Evaluating the Performance of SDGSAT-1 GLI Data in Urban Built-Up Area Extraction From the Perspective of Urban Morphology and City Scale: A Case Study of 15 Cities in China

The SDGSAT-1 satellite, independently developed and launched by China, features a higher spatial resolution in its Glimmer Imager (GLI) data than the currently widely used nighttime light (NTL) remote sensing data, making it possible for high-precision extraction of urban built-up areas (UBAs). However, recent empirical studies have revealed that the higher resolution of the NTL data does not inherently guarantee enhanced accuracy of the UBA extraction, as urban morphology and city scale significantly influence the performance. Therefore, comprehensively investigating the UBA extraction performance of NTL data across cities with different morphologies and scales offers valuable guidance for optimizing urban remote sensing methodologies in diverse city contexts. Across 15 representative cities with various urban characteristics in China, this study conducted a comprehensive evaluation of the UBA extraction performance of SDGSAT-1 GLI data for the first time, with the NPP-VIIRS NTL data for comparison, and discussed how urban morphology and city scale affect the performance in UBA extraction. The results indicated that (1) In cities with different morphologies, the UBA extraction performance of SDGSAT-1 data follows from high to low: striped cities, clustered cities, and agglomerate cities. In cities with different scales, the UBA extraction performance of SDGSAT-1 data is megacities > super cities > medium cities > small cities > large cities. (2) Compared with NPP-VIIRS data, SDGSAT-1's performance in UBA extraction is not always superior and varies with the urban morphology and city scale as well as the extraction method, which necessitates context-sensitive selection for NTL data and method in UBA extraction.