区域森林碳分布空间估计误差定量分析

A Quantitative Analysis on Spatial Estimation Error of Regional Aboveground Forest Carbon Distribution

  • 摘要: 区域森林碳(地上部分)分布的空间估计存在多种误差来源,其直接影响估计结果的精度。采用森林资源清查样地数据与Landsat TM影像数据相结合的方法对区域森林碳(地上部分)分布进行空间估计的误差来源,采取相对误差的形式对其定量化,根据不确定度的合成与分配方法获得空间估计总误差及各误差来源在总误差中的比例分配。基于临安市2004年森林资源清查和Landsat TM影像数据的森林碳(地上部分)分布空间估计误差分析结果显示,区域森林碳(地上部分)分布空间估计的总误差为10.15%,各误差所占比例为:抽样误差63.5%、遥感影像坐标校正定位误差22.9%、树高测量误差9.4%、生物量模型误差3.4%、胸径测量误差0.8%。

     

    Abstract: There are several error sources in the spatial estimation process of regional aboveground forest carbon distribution, which directly impact the precision of the estimation results. This paper addressed the error sources by means of combining NFI plot data with Landsat Thematic Mapper images to estimate the regional aboveground forest carbon distribution. According to the uncertainty synthesis and distribution method to obtain the total error of spatial estimation and the proportions of various error sources which were quantified by the relative error. The error analyses on spatial estimation of regional aboveground forest carbon distribution based on 2004 NFI plot data and Landsat Thematic Mapper images from Lin′an Municipality, Zhejiang Province showed that the total error on spatial estimation of regional aboveground forest carbon distribution was 10.15% The proportions of each error were sampling error (63.5%), locating error of remote images coordination correction (229%), tree height measuring error (9.4%), biomass model error (3.4%) and DBH measurement error (0.8%).

     

/

返回文章
返回