形变观测中小数据集PSInSAR的误差分析与初步应用  

Error Analysis and Preliminary Application of PS InSAR in Deformation Detection on Small Stacks

作　　者：唐攀攀[1,2,3] 单新建[2] 王长林[4] 张桂芳[2] 

机构地区：[1]中国科学院遥感应用研究所,北京100010 [2]中国地震局地质研究所地震动力学国家重点实验室,北京100029 [3]中国科学院研究生院,北京100049 [4]中国科学院对地观测与数字地球中心,北京100094

出　　处：《地震》2011年第3期92-102,共11页Earthquake

基　　金：国家科技支撑项目(2008BAC38B03);国家自然科学基金(40874006);国家行业专项(200708013);中国地震局地质研究所基本科研经费(DF-IGCEA-0607-1-19)共同资助

摘　　要：PS InSAR技术在微小形变的监测中有着独特的优势,但同时受到数据量的严重制约。本文以形变速率稳定的怀来县城为实验区,利用15景ASAR数据,依据主影像选择和高程改正选取了三种略有差异的模型分别进行PS处理,并对形变提取中的相位解缠、基线改正和高程改正进行误差分析。结果表明,小数据集的相位解缠和高程改正在部分PS点上容易发生错误,但形变估计整体上能够正确反映实际情况。三个模型中MS(Multi-Reference Single Re-gression)模型得出结果最为可靠。小数据集解算的整体PS点质量不高,可以在满足PS点密度要求的前提下,降低相位标准偏差阈值以提高解算的可靠性。PS InSAR technique has a special advantage is in monitoring small crustal displacements,but its application is severely restricted by the number of images.This paper selects a town named Huailai which has a stable deformation rate as the test area.We choose three models based on the differences of reference image and height corrections to process 15 ASAR images with PS technique.Additionally,we analyse the errors of unwrapped phases,corrections of baselines and point heights.The result shows that for small stacks,the phases are unwrapped falsely on some points as well as the height corrections,but the deformation estimation is right in the whole.Among the three models,the result of Model Two（Multi-Reference Single Regression）is the most reliable.The quality of points is very low on the whole for small stacks,and the reliability of solution can be increased by lowering the phase standard deviation threshold.

关 键 词：PS INSAR 主影像 高程改正 相位解缠 小数据集 

分 类 号：P315[天文地球—地震学]