| Drought is one of the major natural disasters with frequent occurrence,extensive impacts,and huge losses all over the world,and it has seriously threatened regional water supply security,food security and ecological security.Due to the influence of multiple factors including temperature,wind speed,relative humidity,evaporation capacity,precipitation distribution,human activities and variation of underlying surface in different periods,a great number of complex uncertainties are unavoidably existing during drought occurring,developing,and evolution processes.In addition,as a fundamental task of regional drought risk management research,the quantitative analysis of regional drought hazard system is primarily focusing on the recognition of water shortage condition of hydrological cycle system during a certain period and description of drought evolution process utilizing the joint probability distribution function of drought characteristic variables involving drought duration,drought intensity,drought area et al.In this study,the drought evolution characteristics and comprehensive assessment of drought hazard system basing on different drought indicators were carried out.Specifically,the Cloud model theory and Copula function,aiming to describing the transition relationship between the qualitative description of uncertainty concept and its quantitative analysis,were introduced to establish the integrated quantitative analysis framework of regional drought hazard system.The proposed quantitative analysis approach of regional drought hazard system basing on Cloud model and Copula function was verified in Anhui and Yunnan Province,and research results were reliable and reasonable,and will provide scientific decision making basis for the development and implement of drought-resistant schemes.Overall,the primary findings of the study can be concluded as below:(1)the spatial and temporal distribution characteristics of drought conceptual clouds of different hazard grades in Anhui Province from 1961 to 2007 was analyzed based on SPI and SSMI drought indicators utilizing the cloud transformation and conceptual zooming algorithms.The application results indicated that,the varying trend of drought hazard grade was basically consistent with the variation of latitude in different regions of Anhui Province,i.e.the drought hazard grade was the most severe in northern Anhui province,mostly ranking grade ?(moderate drought)and grade IV(severe drought),and the drought hazard grade in central Anhui Province was primarily ranking grade ?(slight drought)and grade ?(moderate drought),whereas the drought hazard grade in the southern Anhui Province was mainly ranking grade ?(slight drought).Meanwhile,the higher of the drought hazard grade,the greater of the certainty and stability for the distribution of the drought conceptual clouds.(2)basing on the standardized precipitation index SPI and standardized soil moisture index SSMI,a new comprehensive drought index construction method was proposed integrating the Copula function,by which the statistical drought events were recognized according to the Runs Theory,and then the drought characteristics variables including drought duration,drought intensity,et al.,were obtained,and finally the joint distribution probability and its returning period of drought events corresponding to different drought hazard grades in different area of Anhui Province from 1961 to 2007 were discussed.It can be seen from the calculation results that,the most severe historical drought events in Anhui Province occurred in 1966,1978 and 2001 respectively with the returning period above 8a.Meanwhile,the returning period of historical drought events in northern Anhui Province was varying from 2a to 6a,the drought hazard grade in southern and central Anhui Province were primarily ranking grade ?(slight grade)with the returning period below 2a,and the number of severe drought events in southern Anhui Province was 13,with the returning period above 6a.Therefore,it was evident that the probability for the occurrence of drought events above grade ?(moderate grade)in northern Anhui Province will be much higher than that of central and southern Anhui Province in the future.(3)a comprehensive drought hazard assessment model(NPCC)basing on the integrated coupling of positive normal condition cloud algorithm and copula function was proposed,the model was applied in comprehensive evaluation of drought hazard system in Kunming city,Yunnan Province 1956-2011,and the results showed that,during the past 56 years,the drought events occurred in 21 years,among which,the number of the drought events with grade ?(slight drought)were 10 years,8 years with the drought events of grade ?(moderate drought),and 3 years with the drought events of grade ?(severe drought).Thus it can be seen that the variation trend of characteristic value series of drought hazard grade was basically consistent with the actual variation of observed drought hazard indicators.Meanwhile,the comprehensive probability for the occurrence of future drought hazard events above level ?(slight drought)is 62%,and 27%for the occurrence of future drought hazard events above level ?(moderate drought).So the above results demonstrated that the proposed NPCC model can be applied effectively to recognize and portray the variation characteristics of different drought hazard indicators,and the result is more comprehensive,reasonable and reliable than other traditional single indicators.(4)considering the shortcomings for the assumption of cloud drops satisfying uniform distribution pattern when applying for drought hazard comprehensive evaluation analysis,the maximum information entropy principle was proposed to re-optimize and estimate the probability distribution pattern of cloud drops,and then the drought hazard comprehensive evaluation model coupling the positive normal condition cloud algorithm and the maximum information entropy(PCMEP)was constructed.The application results revealed that,?the drought hazard comprehensive evaluation result using the proposed PCMEP model was basically consistent with the actual variation trend of drought hazard evolution in Kunming city,1956-2011.?a negative exponential distribution pattern was showed between cloud drop certainty calculated by positive normal condition cloud and its occurrence probability,and the more the number of cloud drops,the smoother the fitting curve of the probability distribution of cloud drops under different drought hazard levels and the higher the calculation accuracy.? compared with other methods,based on the PCMEP model,the more obvious differences of the eigenvalues of the comprehensive grade of drought hazard evaluation in different years,the more conducive to identifying and determining the final comprehensive grade of drought hazard.(5)basing on cloud transformation algorithm,adopting monthly scale precipitation and runoff percentage index,a new method of regional drought hazard frequency curve fitting was proposed,the method was applied in the analysis of the frequency of drought hazard in Kunming from 1956 to 2011.The probability of drought hazard events in Kunming city calculated using cloud transformation algorithm was in good agreement with the results calculated by empirical frequency formula.? the probability of occurrence of drought hazard events above slight meteorological drought(Pa<-40%)and hydrological drought(Ra<-40%)will be 44.87%and 31.11%,above moderate meteorological drought(Pa<-60%)and moderate hydrological drought(Ra<-60%)will be 31.55%and 19.98%;above severe meteorological drought(Pa<-80%)and severe hydrological drought(Ra<-80%)will be 14.17%and 10.13%;above extreme meteorological drought(Pa<-95%)and extreme hydrological drought(Ra<-95%)will be 4.93%and 1.11%.? the calculation of probability of drought based on percentage of precipitation is greater than the percentage of runoff.?the proposed drought hazard frequency curve fitting method basing on Cloud Transformation Algorithm can provide new insights for the derivation of frequency distribution function of regional drought hazard system,and the corresponding drought hazard frequency analysis results are more reliable and reasonable,which can be further applied in related hydrological frequency research field. |
PDF Full Text Request