Flood Hazard Assessment Based On Spatial Multi-criteria Decision Making And Heterogeneous Spatial Data Integration

China is one of the countries with the most serious flood disasters in the world.Floods occur frequently,affect a wide range,and have a destructive power,which makes them the most serious natural disasters.Floods pose serious threats to the lives and property and hinder the sustainable development of China’s social economy to a certain extent.It is meaningful to make flood hazard and risk assessment and distinguish the flood prone area for optimal arrangement of flood protection measure and reduction of potential losses.Though the traditional flood hazard assessment based on hydraulic and hydrological model can obtain detailed flood information under certain design flood,the model is hardly applied in the data-scarce watershed due to its high demand of data.This paper tries to conduct research on flood hazard assessment based on multi-criteria decision-making method,which can integrate multi-criteria indicators to make flood hazard assessment based limited information of watershed properties.On the basis of the analysis and summary of the flood hazard criteria used in the current research,this paper proposed and established the "THR" flood hazard evaluation indicator system.The current studies of spatial multi criteria decision making for flood hazard assessment were mostly based on spatial homogeneity assumption,which lacks the consideration of spatial heterogeneity.In view of the problem,this paper analyzed the spatial heterogeneity of criteria indicator,criteria weight and risk preference,and the relevant spatial multi-criteria decision-making methods were introduced and applied to analyze and compare the flood hazard assessment results considering or ignoring the spatial heterogeneity in the decision-making process.Finally,the flood hazard assessment method based on spatial multi criteria and spatial heterogeneous data integration model was proposed,which considered spatial heterogeneity of criteria indicator,criteria weight and risk preference in the decision-making process.And the proposed method was applied to make flood hazard assessment in the study area.The main research contents of the paper are as follows.(1)The "THR" flood hazard evaluation indicator system was put forward,and multiple criteria indicators related to topography,hydrology and flood-resistance were selected.In view of the characteristics of water logging in the study area,the indicator to reflect the information of depression was put forward,and the preliminary analysis was performed to keep criteria concise.The framework and estimation indicator system of multi-criteria flood hazard assessment was improved.The fuzzy analytic hierarchy process was introduced to obtain the global criteria weight,and fuzzy number was used to quantify the subjective judgment semantics.Thus,the uncertainty in semantic quantification in the process of subjective criterion weight acquisition was settled.(2)The local spatial multi criteria decision analysis method based on the moving window was adopted to consider the spatial heterogeneity of the criteria,which is mainly embodied in the standardization of criteria and the local adjustment of the global weight.Through the analysis of the moving window,the effect of grid cells within local window on the focal cell was emphasized,and the criterion was locally normalized within local area.The local entropy weight was calculated based on the moving window,and the global subjective weights were locally adjusted according to the local entropy weight.And the spatially variable criteria weights were obtained.A local weighted spatial multi criteria flood hazard assessment model was established and applied to the study area,and the spatial heterogeneity of criteria and criteria weights was considered in the process of flood hazard assessment.The result was compared with result derived from the traditional multi-criteria flood risk assessment based on the spatial homogeneity hypothesis.The result showed that the flood hazard map considering spatial heterogeneity were spatially dispersed in each hazard class,and on the basis of keeping the low level of overall hazard,more attention was paid to the highlighting of the relative hazard in the local area.(3)The method of ordered weighted multi-criteria decision making for flood hazard assessment was studied,and the characteristics of order weights under different decision strategies were analyzed.The flood hazard of the study area under different global risk preference(α)was tested.The results of flood hazard under different global risk preference were summarized and analyzed,and the problem that global risk preference strategy may cause overestimation or underestimation of the hazard was put forward.Aiming at this problem,a spatially variable risk preference allocation method based on regional importance was proposed,and the spatial heterogeneity of risk preference was taken into account in the decision-making process.Taking the study area as an example,the spatial ordered weighted averaging flood hazard assessment model was established.,and the spatially variable risk preference was used to conduct flood hazard assessment.The result was compared with that derived with the global risk preference strategies,and it showed that adjusting risk preferences based on regional importance tended to highlight the flood hazard in the developed and populated areas,while keeping relatively low level of overall hazard.It was beneficial to shift the focus of flood prevention to the relatively important areas.(4)A spatial multi-criteria and spatial heterogeneous data integration model for flood hazard assessment was proposed,which can consider the spatial heterogeneity of criteria indicator,criterion weights and risk preferences in decision-making process.It improve the model to highlight flood hazard in local and important areas,and the hazard area that need to be noticed was reduced,which was conducive to the optimal allocation of limited flood control resources.The integrated evaluation model was applied to flood hazard assessment in the study area.Through the verification of the evaluation results,the integrated evaluation model showed good practicability and reliability in identifying the flood hazard areas in the study area.Finally,the sensitivity of global criteria weight which is one of the most important parameters in the model was analyzed,and the influence of the change of window size on the evaluation results was also analyzed.