Natural Disaster Risk Assessment Using Information Diffusion and Geographical Information System

Abstract

With the social and economic development, the losses caused by natural disasters were more and more serious. Natural disaster assessment, management and research are the important field, developing direction and hotspot issues on disaster science and geo-science in resent year. However, because most of the natural disasters are a small sample of events, and uncertainty of natural disasters, so natural disaster risk assessment is particularly difficult based on historical data. Information diffusion theory is useful method for natural disaster risk assessment based on small sample even; it is a fuzzy approach to quantitative analysis of the probability of natural disaster risk. Therefore, the information diffusion theory has unique advantages in natural disaster risk assessment and management. This chapter presents a Geographical Information Systems (GIS) and information diffusion theory-based methodology for spatio-temporal risk assessment of natural disasters, taking grassland fire disasters in the Northern China as the case study. Firstly, we discuss connotation and forming mechanism of natural disaster risk, basic theory and framework of natural disaster risk assessment and management. Secondly, we introduce information diffusion theories and Geographical Information Systems (GIS) in the form of definitions, theorems and applications comprehensively and systemically. Finally, we give the case study on application of information diffusion theory and Geographical Information Systems (GIS) on grassland fire disasters in the grassland area of the Northern China. We employed information matrix to analyze and to quantify fuzzy relationship between the number of annual severe grassland fire disasters and annual burned area. We also evaluated the consequences of grassland fire disaster between 1991 and 2006 based on historical data from 12 Northern China provinces. The results show that the probabilities of annual grassland fire disasters and annual damage rates on different levels increase gradually from southwest to northeast across the Northern China. The annual burned area can be predicted effectively using the number of annual severe grassland fire disasters. The result shows reliability as tested by two-tailed Pearson correlation coefficient. This study contributes as a reference in decision making for prevention of grassland fire disaster and for stockbreeding sustainable development planning. The fuzzy relationship could provide information to make compensation plan for the disaster affected area.