A combination of methods for mapping heat and cool areas in past and current urban landscapes of Poitiers (France)

•Two mapping methods are combined to locate precisely urban heat areas.•The number of land use categories can enhance the heat mitigation index and the localisation of heat areas.•Spatio-temporal evolution of urban heat areas between 1993 and 2020.•Results provide valuable tools for finding optimal solutions for sustainable urban development. As a result of human activity, living organisms are faced with the consequences of both climate and landscape changes. These disturbances are particularly significant in urban environments, where an increase of areas are experiencing a rise in temperatures, creating urban heat islands (UHIs). Knowing the exact location of these areas and the factors involved in their formation would help to guide management measures aimed at reducing their impact. To locate heat and cool areas, this study uses, compares and combines satellite images (i.e. Land Surface Temperature = LST) and spatial modelling of an Heat Mitigation Index (i.e. HMI) in the urban landscape of Poitiers (France). We highlighted that the LST value differs according to the land use category. Indeed, while the highest temperatures were observed for high building density, moist tree vegetation and water areas are rather associated with the lowest temperatures. The results showed that the LST values correlate with the spatial modelling of HMI. Moreover, this correlation increases with the precision of the land cover (i.e. the number of land cover categories taken into account). While LST provides contextual information about heat, the HMI reflects the perceived heat according to the land cover. Thus, HMI modelling therefore showed that it appeared to be better suited to studying variations in highly heterogeneous landscapes such as urban landscapes and appear as a very interesting alternative to LST data when they are not available. Our approach is reinforced and corroborated by the installation of thermometers in the field. In addition, taking advantage of our improved HSI, we calculated the HSI and LST for two dates in 2020 and 1993. This demonstrated (i) the applicability of our method in the analysis of recent and past images and (ii) the contribution of our method to the study of the spatio-temporal evolution of heat and cool areas between two dates. The results of this study could help and guide future local urban planning in order to improve the mitigation and cooling potential of UHIs in the cities of tomorrow.