Spatial statistical analysis and simulation of the urban heat island in high-density central cities

Abstract

The urban heat island (UHI) is a mounting problem in built-up areas, leading to increased temperatures, air pollution, and energy consumption. This paper explores the urban determinants of the UHI, using two-dimensional (2-D) and three-dimensional (3-D) urban information as input to spatial statistical models. The research involves: (a) estimating land surface temperatures, using satellite images, (b) developing a 3-D city model with LiDAR data, (c) generating urban parameters with 2-D/3-D geospatial information, and (d) conducting spatial regression analyses. The data are captured over three grids of square cells – 480m, 240m, and 120m – and characterize the center of Columbus, Ohio. The results show that solar radiations, open spaces, vegetation, building roof-top areas, and water strongly impact surface temperatures, and that spatial regressions are necessary to capture neighboring effects. The best regression is obtained with the general spatial model (GSM), which is then used to simulate the temperature effects of different greening scenarios (green roofs, greened parking and vacant lots, vegetation densification) in the center of Columbus. The results demonstrate the potential of such models to mitigate the UHI through design and land-use policies.