Relating Urban Morphology and Urban Heat Island During Extreme Heat Events in the Kansas City Metropolitan Area

Abstract

Satellite images offer continuous spatial and temporal coverage of surface temperature, thus allowing us to transcend limitations of in-situ measurements and giving us a powerful tool to understand the Urban Heat Island (UHI). This study uses MODIS Land Surface Temperature (LST) at 1 km to examine the relationship between urban morphology and UHI. Using the Kansas City metropolitan area as a case study, we examined MODIS LST data during the summer months of 2002-2017. We found that LST anomalies increase exponentially from 1.5°C at the 90-95 percentile to 3.5°C at the 95-100 percentile. In particular, natural land cover LCZ classes are found to have higher LST anomalies than built-type LCZ classes by up to 5°C. Results suggest that the higher LST anomalies over outlying areas are not due to suburban development during the most extreme heat episodes. We also examined the utility of the LCZ scheme during extreme heat events. We found that the LST response is not statistically different between the various LCZ classes and that the local built environment is not as important in predicting the LST response to increasingly extreme heat events. However, the LST response by most LCZ classes as a function of distance from downtown is statistically significant, with values ranging from -0.08°C/km to -0.01°C/km. The results show that the distance from the city center plays a more important role in helping predict LST response than knowledge of the LCZ class.