IDENTIFYING THE LOCAL SURFACE URBAN HEAT ISLAND THROUGH THE MORPHOLOGY OF THE LAND SURFACE TEMPERATURE

Abstract. Current characterization of the Land Surface Temperature (LST) at city scale insufficiently supports efficient mitigations and adaptations of the Surface Urban Heat Island (SUHI) at local scale. This research intends to delineate the LST variation at local scale where mitigations and adaptations are more feasible. At the local scale, the research helps to identify the local SUHI (LSUHI) at different levels. The concept complies with the planning and design conventions that urban problems are treated with respect to hierarchies or priorities. Technically, the MODerate-resolution Imaging Spectroradiometer satellite image products are used. The continuous and smooth latent LST is first recovered from the raw images. The Multi-Scale Shape Index (MSSI) is then applied to the latent LST to extract morphological indicators. The local scale variation of the LST is quantified by the indicators such that the LSUHI can be identified morphologically. The results are promising. It can potentially be extended to investigate the temporal dynamics of the LST and LSUHI. This research serves to the application of remote sensing, pattern analysis, urban microclimate study, and urban planning at least at 2 levels: (1) it extends the understanding of the SUHI to the local scale, and (2) the characterization at local scale facilitates problem identification and support mitigations and adaptations more efficiently.