中国森林火灾周期振荡的小波分析

The Wavelet Analysis of Periodicity Variation of Forest Fires in China

小波分析具有自适应性和数学显微镜性质,能将交织在一起的不同频率成分组成的复杂的时间序列分解成频率不相同的子序列。本文采用Matlab 6.5软件对1950-2002年中国林火统计资料进行小波分析,采用的小波是Daubechies 10。结果表明,1950-2000年森林火灾受害森林面积总体上呈下降趋势,1995以后受害森林面积基本控制在比较低的水平,年际间变化不大。10年和30年尺度上的变化反应出气候和植被变化对森林火灾的影响,可燃物的变化表现出近10年周期的波动性。1950-2000年中国森林火灾次数整体上是下降的,森林火灾发生次数减少的主要原因是我国对森林火灾的预防、监测和扑救能力都得到显著提高,特别是对人为火源进行了严格控制,近年来林火发生次数保持在相对低的水平。

Wavelet analysis has the attributes of adaptability of itself and mathematics microscope. It can decompose the complex time series into the components with different frequencies. The paper uses the software of Matlab 6.5 to analysis the forest fire statistics data of China during 1950-2002 by using wavelet Daubechies 10. The results show that the burned forest areas are decreasing generally during 1950-2000 and after 1995 the burned areas keep in low amount and its change insignificantly among years. influence of climate and vegetation change on forest fires. The scales of ten years and thirty years have reflected the Fuels show a change process in ten years period. During 1950- 2000, the forest fires decreased generally, which is the result of the improvement of fire prevention, monitoring and attacks. Especially the government pays more attention to manage human fire resources and keep the forest fire in low amount.