3 - UHI drivers and mapping the urban thermal environment

Abstract

The Kolkata metropolitan area is one of the subtropical urban environments experiencing swift and incessant urbanization process that has resulted in significant energetic differentiation in rural–urban domain as manifested in urban heat island (UHI) effect that has persuaded urban climate, biological environment, and socioeconomic aspects of urban society. The preparation of model–based UHI zonation map is the foremost task for UHI risk assessment and hazard mitigation. The present study has adopted the city-scale modeling of UHI by following geographic information system (GIS)–based statistical approach for building UHI maps using remote sensing data and other ancillary data. Initially, the UHI inventory map with 350 random pixels was extracted from monowindow algorithm–derived land surface temperature (LST) map using e-cognition approach. As such, UHI locations in LST map were then split into a ratio of 70/30 for building the UHI models and model validations. Finally, a spatial database of socioeconomic, structural, and radiative drivers with nine UHI conditioning factors have been prepared, including population distribution, land use and land cover, building material, building height, building roof type, building roof reflectance, building age, building association, and road pavement, and these database were extracted from multispectral scanner, thematic mapper of Landsat images, Google Earth historical images, and OpenStreetMap along with intensive rapid visual field survey. The geospatial relationships between UHI inventory's pixel locations and nine conditioning thematic factors were recognized by using four GIS-based statistical models, i.e., analytical hierarchy process (AHP), two-class kernel logistic regression (KLR), support vector machines (SVMs), and spatial multicriteria evaluation (SMCE) model. These models were constructed based on training data set, and model-derived results have been validated and compared with the area under receiver operating characteristic curve, kappa index, and five different statistical evaluation measures to corroborate the differences in overall performance. The results of goodness of fit are 86%, 87%, 85%, and 89%, and corresponding prediction capabilities are 81%, 85%, 82% and 87% for AHP, KLR, SVM, and SMCE models, respectively. The statistical measures show that the SMCE model gives better performance overall and precise results than the AHP, KLR, and SVM models. The KLR and AHP models have produced to some extent better results than the SVM model in provisions of positive spatial prediction values. Hence, the study revealed that SMCE and KLR are the promising physical data mining approach to be considered to map the spatiality of UHI zones. Finally, the contribution of built-up areas to UHI phenomena has been quantified.