基于星载激光雷达ICESat-2的雪深监测方法研究

目前，被动微波数据是积雪深度反演的主要数据源，受其较粗空间分辨率的影响，反演雪深存在较大的不确定性。激光雷达由于其较高的测高精度，在雪深监测方面具有一定的潜力，基于星载激光雷达ICESat-2数据对北疆地区2018年10月至2019年9月积雪季的雪深进行了提取。由于很难获取ICESat-2轨迹点的雪深观测资料，因此首先利用地面雪深观测数据对目前流行的被动微波雪深反演产品进行验证，获取精度可靠的雪深产品并与发展的ICESat-2监测雪深数据产品进行对比。结果表明：AMSR2雪深产品在北疆地区误差较大，整体存在高估现象，中国雪深长时间序列雪深产品精度相对可靠，以作为对比ICESat-2模拟雪深的参考数据；ICESat-2雪深与中国雪深长时间序列雪深产品在空间上以及变化趋势方面吻合度较高，但ICESat-2雪深变化更加连续，说明ICESat-2激光雷达数据不但可以提取区域积雪的深度，对积雪深度的空间变化也比被动微波数据更加敏感，可以获取更加详细的积雪深度空间变化细节，为精细化的积雪深度空间分布提供数据支撑。

Snow depth retrieval from spaceborne LiDAR ICESat-2

At present， passive microwave data is the main data source for snow depth retrieval， and there is large uncertainty in snow depth retrieval due to its coarser spatial resolution. Due to its high measurement accuracy， LiDAR has a specific potential in snow depth retrieval. This study tries to extract the snow depth of the snow season from October 2018 to September 2019 in the northern Xinjiang region based on the spaceborne LiDAR ICESat-2 data. Since it is difficult to obtain snow depth observation data at the ICESat-2 footprints， thus， this study first uses ground snow depth observations to verify the current popular passive microwave snow depth products and then obtains reliable snow depth products， which are compared with the ICESat-2 snow depth data produced in this study. The results show that the AMSR2 snow depth product has a large error and overall overestimation in northern Xinjiang. The long-term sequence of snow depth dataset in China （CHINA_SD） is relatively reliable， and which is used as a reference data for evaluating the simulated snow depth from ICESat-2； ICESat-2 snow depth is highly consistent with the snow depth products of CHINA_SD in terms of space and change trend， but ICESat-2 snow depth changes more continuously， indicating that ICESat-2 can not only extract the snow depth in regional， but also be more sensitive to the spatial change of snow depth than passive microwave data， and can obtain more details of snow depth spatial change， providing data support for refining the spatial distribution of snow depth.