不同龄组的热带森林植被生物量与遥感地学数据之间的相关性分析

CORRELATION ANALYSIS OF LANDSAT TM DATA AND ITS DERIVED DATA, METEOROLOGICAL DATA AND TOPOGRAPHIC DATA WITH THE BIOMASS OF DIFFERENT AGED TROPICAL FORESTS

在森林植被生物量遥感动态监测方面最基础性的研究是探讨生物量与遥感数据及其派生数据、地形数据和气象数据之间的相关性。为此 ,以我国云南省西双版纳的热带森林植被为例 ,分别对幼龄林、中龄林、近熟林和成过熟林的生物量与其对应的LANDSATTM数据及其派生数据、气象数据和地形数据之间的相关性进行了分析。首先 ,利用森林资源连续清查的林业固定样地数据 ,通过各树种组的各器官生物量估算模型计算出各样地森林植被的生物量 ,并根据样地的坐标来建立样地GIS数据库。然后 ,利用地形图对遥感图像进行几何校正 ,并对遥感图像进行主成分变换、缨帽变换以及植被指数的计算来产生其派生数据。其次 ,将栅格样地数据、遥感数据 (如LANDSATTM数据 )及其派生数据 (如各种植被指数数据、主成分数据、缨帽变换的亮度、绿度和湿度数据 )、栅格地形数据 (如DEM和坡向 )和栅格气象数据 (包括年平均温度、大于 0℃的积温、年平均降雨量和湿润度 )统一到同一坐标系和投影下 ,并将所有的数据内插为 30m分辨率的格网数据 ,利用样地数据与遥感数据及其派生数据、地形数据和气象数据进行栅格空间叠加分析 ,从而得到各样地的样地数据、遥感数据及其派生数据、地形数据和气象数据。再次 ,根据各样地优势树种所属的龄组将所有?

Correlation analysis of LANDSAT TM data and its derived data, meteoro logical data and topographic data with the biomass of tropical forest vegetation of different ages (Young, middle-aged, near mature, mature and over mature fore sts) were explored in Xishuangbanna, Yunnan Province, China. The analysis includ e d four steps. First, the biomass of each forest was determined from field invent ory data and a GIS database based on the geo-coordinates of each for est sample site. Second, the LANDSAT TM images were geometrically corrected usin g topographic maps. The derived data were derived from the LANDSAT TM ima ges using principal component analysis, tasseled cap transform and vegetation index analysis. Third, the data, including LANDSAT TM data and its derived data, to pographic data, such as DEM and aspect, and climatic data, such as annual av erag e temperature, annual average accumulative temperature above zero degree, annua l average precipitation and humidity, were referenced to the same projection and coordinate system and interpolated across a grid at a resolution of 30 meter. The LANDSAT TM data and its derived data, the topographic data and the cl imat ic data for the samples were achieved by using overlay analysis. Fourth, all of the data were overlaid onto the different-aged forests. Finally, correlations am ong the LANDSAT TM and its derived data, meteorological data, topographical data and forest biomass were analyzed for each forest type. Our results are as follo ws: 1) Correlations between the biomass of the young forest and LANDSAT TM1 and LANDSAT TM6 were significant at the 0.05 level and both correlations reache d - 0.33. 2) The correlation between the biomass of the middle-aged forest and preci pitation was significant at the 0.05 level and reached 0.33. 3) The correlation s between the biomass of the near mature forest and VI3, TM4 and Bright Index we re significant at the level 0.05 and were 0.50, -0.45 and -0.45, respectively. 4) Correlations of the biomass of mature and over mature forest and the second p rincipal component, which were significant at the 0.05 level, reached -0.46. 5) The highest correlation existed between the near mature forest biomass and VI3.