Spatial interaction model of energy demand of buildings and satellite thermal imageries using Geographically Weighted Regression analysis

The Energy Performance Certificate (EPC) is an important information tool to improve the energy performance (EP) of buildings. However, establishing the EP of building is tedious, time-consuming, and numerous input parameters are required in its estimation. However, the usefulness of EPC for the implementation of customized solutions by the supply-side actors require that EPCs are available for all buildings, easily accessible, credible, and recent. However, this is not the case at present. This could be addressed by employing remote sensing dataset along with GIS based spatial analysis techniques. In the present study, the spatial regression analysis technique is implemented in identifying the spatial relation between the input variables and the EP of selected 4541 buildings within Växjö municipality, Sweden. The input variables used in the study include the land surface temperature (LST) maps of summer and spring of 2020 derived through the thermal band of Landsat 8 satellite data, built-up and openland neighbourhood maps prepared from the land use/land cover map 2020 of the study region. Building topology including year of construction, type, category, and complexity of buildings are also used to identify the relation between the input variables and the EP of those selected buildings. Results of spatial regression analysis reveal a significant positive relation between the LST and EP of buildings (regression co-efficient are 0.86 and 0.95 in spring and summer respectively). The stronger correlation in summer could be because of the availability of higher intensity of solar radiation which gets absorbed by the built-up regions. Results suggest that the LST maps derived from satellite imageries could provide information on the EP of buildings. This could be beneficial to local decision makers and policy regulators in identifying the buildings with lower EP with better accuracy with less dependence on EPC data which are sometimes not available or not updated. The results could also be beneficial to investment bankers, real estate companies during the purchase and sale of a building. Policy makers and renovation companies could get benefited with the results in preliminary identification of the potential hotspots for district energy renovation where the EP of buildings is poorer. This could help achieve the goal of sustainable urban planning targeting energy reduction, climate adaptation, through implementation of effective energy management strategies in t