Exploring Heat Variations in Columbia, South Carolina

As the leading cause of weather-related fatalities in the United States, the concept of heat as a hazard has been studied for decades, especially in the urban environment. Urban heat island research is limited by the fact that most investigate land surface temperature rather than air temperature and other variables that affect human comfort including humidity. This thesis utilizes a public network of weather stations in Richland County, South Carolina to investigate differences in heat between the city of Columbia, and more rural parts of the county. The study period was June, July, and August of 2022. Results indicate that through the summer average, there is a range of afternoon heat index deviations of 8.1°F. The afternoon hot spots lie in downtown Columbia, as well as the more rural southwestern portion of the county. Heat Index, temperature, relative humidity, and dew point deviations were investigated for both afternoon and morning. To understand what land use variables best predict these deviations, twelve variables were found including Normalized Difference Vegetation Index (NDVI) and percent impervious surface each at three scales. Results here indicate that afternoon values are difficult to predict, possibly a result of the 10- meter station height with afternoon wind and vertical mixing. Land use variables performed better with morning heat variables, especially NDVI. Finally, multiple linear regressions were created to predict morning and afternoon heat variables, with more success in the morning accuracy. This project is a step towards creating local heat models based on air temperature and humidity rather than land surface temperature.