Description

[Synopsis]

In this study, authors improved the accuracy of snow depth estimation and established an index which represented melting situation of deposited snow from microwave brightness temperature, observed by a space born passive microwave radiometer (SSM/I). A daily dataset consisting of ground measured snow depth and microwave brightness temperature for 339 points over North East Asia, from October 1993 to April 1998 was produced. Data archives produced by NCDC, NSIDC and IMH in Mongolia were used as the data sources.
At most of the points, the difference of brightness temperatures between 19GHz and 37GHz ( ΔT ) indicated almost linear relationship with snow depth. Its slope and offset did not change much on a year to year basis although they changed markedly from place to place. Based on this fact, a multi-linear formula that determined the parameters from five geographic information (climatic maximum snow depth, vegetation index, polarization index of 37GHz in February, brightness temperature of 19GHz in February, and altitude) was established. Using the pair of raster map of the slope and offset generated with the equation, one can estimate snow depth from satellite image more accurately.
It was found that difference of ΔT in the morning and that in the preceding evening, was able to detect the melting snow pack.