不同空间尺度的内蒙古森林火点及面积分布格局研究

Patterns of Forest Fire Point and Area Distribution in Inner Mongolia at Different Spatial Scales

  • 摘要: 通过选用空间变异性分析(变异系数)、全局空间自相关分析、热点分析和尺度方差分析等方法,研究1981—2020年内蒙古森林火点及面积在盟市和旗县2个空间尺度上分布的变异性、关联性及尺度效应。结果表明:盟市尺度林火面积变异系数总体呈上升趋势;旗县尺度林火面积变异系数总体呈下降趋势,旗县尺度空间变异性更大。40 a全局森林火点Moran's I指数为0.99,双空间尺度Moran's I指数均呈下降趋势。其中,盟市级别最大值在1991—1995年,为0.46;旗县级别最大值在1986—1990年,为0.61。双尺度森林火点在空间上均呈现高度聚集性。旗县级尺度更显著地展示了森林火灾的空间分布差异及较小区域的森林火灾热点类型;盟市级尺度揭示了森林火灾的总体空间分布规律,热点类型明显缺乏多样性。盟市尺度热点区主要分布在呼伦贝尔市、兴安盟,冷点区不明显。旗县尺度均存在热点聚集区,林火发生水平较高。不同空间尺度的尺度方差表现为旗县 > 盟市,旗县级尺度林火面积差异对全内蒙古自治区的总体差异贡献率远超过盟市尺度。1981—2020年林火面积在双空间尺度呈现一定的分布规律,火点空间聚集性呈减弱趋势,森林火点与林火面积之间不完全呈正相关关系。内蒙古东部地区由于自然因素和人为因素造成火灾发生次数高于中西部,相关部门应继续加大防火力度。随着尺度的减小,林火空间分布模式愈加明显、聚集度增强,未来需进一步缩小研究范围。

     

    Abstract: The variability, correlation and scale effects of the distribution of forest fires and areas in Inner Mongolia at two spatial scales, namely, league/city and banner/county, from 1981 to 2020 were investigated by choosing methods such as spatial variability analysis(coefficient of variation), global spatial autocorrelation, hotspot analysis and scale analysis of variance(ANOVA). The results showed that the coefficient of variation of forest fire area at the league/city scale showed an increasing trend; the coefficient of variation of forest fire area at the banner/county scale showed a decreasing trend, and the spatial variability at the banner/county scale was even larger. the Moran's I index of forest fire points in the whole bureau was 0.99 for 40 a. Moran's I index at both spatial scales showed a decreasing trend; among them, the maximum value of Moran's I index at league/city level was 0.46 in 1991–1995 and the maximum value of Moran's I index at banner/county level was 0.61 in 1986–1990. The banner/county scale more significantly shows the spatial distribution differences of forest fires and the hotspot types of forest fires in smaller areas; the league/city scale reveals the overall spatial distribution pattern of forest fires, and the hotspot types obviously lack diversity. The hotspot areas at the league/city scales are mainly distributed in Hulunbeier City and Xing'an League; the cold spot areas are not obvious. Hot spot aggregation areas existed in all flag/county scales, and the level of forest fire occurrence was high. The order of scale variance at different spatial scales was: banner/county > league/city. The contribution of the variance of forest fire area at the flag/county scale to the overall variance of the whole Inner Mongolia Autonomous Region is much higher than that at the league/city scale. The forest fire area at the dual spatial scales from 1981 to 2020 shows a certain distribution pattern, and the spatial agglomeration of fire points shows a weakening trend. There is not exactly a positive correlation between forest fire points and forest fire area. The number of fires in the eastern part of Inner Mongolia is higher than that in the central and western parts due to natural and anthropogenic factors, and the relevant departments should continue to increase fire prevention efforts. As the scale decreases, the spatial distribution pattern of forest fires becomes more obvious and the degree of aggregation increases, so it is essential to further narrow the scope of the study.

     

/

返回文章
返回