基于HJ-1B卫星数据的疏勒河上游流域地表温度反演

地表温度是描述地表和大气之间物质交换和能量平衡的关键参数之一. 利用疏勒河上游流域2010年7月25日过境的HJ-1B卫星数据, 采用修正的Jimenez-Mufloz&Sobrino普适性单通道算法反演了疏勒河上游流域的地表温度. 与MODIS LST产品的对比验证表明, 该算法在研究区有较高的精度. 同时, 对不同土地利用类型的地表温度的分析表明, 土地利用类型不同, 其地表温度存在一定的差异, 其中, 稀疏草地的地表温度最高, 裸岩次之. 研究发现, 在河流和冰川区, 修正的单通道算法可能高估了表面温度. 同时, 对模型反演所需的关键地表参数的敏感性分析发现, 算法在一定范围(0~0.01)内对比辐射率不太敏感, 但超过该范围则变得非常敏感, 而对大气水汽含量比较敏感. 因此, 一定程度上, 地表比辐射率和大气水汽含量的精度对HJ-1B卫星数据反演地表温度的准确性起主要作用.

Land surface temperature retrieved from HJ-1B satellite data in the upper reaches of the Shule River

Land surface temperature (LST) is one of the important parameters for describing material exchange and energy balance between land surface and atmosphere. Based on the HJ-1B satellite data in the upper reaches of the Shule River on July 25, 2010, LST of the upper reaches of the Shule River was derived with the modified Jimenez-Mufloz & Sobrino (JM&S) generalized single-channel algorithm. Compared with MODIS LST product, the algorithm has high accuracy. Analyzing the LST of different land use types indicates that LST of different land use types have some differences. The LST from HJ-1B of sparse grassland has the highest average LST, following by bare rock. It is found that the modified generalized single-channel method may overestimate LST on glaciers and rivers. In addition, sensitivity analysis has performed for major land surface parameters, and found that LST is not sensitive to land surface emissivity within a certain range (0-0.01); once beyond the range, it becomes more sensitive; and LST is more sensitive to water vapor content. Therefore, to some degree, the accuracy of land surface emissivity and water vapor content plays a remarkable role to the veracity of LST derived from HJ-1B satellite data.