A Physical Statistic Model for Predicting the Rainfall during Flood Season in Sichuan-Chongqing Region
Ma Zhengfeng 1), and Tan Youbang 2) 1) (Institute of Plateau Meteorology, China Meteorological Administration, Chengdu 610072 ) 2) (Neijiang Meteorological Bureau, Neijiang 629000)
By using 20 meteorology stations anomaly percentage of rainfall data in Sichuan and Chongqing area from June to August, the rainfall distribution patterns of Sichuan basin and its early signals are analyzed. Three rainfall distribution patterns are found in Sichuan basin in main flood period, which are east-west-oscillation, consistent-distribution and south-north-oscillation patterns. According to the three rainfall distribution patterns the relationship between main flood period rainfall in Sichuan-Chongqing region and their signals are as follows: The rainfall of main flood season is more in west and less in east in Sichuan-Chongqing region when the western Pacific warm pool intensity is stronger in spring or the autumn-to-winter geopotential height over Qinghai-Xizang Plateau in last year was higher than normal. Otherwise the rainfall is more in east and less in west. So the east-west-oscillation pattern appears. The rainfall of main flood season is more in Sichuan-Chongqing region when Jan-to-Mar westerlies polar front position tended to be more north or the the winter-to-spring geopotential height on 100hPa over Qinghai-Xizang Plateau in last year was lower than normal. Otherwise the rainfall tend to be less. So the consistent-distribution pattern appears. The rainfall of main flood season is more in south and less in north in Sichuan-Chongqing region when westerlies intensity index is higher in summer of last year. Otherwise the rainfall is more in north and less in south. So the south-north-oscillation pattern appears. By using these strong signals as factor field and basing on the variance weighing, the physical statistic model to predict the rainfall of Sichuan-Chongqing region in main flood season is developed. The result is very confident according to the applicable prediction during year 1996 to 2001.