Abstract:

Snow is an important factor in climate change,thus,the extraction of snow cover by Remote Sensing data is of great significance in theory and practice for its advantage of realtime and widerange.Using HJ1A/1B satellite data,combined with the theory of snow remote sensing monitoring,this paper fully analyzed spectral response characteristics of HJ1A/1B satellite data.Then the widely used Normalized Difference Snow Index (NDSI) was introduced to the HJ1A/1B satellite,and a snow cover extraction method HJNDSI which integrated CCD and IRS image was proposed.To avoid the data acquisition problem in the two different sensors caused by the differences of swath width,scanning and other factors,HJNDSI was improved in this paper.Only based on IRS data,an improved extraction method based on the Modified Normalized Difference Snow Index (MNDSI) was put forward.Using multitemporal HJ satellite data,the parameters in the two models were identified.Finally,taking Pulan Xizang as the study area,extraction accuracy of the two results by HJNDSI and HJMNDSI methods was assessed.They were 9766% and 9492%,respectively,and both of them could meet the needs of practical applications.However,HJMNDSI method could guarantee a larger range of snow monitoring in this study.