Examining Urban Heat Island Relations to Land Use and Air Pollution: Multiple Endmember Spectral Mixture Analysis for Thermal Remote Sensing

Abstract

This paper proposes an integration of Spectral Mixture Analysis and Endmember Remote Sensing Indices to derive land surface temperature (LST), to identify urban heat islands (UHI) and to investigate their relationships to land use/land cover (LULC) and air-pollution in Tabriz city, Iran. LST values are assessed from ASTER remotely sensed thermal images to appraise the intensity of UHIs through a statistical approach. The results indicate that LST is highly influenced by LULC and that UHIs are closely linked to LST and LULC. As expected, LST is sensitive to vegetation and moisture and low temperatures are found in water bodies and vegetated areas. High temperatures are related to construction zones and industrial sites which are not necessarily located in the city centre. While the results demonstrate a high correlation between UHIs and air-pollution spatial analysis discloses a different pattern: when correlating identified UHI zones with air pollution the market area of Tabriz city reveals the highest measured PM values.