Annual maps of rural settlements in Beijing-Tianjin-Hebei region, China (1990 to 2020)

Using the Landsat-derived impervious surface area (ISA) data, the prolonged artificial Nighttime-light (NTL) data, and the Point of Interest (POI) data on the Google Earth Engine, we built the first annual maps of rural settlements (RUS) in the rapid urbanization region of Beijing-Tianjin-Hebei (BTH) in North China during 1990-2020. The RUS mapping model proposed in this study on the basis of referring to similar studies consisted of three basic steps, including kernel density estimation, initial boundary delineation, and post-processing. The location-based and the area-based comparison verified the effectiveness of our results, with a mean overall accuracy of 85% and a mean correlation value of 0.88, respectively.本研究依托谷歌地球引擎（Google Earth Engine）平台，采用基于陆地卫星（Landsat）反演的不透水面面积（impervious surface area，ISA）数据、长时间序列人工夜间灯光（Nighttime-light，NTL）数据与兴趣点（Point of Interest，POI）数据，构建了1990-2020年华北快速城市化区域京津冀（Beijing-Tianjin-Hebei，BTH）的首套乡村居民点（rural settlements，RUS）年度分布图集。本研究提出的RUS制图模型参考同类研究成果，包含三项核心步骤：核密度估计、初始边界划定与后处理。通过基于空间位置与地表面积的双重验证方法对研究结果进行有效性检验，结果显示其平均总体精度达85%，平均相关系数为0.88。