Mapping 10-m global impervious surface area (GISA-10m) using multi-source geospatial data.

Artificial impervious surface area (ISA) documents human footprints. Accurate, timely, and detailed ISA datasets are therefore essential for global climate change and urban planning. However, due to the lack of sufficient training samples and operational mapping methods, global ISA mapping at 10-m resolution is still lacking. To this end, we proposed a global ISA mapping method leveraging multi-source geospatial data. Based on the existing satellite-derived ISA maps and the crowdsourcing OpenStreetMap (OSM), 58 million training samples were extracted via a series of temporal, spatial, spectral, and geometric rules. Combined with over 2.7 million Sentinel optical and radar images on the Google Earth Engine, we produced the 10 m global ISA dataset (GISA-10m). Based on the test samples that are independent to the training set, GISA- 10m embraced an overall accuracy greater than 86%. In addition, the GISA-10m was comprehensively compared with the existing global ISA datasets, and the superiority of GISA-10m was demonstrated. It was found that China and the United States embraced the largest ISA and road area. The global rural ISA was 2.2 times that of urban while rural road area was 1.5 times larger than that of urban region. The global road area accounted for 14.2% of the global ISA, 57.9% of which was located in the top ten countries. Generally, the produced GISA-10m dataset and the proposed sampling and mapping method are able to achieve rapid and efficient global mapping, and have potential for detecting other land covers. It was also indicated that global ISA mapping can be improved by incorporating refined OSM data. GISA-10m can be used as a fundamental parameter for Earth system science, and provide valuable support for of urban planning and water cycle study. [ABSTRACT FROM AUTHOR]