Construction And Applicability Of Comprehensive Drought Index

Drought is a frequent natural disaster.With the change of global climate,extreme drought events are happening more and more frequently,which has brought a great impact on the development of agriculture,industry and social economy in China.Drought index is an important tool for monitoring and evaluating the occurrence,development and evolution of drought.The traditional single-type drought index focuses on one aspect of drought and is difficult to describe the complex characteristics of drought development and transmission.The comprehensive drought index based on weight method,principal component analysis method,Copula function method and other methods has gradually attracted more and more attention.However,there are few studies on the comparison and applicability analysis of various synthetic drought index construction methods.Therefore,the comparative study on the construction method and applicability of the comprehensive drought index can provide a basis for comprehensively and systematically guiding the drought prevention and relief work.This study takes the upper and middle reaches of Heihe River Basin as the research object,calculating three kinds of single type of drought index based on parametric method and nonparametric method,including characterization of meteorological drought index(SPEI),standardized precipitation evaporation characterization of the standardization of agricultural drought soil moisture index(SSMI)and characterization of hydrology and the standardization of runoff drought index(SRI).By using entropy weight method,principal component analysis method,Copula function method,principal component analysis-Copula function method and kernel entropy component analysis method combined with three single types of drought indexes,five kinds of comprehensive drought indexes were constructed.Using Pearson correlation coefficient analysis of comprehensive drought index and drought index,the correlation between single type comparing typical historical drought events verify the applicability of the comprehensive drought index,and using the M-K trend test,wavelet analysis methods to analyze the single type of drought index and the comprehensive drought index of the most applicable characterization of space-time evolution characteristics of drought.The main conclusions are as follows:(1)Based on the data from meteorological stations,multi-source remote sensing and hydrological stations,we calculated the SPEI,SSMI and SRI at the resolution of 0.125°×0.125° at the seasonal and annual scales,and analyzed the trend evolution,cycle evolution,chronological evolution and spatial evolution of the three single types of drought indices in the middle and upper reaches of Heihe River Basin.The results showed that 89% of the meteorological drought areas showed increased drought,and 45% of the meteorological drought areas were significantly increased drought.94% of the agricultural drought areas showed drought relief,and 64% of the areas showed significant drought relief.The hydrological drought was alleviated in the whole basin,and the drought was significantly alleviated in 24% of the regions.The three types of drought indices showed a long period of33?45 years,a medium and long period of 28?30 years,a medium and long period of 11?18years and a short period of 3?10 years.In the four periods of 1970 s,1980s,1990 s and the beginning of 21 st century,the severity of meteorological drought is 1990 s > the beginning of21 st century >1980s >1970s,the severity of agricultural drought and hydrological drought is1990 s > 1970 s >1980s >the beginning of 21 st century.The frequency of light and medium drought was less than 45%,while the frequency of heavy and special drought was less than10%.The meteorological drought should be focused on the adjacent downstream region,the agricultural drought should be focused on Qilian County and Sunan County,and the hydrological drought should be focused on Qilian County and the middle reaches region.(2)Based on the three single types of drought indices,five comprehensive drought indices were constructed.The characteristics of drought represented by each index were analyzed and their applicability was verified.The comprehensive drought index constructed by entropy weight method,principal component analysis method and Copula function method showed that the drought trend was decreased,while the comprehensive drought index constructed by principal component analysis-Copula function method and kernel entropy component analysis showed that the drought trend was increased.The comprehensive drought index constructed by entropy weight method and principal component analysis method fluctuates less,while the comprehensive drought index constructed by the other three methods fluctuates more.In addition,the corresponding index value of Copula function method is smaller than that of the other four methods on the whole.The average values of Pearson correlation coefficients between five comprehensive drought indices and three single drought indices were in descending order as follows: Copula function method > entropy weight method > principal component analysis > principal component analysis and Copula function method > kernel entropy component analysis;The optimal ranking of typical historical drought events in spring 1969,autumn 1997 and winter 2009 is as follows: Copula function method > kernel entropy component analysis > principal component analysis-Copula function method > principal component analysis > entropy weight method.(3)The comprehensive drought index constructed by Copula function method is the most suitable one among the five comprehensive drought indices.The comprehensive trend of drought was 70% of the basin was reduced,and 32% of the basin was significantly reduced.The evolution of the comprehensive drought cycle is dominated by the medium and long period of 25?30 years and the long period of 35?40 years.In four different periods,the severity of drought is 1990s>1970s>1980s>the beginning of 21 st century;The frequency of mild and moderate drought is 25%?60%,and attention should be paid to the adjacent downstream areas.The frequency of severe and special drought is 10%?60%,which should be focused on Qilian,Sunan and Minle counties.