基于可变模糊算法的大连市农业干旱风险评价

农业干旱风险评价是近年来研究热点之一,它不仅是农业干旱风险管理的基础和前提,也是农业干旱风险区划和灾前损失预评估的理论基础。干旱风险受随机性和模糊性等因素的影响,传统的风险评价方法存在这样那样的问题。本文首先基于农业干旱风险理论,从农业干旱致灾因子危险性和承灾体脆弱性两方面分析大连市农业干旱风险,计算大连市2000年-2007年8年的农业干旱风险指数,显示2002年农业干旱风险指数高达4.7366,发生农业旱灾概率最大;然后根据干旱风险指数的结果,构建农业干旱风险综合评价指标体系,应用基于离差最大化的组合赋权法确定指标权重,利用可变模糊法对2002年农业干旱风险进行综合评价,显示大连市2002年农业干旱风险等级为高风险,易发生重度干旱。研究结论表明:本文农业干旱风险的评价方法较可行,计算结果较准确,为区域农业干旱评价提供了一种考虑多因素的综合方法。

Agricultural Drought Risk Evaluation in Dalian Based on the Variable Fuzzy Method

With the development of economy, drought, especially agricultural drought, has become increasingly serious, which would lead to significant losses of economy, society, and the environment. Reducing effects and losses brought by agricultural drought has been a hot spot. In particular, agricultural drought risk evaluation is the key problem. It is therefore indispensable to evaluate the agricultural drought risk since the agricultural drought risk evaluation is not only the basis for agricultural drought risk management, but also a theoretical basis for agricultural drought risk division and loss assessment. Under the influence of fuzziness and randomness, there exist some problems in conventional probabilistic risk analysis methods. To that end, variable fuzzy sets comprehensive evaluation methods are proposed in this study. Using historical data, the authors summarized the characteristics of agricultural drought in Dalian, and analyzed the mechanisms of the agricultural drought disaster. Results show that the agricultural drought in Dalian results from the natural environment and the social environment. Uneven distribution of seasonal rainfall and water shortage are two key factors resulting in agricultural drought. Based on examining factors of agricultural drought and the vulnerability of the hazard bearing body, the agricultural drought risk in Dalian was analyzed, and the corresponding risk indexes from 2000 to 2007 were calculated, in which the agricultural drought risk index in 2002 was 4.7366. It is can be concluded that the occurrence of agricultural drought in 2002 was largest. In addition, five representative indexes were selected to build an agricultural drought risk evaluation index system, including precipitation, drought risk index, river runoff, the degree of crop drought, and the percentage of drought area. Then, the index weights were determined based on the combinatorial weighting method of maximizing deviations. Finally, the agricultural drought risk in 2002 was evaluated comprehensively using the method of variable fuzzy sets. The evaluation results show that the agricultural drought risk was generally higher, with a level of heavy drought. Rating agricultural drought damage is an important part. The model seems to be reliable and feasible, which would provide a promising tool to evaluate the agricultural drought risk. The evaluation results can provide reference for decision making.