The Characteristic Analysis Of Drought Time Series Based On Fractal Theory

Drought is one of the most common natural disasters in our country,it has the characteristics of long duration,widespread and huge influence. Drought not only restricts the social and economic development in our country,but also destroys the ecological balance.Therefore,research on drought data has very important practical significance. The occurrence of drought is the result of joint action which contains natural, social and many other factors. Drought data often presents complexity and nonlinearity, so the conventional data analysis and modeling method can not reveal the essential information. This dissertation puts forward the subject of drought time series that based on the theory of fractal characteristic and apply this method to the study of drought data, according to the characteristics of the drought data.Firstly, this dissertation takes the drought inundated area which occurs in China between 1978 and 2013 as an example, which uses R/S analytical method to do single fractal analysis for time series of drought inundated area. The result shows that the sequence of data has an obvious positive correlation and the fractal characteristic is obvious; Secondly, this dissertation discusses the multifractal characteristics of drought time series which uses multifractal analytical method. It confirms that drought time series have more refined multifractal structure and we get the corresponding multifractal spectrum; Thirdly, this dissertation uses the drought time series which occurred from 1974 to 2004 as the research object and uses fractal interpolation method to do data interpolation whose data comes from a sequence. The result shows that just using a small amount of known data can completely achieve describing the changing process of the historical data and the results are ideal. Finally, apply the fractal interpolation prediction model which is established to forecast the drought disaster area in 2005.Therefore it is quite reasonable to use the fractal theory to analyze the data of drought time series.