Analysis Of Drought Characteristics In Southwest China Based On Joint Distribution Function

In recent years, the frequent drought in southwest exception events brought huge economic losses to the local. At the same time, it had attracted attention of people from all walks of life. Based on the standardized precipitation index data of 89 meteorological stations in southwest China (Sichuan province, Yunnan province, Guizhou province, Chongqing) during 1961-2010, the probability and interval time were analyzed by employing the theory of run and the Copula function. The droughts with duration less than 4 months and the severity less than 3 account for 80% of all the droughts in southwest China. Drought types are mainly light within a month, moderate within a month, severe within a season and severe within inter-seasons. The frequency of occurring of drought with a duration less than 3 months and the severity less than 1 is higher in the eastern and western areas, and the central region has higher frequencies of occurrence of drought with a duration greater than 3 months and with severity greater than 1.5. Return periods of drought in eastern and western areas are longer than that of central region. More obvious spatial difference appears with longer duration and greater severity. The probability differences of the same drought type in different climatic types are within 0.05.The stability of distribution parameters needs larger sample size. The sample size is greater than 50 in some regions and the requirement of sample size of each parameter is not consistent. The sample size of severity distribution parameters is the largest. The probability and return period of which the size of sample is about 10 has no significant difference (the significant level is 0.05) to the size of sample is 40 in most of region. With the results as the standard, statistical model can greatly reduce the requirements of the sample size.And then it demonstrates that the distribution function of drought duration and drought severity still can be established in the lack of measurement date and the inconsistency of starting and ending time. Climate warming has no impact on the minimum of sample size. The fluctuation is mostly between-5 to 5. Statistical model has a certain stability. Meanwhile, the division of climate state reduces the need for distribution test sample size and makes it easier to build models. The theoretical distribution of affected area and affected area rate is consistent with its statistical distribution in Huize county and Mengzi county. The distribution function can pass the KS test (significant level is 0.01). The 3-D distribution function including drought severity, drought duration and affected area rate was built to describe the probability of different drought type. The results show, joint probability has a great increase with increase of affected area rate. If the affected area rate is between 10% and 30%, the probability between the drought duration level 1-2 is the largest. Besides, if the affected area rate is greater than 80%, the probability of drought duration level 3-4 and drought severity 4 is the largest. The complex network containing teleconnection areas and the key areas on sea was built and the results show that the areas of the East Asia-Pacific 1 and 2 and the Pacific/North American pattern 1 and 4 have lager degree. That is to say these four atmosphere regions have stronger effect to other regions in the atmosphere and sea, the North Pacific Oscillation 1 and the East Asia-Pacific 3 have low degree, so these two regions have low effect in the region interaction. In addition, it has been proved that the network has certain robustness by removing nodes in turn.