Mapping snow cover using HJ-1B infrared remote sensing(IRS) spectroradiometer images

Snow plays an important role in global hydrologic and climate changes. After 40 years of research on using multi-spectral remote-sensing data to monitor snow, researchers developed a large number of algorithms and products, such as snow mapping, snow depth inversion and snow albedo. However, algorithms for snow cover extraction from Chinese satellite images are yet to be developed. In this study, we firstly verified the accuracy of HJ-MNDSI snow cover mapping method, and then determined the appropriate thresholds by comparing with the results with the Landsat Thematic Mapper (TM) data. Then we produce snow cover products nearly seven years and calculate snow pixels and its proportion. Finally, according to the historical data, we give a forecast method by using MATLAB to fit the curve based on Fourier. The results provide scientific basis for the rational using of water resources and climate or environment change researches of the Tianshan Mountain.