| Flash flood is one of the most dangerous disasters,causing seriously threatens on the safety of people's lives and property.How to comprehensively and objectively describe the falsh flood pattern and its evolution law,deeply understand the cause mechanism and mechanism evolution process of flash flood,scientifically evaluate the vulnerability of flash flood in different regions,and efficiently and rationally give the optimized allocation of flash flood prevention funds is of great significance to reduce the loss of it.In view of this,this paper proposes four general frameworks for the cognition and prevention of flash flood,including multi-scale distribution pattern and evolution analysis framework for flash flood,cause mechanism and mechanism evolution analysis framework for flash flood,multiple vulnerability evaluation framework for flash flood,and support funds optimal allocation framework for flash flood.The four frameworks provide all-round research on the prevention and control of flash flood from the perspectives of the cognition,mechanism analysis,vulnerability assessment and resource allocation.Meanwhile,the four frameworks are applied to Hainan province,presenting the comprehensive analysis and research results for the flash flood here.The main research contents and achievements of this paper are as follows:(1)The multi-scale distribution pattern and evolution analysis framework for flash flood is constructed.Based on the three spatial scales of county,township and small river basin,the framework analyzes the pattern and evolution of flash flood.In the fhash flood pattern analysis,based on statistical analysis,GIS technology,Lorentz curve,Gini coefficient,global and local correlation analysis and other technical means,the flash flood difference analysis,geographical location analysis,distribution equilibrium analysis and spatial correlation analysis are discussed.In the analysis of pattern evolution for flash flood,the evolution of flash flood pattern with time is analyzed based on M-K test and Markov chain,and the evolution in spatial pattern is analyzed based on global and local correlation analysis.The framework is applied in Hainan Island and the research results are shown as flollows: The overall difference,geographical distribution,equilibrium and spatial correlation of flash flood under different spatial scales are both internal and consistent,while also have some different rules,reducing the space scale can find more details of it.The number of flash flood in Hainan experiences sudden changes in 1995.After 1996,the number of flash floods in each year increases significantly.The evolution patterns of flash flood in different counties and different stages are various,and the situation of flash floods in 1996-2014 is more complicated than that in 1905-1995.In recent years,the spatial accumulation pattern of flash floods in Hainan province is more significant,and the spatial accumulation evolution pattern of township scale is most obvious.The local spatial aggregation patterns of flash floods in the two-stage and three-scale have significantly changed.(2)Cause mechanism and mechanism evolution analysis framework for flash flood is established.Based on the literature summary,the framework gives an index system for flash flood causes analysis including the environmental factors,social factors and rainfall factors,total of 14 indicators.The framework presents typical stage division of flash floods by the M-K test.Then the cause mechanism and evolution of mountain flood disaster are analyzed from the angle of the single factor and double factors,combining different partition to the static factors and dynamic factors.The process is coupling geographical detector and optimization of hot spots analysis.Taking Hainan province as the research area,the research results show as follows: Firtstly,the main driving forces of flash floods are natural factors like Ele and Soil in both periods,and the driving forces from them are on the rise.Secondly,the influences of short-term heavy rainfall(maximum 10 minutes rainfall and maximum 1hour rainfall)and annual rainfall on flash floods are becoming more and more serious.Thirdly,the driving forces from social factors to flash flood seem to be small,but the impact of growing population density on that is significantly increasing.(3)A multiple vulnerability evaluation framework for flash flood is constructed.The concept multiple vulnerability is proposed in the framework.Based on previous research results about vulnerability,the vulnerability assessment index system is built A new methord for index weight determination is put forward based on geographic detector.An inhomogeneity coefficient for quantifying the inhomogeneity between the indicators for vulnerability,based on the wooden barrel theory and entropy method is proposed.Combining with comprehensive index method,the three elements of vulnerability of exposure,sensitivity and adaptive capacity are evaluated.Further,giving full consideration of the relationship between the three elements of vulnerability and based on the coordinated development degree model,six types of vulnerability assessment multi-measure measure four class level are put forward for better flash flood vulnerability assessment.The framework is applied to Hainan island and the reults are shown as follows: the multiple vulnerability assessment can find the potential high-risk areas in the region,and there are still some potential disaster-damaged areas in Hainan.Most parts of Hainan province still have large potential disaster frequency.In terms of the relationship of sensitivity and adaptive capacity,most of the catchments are selfmanaged type.Based on the results of multiple vulnerability assessment results,this paper further puts forward the "three,two and one" decision-making idea for flash flood support,which gives priority to three negative regions,two negative regions,and the last single negative regions,providing necessary decision-making reference for resources allocation.(4)Support funds optimal allocation framework for flash flood is established.Based on the state analysis,this framework fully considers the impact of vulnerability on the allocation of funds for flash flood prevention and control,coupling with the multiple vulnerability assessment and the multi-objective funds optimization allocation model.The priority support areas are determined based on the "three,two and one" decision-making idea.According to the analysis of historical data and geographic detector,investment facilities and economic coefficient can be determined.Then the Fund Allocation Optimization Model(FAOM)model is established and the elite genetic algorithm and TOPSIS are used to quantize and solve the model,giving an optimal fund allocation scheme.The framework is applied to optimize the allocation of funds for flash flood prevention and control in Hainan province.The optimal fund allocation scheme for flash flood prevention in three spatial scales of Hainan province is presented.The research results show that the total number of small basins in hainan province changed from negative type to positive type is 672,and the benefit area is 4486 km2?Comparing with Flash Flood Intensity Scheme(FFIS)and the Control Villages Intensity Scheme(CVIS)for fund allocation,the FAOM scheme and the two schemes have internal consistency for the fund allocation.Moreover,the comprehensive benefits are significantly improved by the FAOM model,which indicates the rationality and superiority of the framework proposed.Finally,this paper puts forward the in-depth application idea of the framework for progressive identification and phased support in flash flood prevention,which provides reference for scientific,comprehensive and sustainable optimal fund allocation for flash flood reduction. |
PDF Full Text Request