Acta Geodaetica et Cartographica Sinica ›› 2021, Vol. 50 ›› Issue (9): 1222-1239.doi: 10.11947/j.AGCS.2021.20200610

• Geodesy and Navigation • Previous Articles     Next Articles

Estimation of 3D coseismic deformation with InSAR: an improved SM-VCE method by window optimization

LIU Jihong, HU Jun, LI Zhiwei, ZHU Jianjun   

  1. School of Geosciences and Info-Physics, Central South University, Changsha 410083, China
  • Received:2020-12-22 Revised:2021-06-07 Published:2021-10-09
  • Supported by:
    The National Natural Science Foundation of China (No. 42030112); The National Key Research and Development Program of China (No. 2018YFC1505103); The Hunan Natural Science Foundation (No. 2020JJ2043); The Special Funds for the Construction of Hunan Innovative Province (No. 2019GK5006); The Hunan Provincial Innovation Foundation For Postgraduate (No. CX20190067); The Fundamental Research Funds for the Central Universities of Central South University (Nos. 2018zzts684; 2019zzts011)

Abstract: The 3D coseismic deformations derived from the interferometric synthetic aperture radar (InSAR) technique is of great significance for interpreting the characteristic of coseismic movement and inversing the fault slip. Recently, a method for estimating 3D surface deformations with InSAR based on the strain model and variance component estimation (SM, VCE, SM-VCE) is proposed, in which a lot of observation functions are established within a fixed-size window based on the SM, making it possible that the VCE is used to determine the weighting factor of different kinds of InSAR measurements. Compared with the pixel-by-pixel weighted least squares method, this window-based SM-VCE method can obtain a more reliable 3D deformation field. In the original SM-VCE method, the window size is designed to be constant for all points and the observations within the window are considered to be equal-accuracy. These rules are simple, but is easy to be violated for estimating the 3D coseismic deformation. Particularly, the case is easy to occur that the InSAR technique cannot obtain valid measurements in the near fault region, resulting in fact that the original SM-VCE method with a fixed-size window would fail to derive the near-field 3D coseismic deformation. Even if the near-field observations are available, the window-based SM-VCE method cannot estimate accurate 3D deformation due to the incorporation of inhomogeneous points across the fault. Besides, the accuracy of InSAR measurements within the window is generally various due to the decorrelation noise, which is not considered in the original SM-VCE method. In this paper, the surrounding points are selected based on a window with adaptive size as well as the fault lines so that the 3D deformation in the near fault zone can be obtained. Furthermore, an iterative weighted least squares method is employed to determine the relative weight of InSAR measurements within the window before the implementation of VCE. Simulation and real experiments of the 2019 Mw7.1 Ridgecrest earthquake are conducted with the Sentinel-1 data, demonstrating that the improved SM-VCE method by window optimization in this paper can obtain a more accurate and complete 3D deformation field compared with the original SM-VCE method.

Key words: InSAR, 3D deformation, SM-VCE, window optimization, IWLS

CLC Number: