基于InSAR和ICESat的南极冰盖地区DEM提取和精度分析

结合InSAR和ICESat测高数据,以东南极PANDA断面4个实验区为例,进行DEM生成研究.在干涉相位转换成高程前,引入ICESat测高数据作为控制点优化干涉对基线,消除基线线性误差趋势的影响.利用控制点之外的ICESat测高数据分析4个实验区的DEM精度及其差异,并探讨了引起DEM误差的原因.冰流和地貌特征是影响InSAR生成冰盖DEM精度的重要因素.针对冰流的影响,从理论上进行了分析并结合冰流速数据进行了分析验证.最后利用克里金插值法改正InSAR DEM残余误差,并利用GPS实测控制点对改正效果进行验证.结果表明:对于高纬区流速较小且分布一致的区域,改正效果很好,DEM精度可达3 m;而对于冰流速较大且复杂的地区,需采用多基线等算法进一步消弱冰流引入的误差.

DEM generation and precision analysis of Antarctic ice sheet based on InSAR and ICESat data

InSAR and ICESat data were combined to generate DEMs of four test sites along PANDA section. In order to eliminate the residual linear trend error, ICESat data was introduced to refine the baselines of the interferometric pairs as control points. Additional ICESat footprints were used to evaluate the accuracies of the DEMs. Accuracy differences of the four DEMs were analyzed and the possible reasons causing DEM errors were discussed. Topographical features and ice streams were the main factors influencing the accuracy of DEMs derived from InSAR on Antarctic Ice Sheet. The effect induced by ice streams was analyzed theoretically. Then analysis and validation were performed considering the impact of ice velocity. Furthermore, Kriging interpolation method was applied to correct the InSAR DEMs and GPS points were used to validate the result. The result showed that corrections performed very well in high latitude areas where ice velocity was small and of consistent distribution, and the accuracy of DEMs can be up to 3-m in these areas. New method like multi-baseline InSAR should be employed to improve the DEMs in areas with large and complex ice velocity.