Extraction of Snow Cover Fraction and Snow Depth in South Korea from MODIS images

摘要

Observations in the visible and infrared spectral bands from the MODerate resolution Imaging Spectroradiometer (MODIS) have been used to derive snow depth from snow albedo. The information on the snow depth has been widely used in hydrological and numerical weather prediction, disaster managements and other environmental applications. Among the remote-sensed imaging systems, the Terra/Aqua MODIS satellite imaging systems have demonstrated success in surface monitoring, because they have been acquired for over ten years with 98-99 minutes repeat cycle. In this study, we create snow depth maps for three heavy snowfall events in South Korea, which occurred in January 2010, February 2011 and December 2012. In order to estimate the snow depths from satellite images, the following three steps are needed: 1) the generation of a snow cover map from the Normalized Difference Snow Index (NDSI), 2) the creation of the snow fraction cover map from the generated snow cover map, and 3) the production of the snow depth map from the snow fraction cover map. In order to establish a statistical relationship between satellite-observed snow fraction and snow depth, we have used snow depth data over the South Korea reported from NCDSS (National Climate Data Service System) stations. For the validation of the results, we compare the snow depths estimated from MODIS images with the in-situ observations.