Spatial And Temporal Variation Characteristics And Precipitation In Xi’an Region Prediction Research

Xi’an, located in the middle of the guanzhong plain and belongs to the warm temperatesemi-arid, semi-humid continental monsoon climate, is the city short of water resources.Asthe rainfall is one of the recharge source to the water resource and its change has a greatinfluence on the water, so study on the space-time change trend and distributioncharacteristics of the rainfall in Xi ’an region could provide important scientific basis foranalysis and calculation of water resources，the reasonable development and utilization ofwater resources, as well as the construction of water conservancy projects.Combined with “the great changes of water cycle caused by Xi’an urban construction”project and basis on the relevant research inside and outside countries, this paper fullyanalyzed the rainfall temporal variation characteristics of Xi’an, and predicted the rainfall offollow year. The main research results are as follows:(1) Based on the linear trend method,sliding average method, anomaly methodand accumulative anomalies method, the trend of rainfall,temperature and evaporation of7weather stations in Xi ’an region were analyzed. Then gave trend judgements of rainfall,temperature and evaporation according to the Mann-Kendall statistical method. The resultstells that the average temperature and rainfall are on the rise and the changing rate arerespectively0.02℃/10a and9.90mm/10a, the average evaporation is on the decline and thechanging rate is-85.61mm/10a.(2) The results of the Mann-Kendall mutation test shows that the mutational point ofannual mean temperature is at2010, the mutational point of annual mean evaporation is at2001, the mutational point of annual mean rainfall at1979.Besides,the mutational points insummer, spring,autumn and winter of the3aspects above were also analyzed.(3) The spatial interpolation figures of annual mean rainfall, temperature and evaporationbased on the Kringing interpolation method indicate that the space variation of annual meanrainfall increasing gradually from northwest to southeast, the annual mean temperaturedecreasing gradually from northwest to southeast and the annual mean evaporation decreasinggradually from north to south. (4) According to differences analysis of rainfall（1975～2012）, temperature andevaporation（1993～2012）, the lowest temperature existed in1993while the highest existedin2006and the extreme ratio of these weather station was1.2, and the largest rainfall andevaporation existed in1983and1997, the year fewest rainfall and evaporation exited wasuncertain. Besides, the evaporation extreme ratio of these weather stations were between1.3and1.6while the rainfall’s were between2.5and3.8.(5) The correlation analysis of climate factor tells that the correlation betweentemperature and evaporation is positive while the correlation between temperature and rainfalland the correlation between rainfall and evaporation were uncertain.(6) Two methods with BP neural network were used to forecast5years rainfall and floodseason rainfall in this paper. One was based on the influence factors, the other was base on theTime Series Analysis. It is concluded that both of the two methods were suitable for5yearsrainfall forecast and flood season rainfall forecast, but the accuracy of the latter one is higherthan the former one.