The Research On The Estimation Method Of Long-wave Effective Radiation Over Land Surface In China Based On Remotely Sensed Data

Surface long-wave effective radiation is the difference of Long-wave downward radiation and Long-wave upward radiation. It is an important form of the energy exchange of earth-atmosphere system. Currently, the most difficulty for relevant researches is Long-wave radiation is not logged by conventional meteorological stations. Therefore, empirical equation is still a mainstream choice. This kind of method is easy to use and can be applied extensively. But certain error will be caused due to the unevenly spatial distribution and limited quantity of conventional meteorological stations. With the development of Satellite Remote Sensing Technology in recent years, a lot of scholars have started to study how to use remotely sensed data to estimate surface long-wave effective radiation directly. Since the retrieve process of satellite has not been understood completely and the uncertain retrieve precision of satellite product, certain error will be caused.Therefore, this paper established a new method to estimate surface long-wave effective radiation. This method combined empirical equation method with directly retrieve method by using remotely sensed data by introducing remotely sensed digital product to calculate several parameters that will be used in the empirical equation based on empirical equation method. This paper firstly corrected remotely sensed digital product by using the data logged by conventional meteorological stations, which is seen as the true value, in order to improve the precision of Satellite digital product. Then, this paper established certain estimation models by remote sensing for certain parameters ready to be used in empirical equation and estimated these parameters based on different kinds of remote sensing satellite digital product. Finally, the spatialization simulation of long-wave effective radiation over land surface in china was accomplished.In this paper, the following tasks have been accomplished:1. Since there is a good positive linear correlation between surface vapor pressure and MODIS near-infrared precipitable water vapor product, this paper accomplished the estimation model by remote sensing for surface vapor pressure based on MOD05. And the spatial distribution dataset of yearly and monthly surface vapor pressure in2001was accomplished. After the validation by using data logged by intensive automatic weather stations, the yearly relative error of model is2.4%;2. Since there is a good negative linear correlation between sunshine duration of meteorological stations and MODIS cloud cover product, this paper accomplished the estimation model by remote sensing for sunshine duration based on MOD06. And the spatial distribution dataset of yearly and monthly sunshine duration in2001was accomplished. After the validation by using data logged by intensive automatic weather stations, the yearly relative error of model is9.45%;3. This paper also accomplished the estimation model by remote sensing for surface emissivity based on MODIS NDVI product. And the spatial distribution dataset of yearly and monthly surface emissivity in2001was accomplished. The estimation model has a strong operability and can better express the difference of land covers, such as bare soil pixel and vegetation pixel. Meanwhile, the simulation of yearly and monthly mean temperature in2001was accomplished by using the distributed model of monthly mean temperature. The error of model is within1centigrade;4. Finally, the spatial simulation of yearly and monthly surface long-wave effective radiation in2001was accomplished. The regular pattern of spatial distribution of estimation result and the characteristics reflected by certain areas in estimation result are akin to many researchers’results. After the comparison between NCEP reanalysis data and estimation result, we find the macroscopical trend reflected by the former and the latter are the same; but estimation result can better express regional details than NCEP reanalysis data.