中国区域多源土地覆被遥感产品精度分析与验证

Precision analysis and validation of multi-sources landcover products derived from remote sensing in China

国家及区域尺度的土地覆被信息对于解决环境演变、生物多样性保护、生态系统评价及环境建模等一系列问题起着至关重要的作用。该文以中国科学院CHINA2000数据为参考,从国家尺度比较了当前全球4种土地覆被遥感产品的分类精度,以便解释验证数据与参考数据在空间及专题上的一致性和异质性。结果表明:与参考数据相比,4种遥感产品的土地覆被类别的空间分布,从整体上表现出较高的一致性,但在局部区域差异性较大,特别是在景观异质性较强的西南和东南地区;GLC2000遥感产品从相关系数和总体精度方面都表现出与参考数据较好的一致性,其相关系数为0.92,总体精度为55.86%,GLOBCOVER产品则表现出与参考数据较差的一致性;4种产品与参考数据在空间上表现出明显的混淆现象,混淆主要发生于林地、灌木、草地和耕地之间,这表明粗空间分辨率遥感产品在识别叶类土地覆被类型的能力上仍需要改进,该文为中国用户合理利用遥感产品提供了科学合理的依据。

Precise state and regional land cover mapping are foundational data in the studies of land surface processes and modeling, such as climate change, biodiversity conservation, ecosystem assessment, and environmental modeling. This study aimed to perform the validation and comparison of the four global land cover datasets, and examine the suitability and accuracy of different coarse spatial resolution datasets in mapping land cover categories across China. The results indicated that the four land cover products demonstrated the consistency in whole scale, but in local scale, display a conspicuous difference especially in southeast and southwest zones. GLC2000 had the best consistency in the relative index and overall accuracy, which were 0.92 and 55.86% respectively; the GLOBCOVER had the worst consistency compared to the reference data , the relative index and overall accuracy were 44.71% and 0.31. The four land cover products showed the spatial confusion especially in forest, shrub, cropland and grassland, which implied the coarse resolution remote sensing data was helpless in discerning the land cover with the leaf type. The result will provide a reasonable reference for using these datas.