Mainland China, The Net Surface Solar Radiation, Satellite Remote Sensing Research

Recently, with the improvement of sensor's precision and the development of remote sensing technology, the research to derive the global radiation budget from satellite data has a great progress (e.g. the CERES/EOS of USA). The net surface solar radiation (NSSR) as one of the most important portion of global radiation budget has a great action to solar energy absorbed by surface and then affect the surface radiation budget and atmospheric dynamical process. The global distribution of NSSR can be provided by means of satellite remote sensing.In this paper a new parameterization derived by Li Zhan-qing(Li model) in 1995 was used to compute the net surface solar radiation for 18 stations of Chinese main land on some days over January and February in 2001. After compared the results of Li model with those derived from A model (Li-Leighton model) and B model (Langley parameter model) for clear-sky and cloud-sky respectively, we founded that the relative difference of the two model are below 6%. From a large of computation we can concluded that Li-model can instead of A model on clear-sky or B model on cloud-sky condition, because A model can be used only on clear-sky condition and B model used on cloud-sky condition.Atmospheric precipitable water vapor is one of the most important parameter to Li model. In order to verify the reliability of MODIS atmospheric precipitable water vapor, we used the NCEP atmospheric water vapor data. The comparison showed that the difference between two data is less ±0.5cm in Chinese main land area, and the same time we found that MODIS atmospheric precipitable water vapor is greater than the NCEP data in the area of northwest and north of China and less it in the area of southeast of China. If we multiply MODIS atmospheric precipitable water vapor data with 0.933983 and 1.07686 respectively on above area the difference will below to ±0.2cm . The MODIS atmospheric precipitable water vapor is precise enough as the input parameter toLi-model.