Study On The Cloud Processing With MODIS Data And The Retrieval Of LST

In this paper, the cloud processing and the retrieval of Land Surface Temperature with MODIS data are researched Due to the influence of atmosphere condition, the MODIS remote sensing data is disturbed by the cloud noise frequently, not only the further processing and recognition of image but also the precision of quantitative retrieve of the land surface parameter is affected. So, the cloud processing is one of the most important stages for improving the quality of the image before making analysis and applications. Firstly, the MODIS data is preprocessed by remote sensing image processing software ENVI. Based on the theory of remote sensing, the radiation characteristic of different underlying surface including clound, water, vegetation, and land on local area is obtained. Through the statistic characteristic and thresholds that are useful for the target area are found, the thin cloud area and the thick cloud area are detected. Due to the different imaging principle and imaging characteristic of thin cloud and thick cloud, it is necessary to use different methods to remove the cloud in MODIS images. For thin cloud coverage areas, an improved homomorphism filtering method is used to remove thin cloud. This method not only removes the information of thin cloud areas, but also ensures the information of thin cloud-free areas unchanged. For thick cloud, the remote sensing image under the thick cloud areas with hardly any surface information, the normal algorithm has little effect on cloud removal, so substitution is the only way to eliminate the effect of thick cloud.The thick cloud zones of the base image are replaced by the same-location cloud-free zones on other images. It is found that the effect of removing cloud is very significant.For the retrieval of land surface temperature(LST), firstly, the atmospheric transmittance is retrieved and corrected with the MODIS data, then Secondly, the vegetation coverage is obtained through the calculation of vegetation index, estimated the land surface radio radiance. Finally, the LST is retrieved through the split-window algorithm and the effect of cloud on the retrieval results is analyzed.