China Regional Rainfall Changes Of Space-time Mode Analysis

China is a frequent occurrence in the state of climate disasters. Under the background of global warming and extreme weather disasters in China, it is important to study space-time variation characteristics of Chinese regional rainfall. In this paper, based on the rainfall time series data of 160 observation stations in the country from 1951 to 1999, the space-time variation of rainfall pattern in Chinese seven regions were analyzed. It was divided into two parts:one is the research of time series, the other is the research of spatial distribution.The study of rainfall time series is mainly trend analysis, using trend analysis method parameters test and non-parametric test. The study of the trend analysis showed that Chinese regional rainfall characteristics are very obvious. In northeast area's 17 observatories,7 observatories passed significant test and had the remarkable trend. There were 4 stations showing upward trend and 1 station showing downward trend. In Huabei region's 22 observatories,9 observatories passed significant test and had the remarkable trend. There were 2 stations showing upward trend and 7 stations showing downward trend. In East China area's 26 observatories,20 observatories passed significant test and had the remarkable trend. There were 9 stations showing upward trend and 11 stations showing downward trend. In Huazhong area's 22 observatories,12 observatories passed significant test and had the remarkable trend. There were 8 stations showing upward trend and 4 stations showing downward trend. In South China area's 14 observatories,5 observatories passed significant test and had the remarkable trend. There were 4 stations showing upward trend and 1 station showing downward trend. In Southwest area's 28 observatories,17 observatories passed significant test and had the remarkable trend. There were 8 stations showing upward trend and 9 stations showing downward trend. In Northwest area's 31 observatories,20 observatories passed significant test and had the remarkable trend. There were 11 stations showing upward trend and 9 stations showing downward trend. The characteristic of Chinese regional rainfall is notable increase from January to March in east China. In north China the trend is significant increase in May, and significant reduction in July. The trend of central China area, on the contrary, is obvious reduction in May, and obvious increase in July. In east China the rainfall generally reduce from August to December. The rainfall trend in western region of China is not obvious, most months showing increased.In ArcGIS, the rainfall data was interpolated and the result was verified using kriging interpolation, inverse distance weighted interpolation. Generally, the root mean square error of the data can reflect the use of samples and the extreme sensitivity of the valuation. The smaller between average standard error and root mean square error, the better interpolation was, especially the root mean square error the smaller the better. Through validation we could see that the average standard error of simple kriging was minimum, and the average standard error of ordinary kriging was maximum. Because the root mean square error is more important, ordinary kriging interpolation was better than inverse distance weighting, which was better than simple Kriging method.