的导数方法与Daubechies小波分析方法在信号分析过程中的异同,证明了这两种方法在断裂识别中的有效性.提出将这两种断裂识别方法相结合进行断裂识别的新方法.通过对漠河盆地高精度重力剖面测线数据使用经典断裂反演方法和Daubechies多尺度小波分析相结合的方法划分出10条断裂构造,并在漠河盆地中依据平面重力异常利用上述方法识别出四组断裂构造,详细分析了由剖面数据及平面数据获得的各个断裂的构造特征.通过在漠河盆地高精度实测数据中的应用,多尺度小波分析与经典方法相结合的断裂构造识别方法可有效的避免单用经典方法时产生的多个离散点的影响及对断裂位置划分的干扰.该方法可明显去除一些不收敛的反演点位,使断裂图的清晰度得到大幅度提高.The extraction of the faults is one of the critical technique in the processing and interpretation of the potential field data. Accordingly, How to more effectively extract the boundary information, and suppress the noise in the data become key research objectives for the interpreters. The similarities and differences between the original inversion and wavelet transform method named Daubechies in signal analysis process which is used to detection the faults in gravity data are discussed in this paper in order to prove the identification of the method. Ten faults are detected by the use of the combination of the original inversion and wavelet transform in high accurate data from a profile in Mohe Basin. Four groups of faults are identified by the method described above with the use of the plane gravity anomalies in Mohe Basin, and the characteristics of the fault are analyzed in details. The method of the combination of the Multi-scale wavelet and original inversion to detect the fault can avoid the influence of the discrete points generated by the use of the original inversion individually which affects the ascertainment of the location of the faults. This method can extract the points which don＇t converge obviously. The definition of the fault map can enhance enormously.

关 键 词：多尺度小波分析 断裂识别 漠河盆地 

分 类 号：P312[天文地球—固体地球物理学] P541[天文地球—地球物理学]