The Methods Of The Space-Time Analysis And The Risk Model Building Of The Forest Fire In Fangshan Country, Beijing

To answer the request of the praction of the forest fire prevention after reading a great deal of domestic and international documents. the subject of this dissertation was named as "the timespace analysis method and the building of the risk models of forest fire".Two problem will be resolved In this research, the first, to find out when and where the forest fire would take place, what having been pay great attention by the forest manager, the second, to build up a forest fire risk model which can be used in everywhere and at anytime,for allthough there are a great number of forest fire forecast model currently, but few can be adopted in the praction, because of the different condition of different region.The basic train in this paper is that after the data having been processed under normal regulation, including the geography, weather, the forest space distributes data and its social and economic datum.Applying the dining data and knowledges method to analyses these datum, include mul-statistics, the artificial nerve network and wavelet analysis etc.,to find out the relations between the forest fire occurrence with its environmental factors , and to choose the major factors those affected the forest fire's occurrence and spread greatly, based on above results ,the space-time risk model will be modeling, the model includes two parts of contents, the first part is the forest fire ignition risk model , and the other part is the forest fire spread risk model.It is the space-time risk model of forest fire to synthesize both.Fangshan district as research region, its forest fire data, weather data, economic and cultural data from 1986 to 2004 be got.In this research, applied the modern mathematics methods and a great deal of computer softwares, include multivariate statistical, time sequence analysis in mathematics method , geography statistics, grey systematic forecast method, little wave analysis, nerve network etc.; Applied software include ArcviewGIS, ArcMap, SaTscan and MatLab , Spss etc.. When applying ArcViewGIS, the plenty of development modulars of 3th party, as xtools, DNRsampellingtools, SDM etc. was applied.the processes to biuld the space-time risk model are as below.to analyses occurence's trend and periods of the forest fire applying the analytical method of wavelet, and to build up some time squence forecast model of the fire occurrence, with physical mathematics method and the method of geographical statistics( forest fiery quality-heart moving) to study theinfluence of the meteorology factors on the space distribution of the forest fire, to selecte the major factors that influence the occurence of the forest fire influence with major component analysis, and to determine the coefficient of model with the artificial nerve network analysis.Research result shows, the major factors those affect the fire ignition are elevation, aspects,slope,the distance away from the road, population density and relative humidity.the major factors that impact the spread risk are elevation , aspects , slope, the windspread. the most of the fire occurency from 100 to 600 meters in elevation, take 60.4% of the total; within 1353 meters away from road, take 61.0% of the total; from 500 to 1500 meters away from village, take 56.2% of the total; below 15 degrees of the slope, take 72.0% of the total; the forest fire on eastern, eastern south and south slope takes 71.8% of the total. There is a linear relation between the risk and slope, with elevation conic. Tree species affects forest fire intensity , the order according to its flammability from easy to disaster is dahurian larch(larix gmelini), Chinese pine(pinus tabulaeformis), quercus, arborvitae(platycladus orientalis) , mountain apricot(prunus armeniaca var. pendula) , poplar(populus), false acasia(Robinia pseudocacia), populus tremula(pupulus davidiana) , white birch(betula platyphylla).relative humidity impacts on the occurency of the fire directly while other weather factors take affects on the occurency of the fire by affecting the relative humidity.Among these factors, the...