Retrieving snow cover in forests of Qilian Mountains from Landsat optional land imager(OLI)

Snow is an important component of the earth surface, it has a significant role in the regional climate change, natural environment and human daily lives. Improving the techniques for global and regional snow cover mapping may benefit both environment interests and hydrological application. In the past several decades, satellite remote sensing is widely used in monitoring the snow cover because it enables observations of large and remote areas. Normalized difference snow index (NDSI) is an effective index in snow cover mapping at large scale, but in forest zone of mountains, the accuracy of snow identifying with NDSI is low. In this paper, some typical Qinghai spruce forest zones of Qilian Mountain in upper Heihe River basin is chosen as example regions. By comparing the Landsat OLI spectral differences between forest with snow and forest without snow, we propose a spectral band ratio named normalized difference forest snow index (NDFSI). Statistical results show that, in winter Landsat8 OLI image, the value of NDFSI about forest with snow has a relatively stable distribution, and it is obviously higher than forest without snow. By choosing the suitable threshold, the snow-cover area in forest can be effectively recognized.