The Influence Of Climate Change To Forest Fire And Classification Of Forest Fuel Based On Remote Sensing In Chongqing

Forest fire is a disaster with most destructive force in the factors of harming forest, causingirreparable damage to the forest.The climate is an important factor of influencing the forest fire.Because of the influence of solar activity, the EI Nino phenomenon reinforced constantly, leadingto global climate anomaly. Due to the influence of the climate conditions, the probability ofcausing forest fire greatly improved.As the study area in Chongqing, using meteorological elements in Chongqing forest and thecollected historical data of forest fire, we get the forest fire’s space-time pattern and its changetrend under the background of climate change, including: the inter-annual change of Chongqingforest meteorological elements in1990-2006; the inter-annual change and inter-monthly change ofChongqing forest fire in2000-2010; the forest fire burned area and fire seat and so on in2000-2010. Using each factor of the fire climate index system (FWI) in Canadian fire ratingsystem as the basic evaluation factors, we study the factor’s applicability of FWI system inChongqing, furthermore we classify the combustible material in Chongqing forest by remotesensing.The results of the study indicate that:(1) A total of11years in2000-2010, Chongqing had658cases of forest fire. The forest firefrequency was stepped distribution; the highest frequency appeared in2006accounting for19.8%of the total number of fire. Forest fire mainly happened in spring and summer, the fire number wasdiminished during April between June, with July of summertime coming, the fire number begun torise significantly. The March and August were the two mouths with the highest fire frequency. In afull day, the forest fire frequency presented normal distribution basically, the forest fire mainlyhappened in10am between18pm, accounting for79%of the total fire number.(2)The largest burned area of Chongqing forest appeared in2006, March and August were twomouths with larger burned area than other mouths. The main reasons of causing forest fire wereburning the grass on waste land and clean burning incineration manure, outdoor smoking andchildren playing with fire. Only the five artificial fire origins cause80%in the total fire number,including burning grass on the waste land accounted for23%, outdoor smoking accounted for18%, incineration manure accounted for16%,clean burning accounted for14%and the childrenplaying with fire accounted for9%. The time of Pine forest fire is the most, accounting for40.6%of the total fire number, followed by the fixed forest of Pine and Cedar, accounting for24.5%ofthe total number.(3)From1990to2006, it showed a trend of a significant rise in temperature, the average of annual temperature for the last5years was0.76℃higher than the first5years’ average. So thereis the potential to increase forest fire. During1990-2006, the annual average relative humidity inChongqing was significantly decreased, including the most obvious downward tendencyappearing in2005-2006, and the frying degree of forest was obvious increased. And thesesituations have the potential to increase the forest fire. The annual average wind speed inChongqing rose significantly during1990-2006, the last5years’ average wind speed increased by0.77m/s than the first5years’. Between1990to2006, that the annual rainfall in Chongqingshowed a decreased trend directly caused the forest fuel moisture reduced. This situation also canincrease forest fire.(4)In the fire period(2005-2006), the values of FWI、FFMC、DMC、DC and BUI weresignificantly higher than no-fire period’s. It indicated under the condition of the temperaturecontinued to rise, the humus layer fuel in fire period was drier than the fuel in no-fire period. Thissituation can lead to increase the probability and intension of forest fire. As a result, the possibilityto cause forest fire increased.(5)In the year of2000, the remote sensing classification accuracy of forest fire fuel inChongqing is: the overall accuracy is84.83%; the Kappa coefficient is0.8178. In the year of2010,the remote sensing classification accuracy of forest fire fuel in Chongqing is: the overall accuracyis83.33%; the Kappa coefficient is0.7995. From2000to2010, the total forest covered area inChongqing was increasing, the total forest coverage rate increased from38.23%to41.75%,increased by3.52%.