Abstract:

Evaluation of water resource security is one of the key countermeasures for managing regional water security.In this paper,Chaohu Lake watershed was used as a case study for this purpose.In order to rationally determine the weights of evaluation indices for regional water security,we developed a new method that differs from regular ones such as Fuzzy Analytic Hierarchy Process (FAHP),entropy weighted method and principal component analysis (PCA),which were poorly proven in effectively describing and evaluating the parameters or indices within the complicated systems.According to our method, structural decomposition for the weights was performed by recognizing the single sample weights and sample set weights.The single sample weights of evaluation indices were computed by using the accelerating genetic algorithm,which is based on fuzzy analytic hierarchy process and minimum relative information entropy principle.The sample set weights of evaluation indices were computed by using the accelerating genetic algorithm,which is based on the projection pursuit method and maximum information entropy principle.Thus,based on above mentioned method,projection pursuit model (CWEPP) for evaluating region water resource security based on the changeable weight and entropy principles was established.In the context of evaluation of water resource security,the CWEPP model was used at local and nationwide levels.The result showed that water resource security level for Chaohu Laked watershed is between moderate and severe water shortages,but in general the level tends to severe water shortages.This indicated that water resource security level for Chaohu Laked watershed is worse than the overall situation of Anhui Province. The result also showed that compared with regular methods,the CWEPP model is an effective evaluation method for region water resource security due to its distinct concepts of mathematics and physics.Thus,it can be widely applied in comprehensive evaluation for resources and environment