Studying Surface Urban Heat Island Phenomenon Using Remote Sensing in Three Metropolitan Areas of Texas, USA
MetadataShow full metadata
The goal of this dissertation was to investigate the Surface Urban Heat Island (SUHI) phenomenon for three metropolitan areas of Texas, USA with remote sensing techniques. A GIS-based Local Climate Zones (LCZs) classification scheme was developed with the aid of airborne Lidar datasets and other freely available GIS data, to map and compare the LCZs for the three metropolitan areas: Dallas-Fort Worth (DFW), Austin, and San Antonio. A decision-making algorithm was built for LCZs mapping, and LCZs datasets were established. By linking remotely sensed land surface temperature (LST) with LCZs, the study investigated the ability of LCZs for studying SUHI phenomenon and analyzes how different LCZs affect the SUHI in three major metropolitan areas. Landsat 8 image data was acquired for July 20, 2015 and used to calculate LST as SUHI measurement. Results indicated that large LST variations were first demonstrated among LCZs characterized by different land cover, and then urban morphological information. The close association between LCZs and LST demonstrated that the LCZs mapping was useful for comparing and investigating the SUHI. The geographically weighted regression (GWR) efficiently and accurately explained the underlying factors that contributed to the SUHI based on spatial variation and thus demonstrates improved utility for characterizing SUHI compared to global regression.