Cloud Removing Method And Accuracy Verification Of Snow Extent Product In High Asia Area

High Asia is located in the middle of Asia,mainly including the main part of the Qinghai-Tibet Plateau and the surrounding part of the mountain,with the largest ice and snow reserves apart from the polar circle in the world.The study of snow characteristics,regularity and snow data can deepen the understanding of response to global climate change in high Asia,scientifically assess snowstorm and that will help improve the hydrological simulation,and the level of water resources management,and provide important scientific basis for the sustainable development planning in High Asia.Aiming at the problem of cloud pollution of optical remote sensing image data and the various accuracy of binary snow products,take the commonly used cloud removing method as a reference,this paper proposes an improved cloud removing method based on the temporal and spatial continuity characteristics of snow,through eight steps to remove all the cloud cover:morning and afternoon data synthesis,near day synthesis,"permanent" snow and land method,nearly four pixels,nearly eight pixels,remove the misjudgment of lake and lake ice,maximum snow and land synthesis of eight days,fit the expected snowline method.And generate the daily cloud-free snow products,the products were evaluated by the meteorological station observation data,the results show that the precision of each step is stable and the verification effect is good,which can solve the problem of cloud pollution in the snow product and generate the product data set.Finally chooses relatively typical three kinds of data,carries out the space-time cross comparative analysis using the same ground reference data set and comprehensive index for validation and comparative analysis.The study provides objective analysis and evaluation of snow cover remote sensing product data in High Asian applications.