A new approach of dynamic monitoring of 5‐day snow cover extent and snow depth based on MODIS and AMSR‐E data from Northern Xinjiang region

Abstract Taking the Northern Xinjiang region as an example, we develop a snow depth model by using the Advanced Microwave Scanning Radiometer‐Earth Observing System (AMSR‐E) horizontal and vertical polarization brightness temperature difference data of 18 and 36 GHz bands and in situ snow depth measurements from 20 climatic stations during the snow seasons November–March) of 2002–2005. This article proposes a method to produce new 5‐day snow cover and snow depth images, using Terra and Aqua Moderate Resolution Imaging Spectroradiometer (MODIS) daily snow cover products and AMSR‐E snow water equivalent and daily brightness temperature products. The results indicate that (1) the brightness temperature difference (Tb18h–Tb36h) provides the most accurate and precise prediction of snow depth; (2) the snow, land and overall classification accuracies of the new images are separately 89.2%, 77.7% and 87.2% and are much better than those of AMSR‐E or MODIS products (in all weather conditions) alone; (3) the snow classification accuracy increases as snow depth increases; and (4) snow accuracies for different land cover types vary as 88%, 92.3%, 79.7% and 80.1% for cropland, grassland, shrub, and urban and built‐up, respectively. We conclude that the new 5‐day snow cover–snow depth images can provide both accurate cloud‐free snow cover extent and the snow depth dynamics, which would lay a scientific basis for water management and prevention of snow‐related disasters in this dry and cold pastoral area. After validations of the algorithms over other regions with different snow and climate conditions, this method would also be used for monitoring snow cover and snow depth elsewhere in the world. Copyright © 2011 John Wiley & Sons, Ltd.