蚌埠市农业旱灾脆弱性综合评价

Comprehensive assessment of agricultural drought vulnerability based on improved cloud similarity: a case study of Bengbu city

农业旱灾系统脆弱性评价在农业旱灾风险管理中具有重要意义。为定量评价蚌埠市农业旱灾脆弱性,运用基于改进云相似度的方法建立了农业旱灾脆弱性评价模型。根据云模型的特点分析评价过程中存在的随机性、模糊性等不确定性,通过期望、熵和超熵3个参数定量分析农业旱灾系统的脆弱性。评价结果表明,基于改进云相似度的综合评价所得均值与模糊综合评价结果大致相同,用熵和超熵两个参数来反映可能取值范围、离散程度以及稳定性等不确定性,可充分考虑脆弱性、风险等的不确定性,为旱灾风险调控与管理提供有力依据。

The agricultural drought causes crop damage and failures,which in turn can also lead to food shortage problems and even famine. Drought risk is composed of the risk of disaster causing factor and vulnerability of disaster-bearing body,in short,it is not easy to control the risk factors and disaster risk only by reducing the vulnerability of disaster-bearing body. Therefore,it is of great significance to assess the vulnerability of the agricultural drought system in the agricultural drought risk management. Bengbu city is a big agricultural city in Anhui Province and even in China,with frequent droughts and serious drought damage. A vulnerability assessment model on the basis of an improved cloud similarity is established to assess Bengbu city agricultural drought system vulnerability according to the characteristics of the cloud model,and fully considering the uncertainty existing in the assessment process; and analyses are made through three digital features： expectation,entropy and the hyper entropy,providing a simple and effective method for quantitative analysis of the uncertainty of the agricultural drought system. The assessment results show that the expectation calculated by the comprehensive assessment method and based on the improved cloud similarity is the same as the result of a fuzzy comprehensive assessment. At the same time,there are two parameters： the entropy and the hyper entropy,reflecting the uncertainty including possible range of values,dispersion degree and stability; and the uncertainty of vulnerability and risk can be fully considered,which can provide a strong basis for drought risk regulation,control and management.