Crowdsourced observation networks have significant potential to enhance our understanding of urban climates. Events such as the July 2022 UK heatwave, which featured the first ever red weather warning for extreme heat and over 2000 excess deaths across the period, underline the importance of accurate and detailed information regarding heat hazard. Such hazards are exacerbated in urban areas, both in their intensity and in the exposure of greater numbers of people.
Yet, existing services often lack dense, urban-scale observations at their foundation. Current climatological baselines likely underestimate both the intensity and spatial extent of urban heat. Such underestimates are expected given that our existing weather station network fails to sample urban areas as effectively; instead providing observations more representative of rural environments. Crowdsourced networks – meteorological sensors owned and operated by citizens or third parties – can offer denser and more geographically diverse data, presenting an opportunity to construct products and services that better reflect lived experiences, especially in urban areas.
This project aims to leverage crowdsourced observations to improve forecasting of conditions conducive to extreme heat hazard, forming the basis of a future warning service. In addition to the improvements in spatial resolution, the proposal would seek to leverage the near-real-time data provided by crowdsourced weather stations to both understand the temporal evolution of urban heat risk and produce more timely and responsive forecasts.