论文部分内容阅读
研究基于逻辑斯蒂回归和Arc GIS,以2000—2010年该地区火点数据为基础,结合气候、植被、地形及人口和社会经济数据,对厦门、漳州和泉州等福建南部地区主要森林城市进行森林火险等级划分。在此基础上,应用不同方法对研究区域进行防火瞭望塔选址分析并加以比较,同时根据火险区划结果,综合考虑高火险区域、公路以及居民点的距离等因素进行林火扑救区规划,实现对研究区域森林火灾的管理区划。结果表明:高程、坡度、日最小相对湿度、距离铁路的距离、距离公路的距离、距离居民点的距离、日最高地表气温、日最低地表气温、日降水量、日照时数、日最高气温和GDP等12个因素与林火的发生存在显著相关,且模型的预测准确率达到74.0%。利用该模型进行火险等级区划,发现该地区高火险面积约占33.58%,且分布不均。防火瞭望塔选址的研究结果显示,传统方法、构建20 km×20 km网格和10 km×10 km网格,对于高火险区域的瞭望可见率分别为25.63%、42.25%和63.38%。利用层次分析法对林火扑救区划分的结果显示,到高火险地区距离、到公路距离和到居民点距离的权重分别为0.620、0.284和0.096,通过Arc GIS加权叠加得到相应的扑救区优化图。研究结果可为我国南方山地区域林火预防与管理工作提供一定的支持。
Based on Logistic Regression and Arc GIS, based on the fire point data of the region from 2000 to 2010, the main forest cities in southern Fujian Province such as Xiamen, Zhangzhou and Quanzhou were studied based on climate, vegetation, topography and population and socio-economic data Forest fire rating. On this basis, different methods were used to analyze the location of the fire-proof watch tower in the study area and compare them. At the same time, according to the result of fire risk zoning, considering the factors such as the distance between high-fire danger area, highway and residential area, Regional management of forest fires in the study area. The results showed that: elevation, slope, daily minimum relative humidity, distance to railway, distance to highway, distance to residential area, daily maximum surface air temperature, daily minimum surface air temperature, daily precipitation, sunshine hours, sunshine temperature and GDP and other 12 factors are significantly related to the occurrence of forest fire, and the prediction accuracy of the model reaches 74.0%. Using this model to classify fire-insurance grade, it is found that the area of high-fire insurance accounts for about 33.58% and is unevenly distributed. The results of the study on the site selection of fire watch towers show that in the traditional method, 20 km × 20 km grids and 10 km × 10 km grids are constructed, and the expected visible rates of high fire danger areas are 25.63%, 42.25% and 63.38% respectively. Using AHP method to divide forest firefighting area, the weights of distance to high fire danger area, distance to highway and distance to residential area were 0.620,0.284 and 0.096 respectively. . The results can provide some support for forest fire prevention and management in the mountainous areas of southern China.