Mapping Urban Heat with Community Science, Machine Learning, and Remote Sensing
Hot summers and heatwaves are deadly, killing more Americans than any other natural hazard, on average, and sending many more to the emergency room. Regionally in the US, urban areas contain the hottest temperatures in comparison to their surrounding countryside because of a phenomenon known as the Urban Heat Island (UHI) effect. This effect, which occurs as a result of materials used in constructing cities mediate the absorption and dissipation of the sun's radiation differently than surrounding non-built areas, and can drive temperature differences across urban-rural gradients of 20°F and more. More recently, researchers are mapping intra-urban variation in temperatures in an effort to understand the mechanisms that produce the hottest areas, which in turn, can help to inform potential urban planning and policies, and reduce exposure to extreme heat, especially among those most health-vulnerable communities.While many methods exist to observe, model, and map urban heat, they differ in terms of the techniques employed, the data they capture, their transferability, and, ultimately, the interventions they can inform. In this webinar, UHI measuring and modeling expert Dr. Vivek Shandas will tease these varying approaches apart, and provide an overview of a relatively new machine learning method that incorporates conventional satellite remote sensing data, and in situ observations of temperature and humidity from community science urban field campaigns [see recent paper here: https://doi.org/10.3390/cli7010005]. This presentation will cover the methods applied, results from past campaigns, and lessons learned over the 20 different urban field campaigns conducted since 2015. The presentation will also touch on the emerging plans for 2020 and how outcomes from previous campaigns are helping to inform its design. Applications of the UHI maps to intervention analysis and city planning and policy are already underway.