A comparison between London and Baghdad surface urban heat islands and possible engineering mitigation solutions

Abstract This study adopts remote sensing techniques to compare the Surface urban Heat Island (SUHI) in Baghdad and London as they represent different climatic conditions, natural environments and levels of urban development. It tests the reported correlation of land surface temperature (LST) with land cover in the literature under different conditions and, based on the findings, suggests engineering mitigation strategies for each city. The land surface was characterized using supervised classification and spectral indices, using the Landsat 8 optical bands (2–7), and the LST was retrieved from Landsat’s thermal band 10 after emissivity calibration. Two Landsat 8 satellite images were used, acquired in July 2013 when maximum surface temperature would be expected in both these capital cities. Image processing included radiometric calibration and atmospheric correction and various land surface indices were then calculated. The independent validation of land cover types was performed using higher spatial resolution optical data, and LST patterns were validated using ASTER thermal images. Land cover types or indices and land surface temperature display high correlations, with most having a positive relationship with LST, but vegetation has a negative relationship. The hottest surface type also differs for the two cities. Consequently, covering the soil in Baghdad with new construction, for example, reduces the surface temperature and hence urban heat island effect, while the same action in London increases it. Thus, engineering solutions to urban heat island issues need to take local factors into account.