Generation of digital elevation model in remote locations using satellite SAR interferometry and the model evaluation

为了利用卫星合成孔径雷达（SAR）干涉测量技术生成适用于地面观测困难的偏远地区的数字高程模型（DEM），并对其进行独立评估，采用卫星SAR三次通过法生成了日本越冬站所在地Lutzow-Holm湾附近冰盖的DEM。与最先进的全球数字高程模型的比较表明，其精度提高了一个数量级，使用地面观测，包括全球定位系统的结果作为参考。另一方面，也发现了一些问题的产生过程中，为了准确地获得接地线，形成的边界之间的接地冰盖和浮冰架，合成孔径雷达干涉测量应用于博杜安冰架以西的Lutzow-Holm湾。根据条纹的空间差异，可以清晰地识别和跟踪接地线。为建立和评价快速运动物体的地面观测方法，对固定海冰上的点和基岩上的点进行了GPS观测。同时观测证实，GPS可以检测海冰的垂直位移的精度为1-2厘米。如上所述，这项研究建立了SAR干涉测量的DEM生成远程位置的可用性。与有效的GPS地面观测相结合，将能够对整个冰盖区进行更精确的监测。

The primary goal of this study is to generate a digital elevation model (DEM) using satellite SAR interferometry for remote regions where ground observations are difficult, and is to evaluate them with independent methods.A DEM of the ice sheets around Lutzow-Holm Bay, where the Japanese wintering station is located, was generated with the satellite SAR three-pass method. Comparisons with the state-of-the-art global DEM showed improvement in its accuracy up to an order, using ground observations including GPS results as references. On the other hand, a few problems in the generation process were clearly identified.To derive a groundling line precisely, which forms the boundary between grounded ice sheet and floating ice shelf, the SAR interferometry was applied to Baudouin Ice Shelf to the west of Lutzow-Holm Bay. The grounding line was clearly identified and tracked from the spatial difference in fringes. The derived vertical displacement was consistent with the one which was inferred from a ocean tide model with inverse barometer effect.To establish and evaluate the methods of ground observations for the rapidly moving objects, GPS observations were conducted between the points on fast sea ice and on ground rocks. Simultaneous observations confirmed that the GPS can detect the vertical displacement of sea ice in an accuracy of 1-2 cm.As above, this study established the availability of the SAR interferometry to DEM generation for remote locations. Combined application with the effective GPS ground observations will enable much precise monitoring of the whole ice-covered area.