以20世纪80年代、90年代、2000年以后三个时期的遥感数据和同期的森林资源清查样地数据为基础,应用遥感信息模型的方法,估算了小兴安岭地区森林生物量,并应用GIS和地统计相结合的方法,研究了小兴安岭地区20世纪80年代、90年代、2000年以后三个时期的森林生物量时空变化、空间自相关性和异质性。结果表明:研究区的生物量自上世纪80年代到2000年以后整体上呈现了波动变化,研究区80年代低等级生物量占绝对优势,生物量值普遍较低,呈现低等级生物量连片化分布,空间自相关程度较高,但80年代的中高等级生物量的随机因素增强,说明人为干扰程度不断加强。研究区90年代的中等级生物量占主要优势,主要是80年代占优势的低等级生物量转化为中等级生物量。10多年的变化说明,生物量整体上呈恢复趋势;由于数据主要集中在90年代末期,天然林保护工程已经启动,森林状态向着好的方向发展,所以生物量值整体上有所提高。研究区2000年以后生物量整体上的空间自相关性不高,但是中高等级生物量相近,并且在各个方向上均匀变化,中等级生物量呈现大片分布,但高等级生物量斑块的规模较小,破碎化仍然存在,人为干扰等随机因素和空间自相关因素引起的空间变异相近,出现稳定趋势。

Based on remote sensing data from three periods of 1980s, 1990s, and the period after 2000 as well as the data of plots in forest resource inventory over the same periods, forest biomass of Xiao Hinggan region was estimated by using the remote sensing information model. With the combination of GIS and geo-statistics, this paper studies the temporal changes in forest biomass, spatial autocorrelation and heterogeneity of Xiao Hinggan region in the three periods of 1980s, 1990s, and the period after 2000. Results indicated that the overall biomass presented fluctuation change in the research area from the 1980s to the 2000s. With relatively low biological value, low-grade biomass was dominant in the 1980s, and there was contiguous distribution of low-value biomass, with high degree of spatial autocorrelation. However, the random factors of medium and higher biomass increased, indicating the man-made interference degree continuously strengthened. In the 1990s the main advantages biomass in the study area was medium biomass, which evolved from dominant low-grade biomass in the 1980s. The changes in the 10 years showed that overall biomass tended to recover. For the data were mainly concentrated in the late 1990s when the Natural Forest Protection Project (NFPP) had been launched that made the forest status towards a good direction, the overall biological value was increased. After 2000 the spatial autocorrelation of overall biomass in the research area was not high, but medium and higher biomass was similar and changed evenly in every direction. Median biomass was distributed widely, while high-value biomass was of small patches with fragmentation, and the spatial variability caused by random factors such as man-made disturbance or the factor of spatial autocorrelation was just similar, and appeared a stability trend.