利用Envisat/ASAR雷达影像处理后获取的2008年11月10日青海大柴旦Mw6.3级地震InSAR同震形变场数据和LN算法反演观测值含有粗差情况下的地震Okada矩形位错模型断层参数;断层深度为19.691 5 km,倾角为56.892 3°,长度为16.499 3km,宽度为6.854 8km,倾滑量为0.832 6 m,中心经度为95.884 7°,中心纬度为37.529 5°,走向方位角为116.411 5°,地震矩为3.108 4×1018 Nm(Mw6.295 0)。反演结果表明,当观测值含有粗差时,LN算法在1范数下可以有效抵抗粗差的干扰,充分利用已有先验信息的同时顾及观测、线性化等原因引起的系数矩阵误差的影响。
While gross errors are contained in the observations,through the use of InSAR coseismic deformation data obtained from the ESA Envisat/ASAR images and with LN algorithm,the fault parameters of Okada rectangle dislocation model in the elastic half-space can be calculated for Qinghai Dacaidan Mw6.3 earthquake on 10 November,2008.The earthquake fault inversion results of depth,dip,length,width,dip-slip,center longitude,center latitude,and strike azimuth are 19.691 5 km,56.892 3°,16.499 3 km,6.854 8 km,0.832 6 m,95.884 7°,37.529 5°,and 116.411 5° respectively.And the inverted geodetic seismic moment is 3.108 4×1018Nm(Mw6.295 0),which is consisting with the seismic results.The inversion results has proved that when the observations are contaminated by gross errors,using LN algorithm to deal with structured total least norm problems in L1 norm can effectively drop off the interference caused by those errors,and can fully make use of the prior information and at the same time consider the effects of coefficient matrix errors caused by reasons such as surveying,linearization and so on.