Multiscale Analysis of Census-Based Land Surface Temperature Variations and Determinants in Indianapolis, United States

Abstract

This study utilized remote sensing and geographic information system data to perform a multiple scale analysis of urban heat island (UHI) at Indianapolis, United States. Land surface temperature (LST) data derived from a Landsat Enhanced Thematic Mapper Plus image of Indianapolis city were used to examine census-based variations and to model their numerical relationships with its urban landscape. Selected variables for the urban landscape included building, roads, water bodies, vegetation, and open space. Correlation analysis and linear regression modeling at each census level (i.e., block, block group, and tract) were performed. The sensitivity of the relationship to aggregation and thus the scale effect were examined. The impact of the provision of different urban landscape features on the UHI pattern at three census levels was also examined. Results show LST had strong positive correlations with buildings and roads, but was greatly negatively correlated with vegetation and water at all levels. Overall, the larger the analytical unit, the stronger the correlation between LST and the urban landscape parameters, and the better the LST modeling. The results of this empirical study establish a method for evaluating urban thermal environment within geographical (census) units.