关键词: 自然灾害, 风险分析, 定量方法, 适用性

Abstract:

Risk analysis method for natural disasters is one of the key questions of risk research, and it directly affects the data needed for analysis, the selection of mathematical models, and the reliability of analysis results. There have been numerous published research on risk analysis methods. The applicability and reliability of these methods may determine whether the result of risk analysis is useful and may influence risk management. Among existing research, relative level of risk has been examined more often than absolute risk. However, the choice of method should be based on objective of the analysis and risk categories. This article reviews the advantages and disadvantages of various existing quantitative risk analysis methods and analyzes their applications and suitability. The following are found through this review. First, relative risk analysis methods may be divided into three categories, that is, probability analysis, expected loss analysis, and scenario simulation. Among the three types of methods, probability analysis methods have been commonly used in long time series data and macro-scale analysis. The limitation to their application is that historical disaster data may not be easily available. Expected loss analysis methods may be easily applied and are mostly used in analysis at medium -spatial scales, but their predictive power is weak. The precision of the methods based on scenario analysis is high, and they are widely applied in micro-scale analysis where basic data are sufficient. Secondly, for absolute risk, loss of life been the focus of attention and may be examined through historical data analysis, mathematical models, and scenario simulation. Those methods based on historical data analysis need to be improved. Usually the methods for analyzing the risk of loss of life from international studies cannot be directly applied in research for China. Mathematical models for analyzing the risk of loss of life still need to be tested because quantification of qualitative data require greater scrutiny. Scenario analysis methods are most promising and represent the future direction of analyzing the risk of loss of life from natural disasters. In addition, economic risk analysis based on expected losses and land use using remote sensing and GIS techniques are commonly practiced. Ecological-environmental risk analysis is relatively weak among the three types of risks due to the difficulty in quantifying ecological-environmental values. In the future, more attention should be paid to scenario analysis and the role of land use for the absolute risk. Last, similarities and differences are found among the analysis methods of both relative risks and absolute risks. No matter what types of risks are being assessed, the choice of method should be based on risk mechanism, scale of study, and application of modern technologies in order to improve the usability of risk analysis methods and the reliability of the result of analysis.

Key words: natural disaster, risk analysis, quantitative methods, applicability