| With the global climate warming, forest fire number and burned areas has increased in recent years, and serious fire occur from time to time, causing great damage to forest resources and people’s lives and property around the world.There are too much and serious natural disasters in our country,and fires that caused by extreme climatic phenomena has brought great damage to people.As one of four major forest region in China,Yunnan has a high forest coverage rate and rich forest resources, and it is also the badly-disastered area of forest fires.It is very significant and urgent to research the changes pattern of temporal and spatial forest fires in Yunnan provinceunder context of climate change, which can provide a scientific basis not only for the medium and long term forest fire forecast but also for macro decision-making of forest fire prevention in the region.This paper firstly analyzed the linear trend of climate change in Yunnan province in China under context of global warming, and studied the distribution patterns of temporal and spatial forest fires and its influencing factors for nearly 30 years in Yunnan province.Through the analysis of Pearson correlation between forest fires during peaking month and the current and previous climate factors across the province, each ecoregion,each state, the lag effect of climate change on forest fires were discussed at three spatial scales i.e.province, ecoregion and state. According to the law that climate factors influcing temporal and spatial distribution of forest fire, fire behavior (fire number, burned area, fire density, forest damage rate)and the current and previous climate factor (moisture coefficient, diurnal temperature range and average wind speed) were chose to build fire-weather relational model across the province, each ecoregion and each state by using exponential equation and logistic equation.The climatic data under scenarios of SRES B2 climate change were corrected by correlated meteorological observation data and simulated climatic data, and the future distribution of forest fires from 2011 to 2040 were predicted on time and space scale.The main conclusions are as follows:(1) The climate presented a trend of warm and dry in nearly 50 years in Yunnan province, especially in the southeast and northeast region. Temperatures in each region of the province were significantly increased. Precipitation decreased significantly in southeast and northeast region, and decreased slightly in other regions and the change is not significant.Average relative humidity decreased significantly in southeast, southwest, east and northeast region, butthe change is not significant in central and northwest region. Average wind speed decrease significantly in southeast and northeast region, and increased significant in northwest region, and the change is not significant in central, southwest and east region.(2) Forest fire number and burned area presented a decreasing trend during 1982-2008 in Yunnan province, and there was a shift in the year of 1988. Forest fire can be divided into two stages, with severe forest fires in 1982-1988 and a significant reduction of forest firein 1989-2008. Forest fires occurred in December to July of next yeaar, and peaked from March to May. The moments of fire ignition were mainly concentrated in 1.pm to 6.pm during a day. Forest fires distributed very evenly on the space and more fires in the 80s which mainly concentrated in the eastern and central region. In the 90 s, the fires were less, and mainly concentrated in the eastern and northern region. After 2000, fires were mainly concentrated in the central and northwest region. In the county unit, the spatial distributions of fires were mainly affected by moisture coefficient, agricultural labor population and road density.(3) Climatic fators have lag effect on forest fire occurrences. The lag time of moisture cofficent affecting forest fire occurrence is 2-4 months, and forest fire occurrence is mainly related to previous 2-4 months moisture cofficent.The lag time of diurnal temperature range affecting forest fire occurrence is 0-1 month, and forest fire occurrence is largely affected by current and previous 0-1 moth average diurnal temperature range.Wind spees have no lag effect on forest fire occurrence, and forest fire occurrence are mainly related to the current average wind speed.Due to the difference in climate, vegetation and landform among regions, the lag time that climatic factors affecting forest fire occurrence is different across regions, with a difference of 1-2 month.(4) The fire - weather relational models used to predict forest fire numbers and burned areas in Yunnan province were buit by using exponential function.The determinate coefficients value of model expressions (R2) are 0.513 and 0.602, respectively, which suggested the model test are very significant.The comparison between simulated and observed value showed that the fit of the models are good.The models used to predict forest fire density and froest damage rate of each region in Yunnan province were built by using logistic function. Although there are difference in model performance among different regions, with determinate coefficients vary from 0.34 to 0.825, the overall fit of the models are good and have reached the significant level.The comparison between simulated and observed value suggested that the fit of the models in majority of regions are good.(5) In order to correct future simulated value of clamatic factors, relational model related observed and simulated value of climatic factors under scenario of SRES B2 were buit in 16 regions in Yunnan province.The model parameters of climatic factors suggested that the error of moisture cofficent between observed and simulated value is small, and the diual temperature range was secondry.The errors of wind speed between simulated and observed value are large.Fire prediction of tempal and spatial distribution in Yunnan province showed that future forest fires during 2011-2040 will present a rising trend under scenario of SRES B2. Fire number and burned areas are projected to increase by 17.9.2% and 13.6% by 2040s in condition of existing fire prevention policy, and severe fires probably occurred in the years of 2014,2030,2036 and 2030. The future spatial distribution of forest fires will distributed very unevenly. Fires will be mainly concentrated in the northwestern and central region, and the north region will suffer a lot. |
PDF Full Text Request