Advanced earth observation methodologies for the study of the Thermal Environment of Cities

Satellite platforms carrying thermal infrared sensors provide continuous monitoring of the land surface temperature (LST) distribution at a spatial resolution ranging from 3–5km to ∼100m. A suite of advanced methodologies for the study of the thermal urban environment using satellite images is being developed at National Observatory of Athens (NOA) in collaboration with leading organizations. The first one is the downscaling of geostationary LST imagery down to 1km spatial resolution using advanced regression algorithms such as Support Vector Machines with Gradient Boosting. The second approach is the evaluation of heat wave hazard and risk based on the intensity, duration and time lag between consecutive events as well as the vulnerability of the population using fuzzy logic. Finally, as comparison of thermal patterns, most notably surface urban heat islands (SUHI), across different cities is hampered by the lack of an appropriate methodology to extract and characterize them, an object-based image analysis procedure has been developed allowing for the calculation of several features (e.g., SUHI area extent and intensity) from the original LST maps. A major challenge is now to examine whether the presented methodologies can serve as standardized approaches which will allow meaningful comparisons of the thermal characteristics between cities.