An error analysis method for snow depth inversion using snow emission model

The snow depth and the amount of regional snow play an important role in the climate change and also affect the management of water resources, and the estimate of the runoff and flood forecasting. At present, snow depth is generally thought to be linearly dependent on the difference of the brightness temperature (i.e. NASA algorithm) in passive microwave remote sensing. But some authors argued according to results of snow emission model simulation that it is not simply linear relationship between them due to the influence of many factors (snow grains radius, land use/cover type). If snow depth is retrieved using the linearly dependent relationship, the error will easily come out. In this paper the graphical representation of probability error is constructed based on the simulated multi-solution curve by snow emission model, which reveals quantitatively and visually the snow depth inversion error due to multi-solution inversion. It will make a significant contribution to reduce the influence of uncertainty and increase the accuracy of snow depth inversion.