2种常用的全球土地利用/覆被历史数据集在中国区域的精度评价

The accuracy evaluation of two common global historical land use/cover datasets in China

【目的】分析全球土地利用/覆被历史数据集在中国区域的精度。【方法】分别搜集耕地、草地的统计资料和空间数据以及林地的统计资料,采用升尺度方法对耕地和草地的空间数据进行处理,分析全球土地利用/覆被历史数据集SAGE和HYDE 3.1与我国统计资料中耕地、林地和草地面积的差异。【结果】SAGE数据集中耕地、草地和林地面积与我国统计数据相差较大,HYDE 3.1数据集中耕地和草地面积与我国统计数据虽然相近,但其反映的耕地和草地面积变化趋势以及空间分布与我国实际情况存在较大差距。【结论】鉴于全球土地利用/覆被历史数据集SAGE和HYDE 3.1在我国的精度较低,因此有必要依据国内的历史统计资料,重建一套适用于中国的高精度土地利用数据集,以服务于我国气候变化研究。

The aim of this research was to analyze the accuracy of global historical land use/cover datasets in China.【Method】Based on collected crop,pasture and forest statistical and spatial data during 1900-2000 and spatial up-scaling method,the accuracy of SAGE and HYDE 3.1 global historical land use/cover historical dataset was compared.【Result】Crop,pasture and forest areas derived from SAGE dataset were largely different from domestic data.Although crop and pasture areas derived from HYDE 3.1 dataset were more accurate than SAGE dataset,the spatial distribution of crop and pasture land was largely different from 1km land use map of China.【Conclusion】Producing higher resolution historical land use/ cover datasets using domestic data is necessary for climate change study in China.