基于MODIS数据的黄河源区近地表气温遥感反演

Retrieving of near land surface temperature in the source region of Yellow River with remote sensing

为探讨应用MODIS数据反演近地表气温的可行性,建立遥感反演实验区近地表温度方程,对比不同下垫面对反演结果的影响,利用2007年全年的MODIS地表温度产品和HOBO气象站实测的气温资料,对黄河源区4种典型草地近地表温度进行研究。结果表明,直接应用MODIS-LST反演近地表气温是可行的,分白天和夜间分别建立模拟方程并对方程反演结果进行验证发现,白天和夜间反演平均误差分别为3.47K和2.12K,反演精度都较高,但白天反演精度不如夜间。不同草地类型对反演结果没有明显影响。

Two equations were established during daytime and nighttime by using the MODIS data in the study area for describing the feasibility of getting the LST （land surface temperature）. The MODIS LST product and ground based temperature data from HOBO climate stations were collected to analyze the LST for four different grassland in Yellow river watershed. This study showed that it was feasibility to use the MODIS LST product to acquire the land surface temperature. The retrieval equations of daytime and night- time were also validated, and the mean error in daytime and nighttime are 3.47 K and 2.12 K, respectively. The accuracy of the equation in nighttime was higher, but the effect of different grasslands was incon- spicuous.