Prediction Of Regional Forest Fire Based On Information Diffusion Theory

With the development of economic and further study of forest fire research, study of forest risk has become a new field of forest fire research. As an important factor affecting economic development of China,forest fires severely damaged eco-system and caused huge economic losses.It is necessary to make research on forest risk. we used a new scientific theory--information diffusion theory to predict and analysis forest fire risk.It aims at advicing to decision maker on management and prevention of forest fire, reducing the losses that forest fire brought and ensuring of people’s lives and property safety as well.With statistical methods and GIS, we analyzed forest fire risk spatial and dynamic characteristics based on data from the year2000to2012of Hunan Province.Based on the theory of information diffusion and calculation programming function of Matlab, using the forest fire number as the index, were calculated the forest fires risk and got forest fire risk zoning map with the help of GIS, the main contents and conclusions of the study are as follows:(1) According to the spatial characteristics of forest fire and the characteristics of the study area, combined with GIS, we analyzed the dynamic and spatial distribution characteristics of forest fire risk in Hunan, and determined the different grades of forest fire distribution area.(2) We calculate risk estimation of different cities in Hunan based on information diffusion theory.Risk estimate value is beyond probability value, which means exceed the incidence of corresponding level of risk. With the increase of risk level, risk estimation value decreases, risk estimates value of Xiangtan first reaches0, while Shaoyang is the last, risk estimation value is not0shows that forest fires may occur in the corresponding level of risk, risk estimation for0means forest fires the corresponding level of risk will not occur, in this way, we predicted times of forest fires occur interval.(3) According to risk estimate value (r) of various levels,we calculate recurrence period which equal to1/r, and then we predicted the recurrence period of forest fire under different risk levels.For example,when times of forest fire a year rich130, Changde occur every2.8years, Huaihua occur every1.9years, Shaoyang occur every1.2years, Loudi occur every2.9years, Xiangtan occur every7years, Yiyang occur every2.5years, Yongzhou occur every1.6years, Yueyang occur2.8years, Zhangjiajie occur every4.1years, Changsha occur every1.5years, Zhuzhou occur every2.6years, Xiangxi occur every4.1years.(4) We got comprehensive risk estimate value by accumulating risk estimate of different risk levels, comebine with accumulation loss caused by forest fire, we got the forest fire risk of research area, risk value (R) equal probability value (P) multiply by the loss (d), got the forest fire risk value in various cities of Hunan provincethe highest value is1744in Shaoyang and the minimum value is18.79in Zhangjiajie,(5) The forest fires in Hunan province is divided into four risk areas mild, moderate, severe and very severe based on forest fire risk value.Shaoyang city is in the very severe risk areas, Changsha, Huaihua, Yongzhou and Chenzhou in a severe risk area, Yueyang, Yiyang, Zhuzhou and Hengyang in the moderate risk area, Xiangtan, Loudi, Changde, Zhangjiajie and Xiangxi in the mild risk area and got the map of forest fire risk zoning in Hunan province by using GIS technology.Study of forest fire risk has not formed a complete system as it is a new field. With the rapid development of science and technology,new theory,new technology and new method will apply in the study of forest fire risk,in this way,study of forest fire risk and management work of forest disaster prevention and mitigation will step into a new level.