High-resolution estimation of urban building energy consumption using SDGSAT-1: A hybrid spatial modeling framework

Rapid urbanization has intensified the spatial differentiation of building energy consumption within cities, creating a growing demand for high-resolution evidence to support urban energy planning and low-carbon governance. Urban building energy modeling has provided important methodological support for building energy assessment, retrofit analysis, and policy formulation, but its application in data-scarce contexts remains constrained by the limited availability of building-level energy records, detailed building attributes, and operational parameters. To address this challenge, this study develops a top-down spatial estimation framework for limited-data conditions by integrating nighttime light and land surface temperature observations from Sustainable Development Science Satellite 1 (SDGSAT-1), Building-stock data, building height information, and official energy statistics. Hefei, a prefecture-level city in eastern China with a permanent population of nearly 10 million, is used as a case study to demonstrate the framework.
The proposed approach generates a 100 m resolution spatial map of urban building operational energy consumption and further examines how planning-related urban form indicators are significantly associated with spatial variations in energy use. Geographically weighted regression is used for spatial downscaling, while geographically weighted correlation, multiscale geographically weighted regression, and zone-specific Random Forest models are used to explore spatially heterogeneous and nonlinear morphology–energy relationships. The results show that public buildings dominated Hefei's operational building energy consumption in 2022, accounting for approximately two-thirds of the total. Building energy use exhibited a clustered and polycentric pattern in the metropolitan core, whereas surrounding county-level areas generally showed lower-intensity and more monocentric distributions. Planning-related urban form indicators, particularly floor area ratio, building height, and building spacing, were significantly associated with spatial variations in estimated building energy consumption, with evident nonlinear and spatially heterogeneous relationships. These findings provide planning-relevant evidence for neighborhood-scale, form-sensitive energy assessment and low-carbon planning in data-constrained cities.
Copyright © 2026. Published by Elsevier B.V.