The clustering analysis method is employed to divide the rainfall field in spring and summer over the ThreeRiver Source region into three areas for predicting the rainfall distribution and rainfall trends for different areas. Based on the correlation relationship of the spring and summer rainfall indexes in the three areas with the Pacific SST in the preceding season, the SST distribution index is defined. When the winter SST over the western Pacific is lower (higher) than normal, the higher (lower) than normal SST distribution in the central and eastern equatorial Pacific is correlated with the decreased (increased) spring rainfall in Areas 1 and 3 of the ThreeRiver Source Region; while the higher (lower) than normal winter SST distribution in the central equatorial Pacific and off the coast of California is correlated with the decreased (increased) summer rainfall in Areas 2 and 3 of the ThreeRiver Source region. Further analyses are conducted of the correlation relationship between the winter SST in Pacific and the spring and summer 500 hPa geopotential height field in the Northern Hemisphere. The results show that when the overall winter Pacific SST index is high (low), the Indian high pressure system, central Siberia trough and Aleutian trough at 500 hPa in spring strengthen (weaken) with increased (decreased) spring rainfall in the ThreeRiver Source region. In contrast, when the overall winter SST index in the central Pacific and off the coast of California is high (low), the Iranian high pressure system and central Siberia ridge in summer strengthen (weaken) with the southward (northward) displacement of the western Pacific subtropical high and decreased (increased) summer rainfall in the ThreeRiver Source region.