基于MODIS及多源地理信息的全球250米全球城区范围产品MGUP(2001-2018)
收藏资源简介:
提出一种包含多源样本细化、局部自适应建模和时空后处理的自动化的全球城市样本提取方案。生成 2001 年至 2018 年的 250 米全球城市范围数据集MGUP(MODIS global urban extent product)。经精度验证,MGUP 在不同年份、不同级别城市和不同大洲取得了较好结果,全球 F-Score达88%。产品间的相互比较展示了MGUP更优的精度,因此可被视为现有基于MODIS城市产品的改进版本,是监测全球城市扩张的可靠数据源。
We propose an automated global urban sample extraction scheme that includes multi-source sample refinement, local adaptive modeling, and spatiotemporal post-processing. We generate the 250-meter global urban extent dataset MGUP (MODIS Global Urban Extent Product) spanning from 2001 to 2018. Through accuracy validation, MGUP achieves favorable performance across different years, urban tiers, and continents, with a global F-Score of 88%. Inter-product comparisons demonstrate the superior accuracy of MGUP, making it a refined version of existing MODIS-based urban products and a reliable data source for monitoring global urban expansion.




