Exploring driving force factors of building energy use and GHG emission using a spatio-temporal regression method

Building energy consumption and GHG emission have significant influence on urban sustainable development. However, research works that investigate the temporal information in energy prediction are limited. This study bridged these gaps and assessed the energy use intensity (EUI) and GHG intensity (GHGI) based on three aspects by considering spatio-temporal dimensions. A GTWR model capturing both spatial and temporal information was employed to assess ten driving forces towards buildings EUI and GHGI. A case study in Seattle was used to demonstrate the effectiveness and validate the proposed approach. Results indicate that (1) The temporal heterogeneity has significant influence on EUI and GHGI, where the value of R2 in EUI experienced a remarkable improvement of 21.82% in the GTWR model compared to the GWR model, and that of for GHGI has been increased 13.92% from GWR model to GTWR model; (2) Buildings with large GFA was found to positively impact the EUI and GHGI, while the population under poverty and the number of floors have a negative impact. The novelty of this study lies in establishing a comprehensive framework for predicting energy usage with spatiotemporal information and quantifying the impacts of the ten driving forces at a regional level.