Influence Of Climate Change On Spatial And Temporal Structure Of Forest Fires In China

Based on analyzing the variety rules of the climate change in China in recent 35 years,time series of climate change characteristics are established to give quantitative description ofthe change trend. And analyzing the spatial and temporal structure of forest fire, and relativeanalysis is done to describe the relationship between two forest fire characteristics, at last,meteorological fire order models were established on the basis of amount of relative andregression calculation. The major conclusions of this study are as follows:(1) The average temperature in the whole country has been increasing obviously in recent35 years. The temperature increasing in Northeast was most obvious. Also, in SouthwesternChina it has been increasing since 1990's after a long-term low temperature period. However, itin Upper-middle Yangtze basin shows unobvious trend. All over China, the temperatureincreasing shows in every season, evidently in winter and spring.(2) Total amount of rainfall in China has the slight increasing trend in recent 35 years.There is an apparent rainfall distribution of "floods in the south and droughts in the north" in1990's. the precipitation in summer and winter increased more obviously.(3) The change of relative humidity in the whole country has slight decreasing trend.Taking account of the change in temperature and precipitation, it can be found that West Chinais becoming warmer and wetter, especially in Northwest China, while Northeast China andNorth China warmer and drier,(4) The result of building regression equation shows that there is significant correlationbetween forest fire and disaster area. The time distribution of forest fire shows a prominentcharacteristic of seasonal variations. The high fire frequencies are always centralized in Marchand April of spring and February of winter.(5) The space distribution of forest fire is quite imbalance in China. In the whole country,it shows more frequent in the East than in the west, and in North China. High risk regions offorest fire are located in the Northeast and Southwest of China.(6) By building egression equation, the results shows those meteorological factors, whichare separately correlated with the forest fire frequency, include: annual average temperatureand precipitation, spring average temperature, winter average month precipitation, winteraverage wet coefficient, and winter average ratio of precipitation and temperature, and averagetemperature and the ratio of precipitation and temperature of winter and spring. And othermeteorological factors, which are separately correlated with the forest fire disaster area,includes: annual average temperature and precipitation, spring average temperature, autumnaverage relative humidity and average month precipitation and ration of precipitation and temperature, and the average temperature, average relative humidity and average wetcoefficient.(7) On the basis of amount of correlation analysis and regression calculation above,meteorological fire order models were established and significance-tested, with meteorologicalfactors are the independent variables and forest fire characteristics as the dependent variables.At last, regression models are assessed by fitting forecasting grade and B score of precipitationprobability, even with analyzing the correlation of the true value and model fitting value. Thenthe optimal models of forest fire frequency and disaster area are separately fixed, whichcorrespondingly are annual five variables regression model and autumn five variable regressionmodels.