Speaker: Jia Hu, Iowa State University

Title: Urban Heat Island Intensity Estimation Using Remote Sensing and Statistical Methods

Abstract: Urban heat island (UHI) is a phenomenon with temperature in urban areas higher than surrounding rural areas. Land surface temperature (LST) from thermal remote sensing images has been widely used due to periodic observations and high spatial coverage of satellite images over large scale and time-series. Accurately estimating UHI intensity is essential in urban thermal environmental studies. In our previous study, we used Gaussian model and satellite LST to quantify the UHI intensity. However, the Gaussian model is not suitable for the UHI estimation in multi-centric cities or cities with complex LST distribution. To solve such a problem, we currently proposed an improved method for UHI intensity estimation by using multiple source remote sensing data and statistic methods. In this method, we considered multiple affecting factors on LST, including morphology, land cover, and the distance to waterbodies in each city. We applied this method to 245 big cities in the United States and compared the estimated UHI intensity with those estimated from Gaussian model and conventional temperature dichotomy method. Results showed that the new method had good performance in estimating UHI and had more universality across national level cities compared to the other two methods.