中文摘要：

卫星时序差分雷达干涉已在区域地表形变监测中展现出极好的应用潜力,但因受卫星寿命及重访周期等因素的限制,单一卫星平台很难为长时间跨度（如5~10年）的时序形变计算与分析提供数据源。为扩展区域地表形变监测的时间跨度,本文提出了基于模型约束的短基线集（MC-SBAS）时序干涉方法,即使用非线性形变模型约束SBAS分析与处理,达到不同卫星平台数据融合和形变监测效率提高的目的。研究选取美国南加州地区为试验区域,使用2002—2005年获取的10幅ERS SAR影像和2004—2008年获取的20幅EnviSat ASAR影像,基于MC-SBAS提取该区域时序形变信息,时间跨度扩展为6.5年;以6个GPS站时序观测数据为参考,对MC-SBAS结果和常规SBAS结果进行了对比验证。分析表明,MC-SBAS、SBAS与GPS形变结果的均方根误差分别为8.7和11.7mm,MC-SBAS形变序列的连续性得到有效改善,而且形变监测精度得到提高。

英文摘要：

The satellite multi-temporal radar interferometry technology has been often widely used to monitor the regional ground deformation.However,due to the limitations of the lifetime and revisit cycle,a single satellite platform cannot provide sufficient data sources for estimating and analyzing the long-time deformation monitoring（e.g.spanning 5to 10years）.In order to extend the time span of deformation monitoring,this paper proposes the model-constrained small baseline subset（MC-SBAS）method by using the multi-platform SAR images.This method can integrate the SAR datasets from different satellites and improve the efficiency of deformation monitoring.The validation is performed by taking the Southern California as the study area and using 10 ERS SAR images（acquired between 2002 and 2005）and 20 EnviSat ASAR images（acquired between2004 and 2008）.In this validation,the spatiotemporal deformation information with the extending time span of6.5years in the study area is extracted by the MC-SBAS method,thus extending the time span of deformation tracking to 6.5years.The quality of deformation results derived from both MC-SBAS and the conventional SBAS method is assessed and compared by using the ground truth data obtained from the 6GPS tracking stations.The validation result shows that the root mean squared error（RMSE）between the MC-SBAS and GPS results is 8.7mm,which is less than the RMSE between the SBAS and GPS results（11.7mm）.In addition,the continuity and accuracy in the deformation time series derived from MC-SBAS are improved remarkably.