中文摘要：

利用InSAR形变数据进行反演研究时,InSAR数据地面空间覆盖范围广,获取的数据点数量多、密度大,包含有较多的误差点,会制约反演效率,影响反演结果的精度.为了提高反演的计算效率,保证迭代收敛,对InSAR形变数据进行了降采样研究,以西藏自治区当雄县2008年Mw 6.3级地震的同震形变数据为例,研究了基于降低分辨率的3种降采样方法：最邻近法、双线性内插法和三次卷积内插法,对3种方法所得结果进行对比分析,结果表明三次卷积内插法最适合InSAR地表形变数据的降采样.为了进一步提高降采样成果质量,对三次卷积内插值法进行改进,改进后获取的降采样形变数据能有效减少采样点数,较好地保存形变场的细节信息.

英文摘要：

As InSAR data covers a wide range of surfaces, numerous data points, with a high density and many error points, the inversion efficiency will be restricted and the accuracy of the results will be affected when InSAR deformation data are used to carry out an inversion. In order to improve the effi- ciency of inversion and ensure the convergence of iterations, a down-sampling study on InSAR deforma- tion data was conducted, taking the co-seismic deformation data of the Mw 6.3 earthquake that occurred in Dangxiong County in the Tibet Autonomous Region in 2008 as an example. The three resolution-re- duced down-sampling methods （the nearest neighbor, the biliner interpolation and the cubic convolution interpolation method） were applied, with the results of these three methods compared and analyzed. The results showed that the cubic convolution interpolation method was the most suitable for down-sampling of InSAR surface deformation data; the cubic convolution interpolation method was improved so as to further improve the outcome quality of down-sampling. After having been improved, the down-sampling deformation data acquired can effectively reduce the number of sampling points, resulting in well-pre- served details of the deformation field.