Research Of Drought Spatiotemporal Distributions And Forecasting Model In The Typical Climatic Regions

Water resource is the basis of economic and social development, and water disaster is amajor factor of restricting China’s economic and social sustainable development, in order toguide the industrial and agricultural economy’ development more prosperous, stable andhealthy with long-term, it’s of great significance that researching different periods’ droughtspatiotemporal distribution and forecast in different climatic regions. According to differentclimatic regions’ precipitation characteristic, this study divides6different time scales of1,3,6,9,12and24months for respectively dealing with91stations’ monthly precipitation andrunoff data of Jianghan Plain, Guanzhong and Sinkiang, and adopts standardized precipitationindex, Z index, standardized runoff index, runoff Z index to estimate different periods’drought degree of the three regions in different time scales. At first, mainly from two aspectsof meteorology and hydrology, it explores different periods’ drought spatiotemporaldistribution and variation law in different climatic regions; Secondly, adopting cause analysisand mathematical statistics analysis and making full use of data’s periodicity andperiodization, this paper proposes a new constructing and forecasting method of fuzzy timeseries model, and it investigates the applicability and practical value of the model. The mainconclusions are as follows:(1) Drought and waterlog’s spatiotemporal distribution and variation law of Jianli drawnfrom standardized precipitation index analysis agrees with the fact that drought and waterlogis frequent and happened alternatively in the region. Standardized precipitation index isrelatively suitable for drought monitoring and evaluation.(2) Sinkiang’s drought and waterlog happened frequently and alternately, drought andwaterlog’s occurrence is symmetrical to each other, but the probabilities of drought andwaterlog happened in the different periods or classifications are different; The spatiality ofdrought and waterlog’s distribution happened in different periods or classifications inSinkiang is rather obvious and intense; The spatial distributions of different droughts andwaterlogs which are in the same period and classification are different and change alternatelyon some extent; The most area of Sinkiang becomes more and more moist and the speeds change intensively with spatial regions in different periods. It is relatively valuable forexploring drought and waterlog’s spatiotemporal distribution and variation law based on SPIand ArcGIS.(3) Standardized runoff index and runoff Z index’s evaluations to drought and waterlogare coincident; Standardized runoff index has a certain applicability in different climaticregions; Kenswat becomes more and more moist in each periods of different time scales butJianli’s wetting is not the same in different periods. The acuteness degree of drought andwaterlog’s changes are different both in Jianli and Kenswat in different time scales as well asthe changes are periodic and alternate on some extent. The standardized runoff index deducedin this paper is possible and it has a relatively good practical effect using in the research ofdrought and waterlog’s distribution and variation law.(4) All of different time scales’ drought and waterlog in Guanzhong may happencontinuously; Drought and waterlog’s average instability is inversely proportional to theprecipitation in Guanzhong at3-monthly and6-monthly time scales; Drought and waterlog’sinstability presents a certain periodicity changing characteristic at all time scales andinstability’s fluctuation extent is proportional to time scale’s length by and large inGuanzhong; At long time scales, Guanzhong’ instabilities are downward. Instability canreflect acuteness degree of drought and waterlog’s change well as well as it is simple and easyto application.(5) Comparing with contrastive model, the paper’s model has better accuracy andstability in most of the periods; Except a few periods, the model’s forecast error and itsfluctuations are small; If data’s periodicity and periodization are more obvious, the model’sforecast accuracy is better. The fuzzy time series model proposed in this paper is reasonable,calculation is simple and it has relatively good accuracy and stability.(6) In2015, the water regime’s evaluation of Guanzhong and Sinkiang’s most regionsbased on SPI will normal as well as drought and waterlog’s spatial variations are relativelyobvious in the two regions, but the spatial variation is slightly larger in Sinkiang; In2015, thewater regime’s evaluation of Manas river basin based on SRI tends to be droughty and itsspatial variation is smaller than to Heihe River basin. The fuzzy time series model hasrelatively good applicability in prediction and practical value.