Method Research Of Snow Cover Information Extraction Based On SAR Images

With climate warming and extreme weather events increasing, snow mapping and snow disaster monitoring have become a research hotspot. Synthetic Aperture Radar(SAR) has a lot of advantages, such as all weather, all time, high resolution, which can compensate for the lack of optical satellites. The main contents of this paper are about snow cover retrieve based on SAR data, and analysis of the results.Interferometric Synthetic Aperture Radar(In SAR) can be used to assess the coherence of the surface. This thesis selects several In SAR data pairs from ERS-1/2 and ASAR to count the coherence of snow and other ground objects. The results indicate that snow, water, ice, forest show low coherence in one day temporal baseline In SAR pairs, while coherence of other ground objects is preserved. Compared with one day interval In SAR data, ground coherence reduce significantly for with 35-day interval and 70-day interval In SAR data.This paper retrieves snow cover in several test regions based on coherence differences. Then, the author calculates the accuracy of snow cover mapping and analyzes the reasons for misclassification areas by comparing the classification result with that derived from optical images. The results show that the In SAR technology can retrieve snow cover effectively. Moreover, the short temporal baseline of In SAR pairs can be helped to improve the accuracy. However, this method requires prior knowledge of the test regions, such as climate and topography. Complex surface has a bad impact on the snow retrieve. This paper also analyzes the impact of layover and shadow of SAR image on the snow cover mapping.In the last chapter of the paper, the author uses In SAR technique to detect snow cover in urban areas. By using several In SAR data pairs fro m ASAR, it counts coherence of some ground objects in Urumqi. Due to human activities, crop growth and other factors, the regions show low coherence except airport and large square. It also calculates coherence value during the snow period, and analyzes coherence change of urban areas after snowfall. It turns out that the coherence of airport and suburban wasteland has a significant decline after snowfall. So we can determine whether there is snow in urban areas by calculating coherence of these special features, and then we will decide we should monitor the snow cover areas or not.