Method Research Of Snow Information Extraction Based On SAR Images In Transmission Corridor Area

Snow and ice information is an important part of the earth surface climate and hydrological study, while the changes of local snow affect local and regional climate system and weather.With the climate problem caused by global warming increasing, snow disaster also had a serious impact in our country, especially occurred in southern China’s blizzards in 2008, sudden,strong and destructive,leading to a large number of high-voltage tower collapsed and transmission lines paralyzed, by which huge economic losses was caused.Synthetic Aperture Radar(SAR) has all-weather,all-time and high-resolution features, which made it having special advantages under a wide range of natural disasters, such as snow disaster,and effectively compensate for the lack of optical satellites.The main content of this paper is snow and ice information extraction based on SAR image in transmission corridor area and analysis and verifying the extraction results.Because the snow disaster that can cause massive damage to the transmission corridor often occurs in the south, the snow will melt and freeze rapidly, which leads to a fast change of snow in time and space.In order to enhance the timeliness of snow extraction, this paper provide a new method to retrieve snow and ice area, which based on the multi-polar ASAR data.In this paper, 14 ASAR images were analyzed, then we use the scattering characteristics and the polarization properties to retrieve wet snow area, based on which the DEM information was used to retrieve dry snow.Finally we realize the extraction of snow and ice in alpine area.In order to match with the MODIS optical data, the results of this paper were scaled.After verification, we found the overall accuracy of the results reached 82.04%.In order to focus on monitoring the serious disaster area in the transimission corridor, we use the Terrasar-x staring spotlight mode datas, whose resolution up to 0.25 m, to retrieve snow information with InSAR technique in the last chapter.By using multiple sets of InSAR datas, we analyzed the coherence of different objects in the Qionghai area,which belongs to the transmission corridor of Yuepu the second-tier.We found in the case of a shorter time interval,except the snow, lake, bushes,other features such as electrical towers, towns and bare soil have maintained a high coherence. According to the coherence difference between snow and other surface features, we take the coherence threshold classification method to retrieve the snow information of Qionghai area.Due to the backscatter coefficient of water is lower than snow’s, we excluded the water area by backscatter coefficient threshold method to get the final snow classification chart.Finally the snow extraction result was compared with the same period of the Googleearth optical image.It shows that the result of snow information extraction with InSAR technique is relatively accurate.However, because the backscattering coefficient and coherence of the bushes are similar to the snow area, this mothod does not distinguish between bushes and snow areas effectively.