Characterizing snow cover interannual variability with Empirical Orthogonal Function (EOF) analysis and its climate effect in the inland region, northwest China

摘要

Satellite images and field temperature data have been used to derive the winter snow cover change in the northwest inland regions for the period 1978-2005. Decomposition of the satellite-derived snow depth field yields several statistically-significant EOFs (modes) and spectrums of variability according to the selection rule. The first three leading EOF modes account for 63%, 5%, and 4% of the total variance respectively. Spatially, their sensitive regions are characterized significantly with slope orientation, EOF1 towards northwest, EOF2 towards south and EOF3 towards southeast. A primary analysis shows the three modes can be explained by the precipitation events induced by the water vapor conveyed by atmospheric circulation originated from northwest, south and southeast oceans accordingly. As for the temporal dimension, EOF1 shows evidently different from EOF2 and EOF3, and services as a stable process during the 27 years. Their temporal processes correlate temperature in a complicated way. On the whole, the temperature correlates EOF 1, EOF2 and EOF3 more significantly, which indicates the impact of climate warming to the snow cover is, to some extent, acted with circulation originated from different orientations, or as a result of their interactions.