Assessing Excessive Heat-Related Risk in European Cities: A Multi-Criteria Approach for Medium-Sized Cities

Urbanization intensifies the Urban Heat Island effect, exposing city dwellers to higher temperatures and increased heat stress risks, particularly during summer. This study proposes a comprehensive approach to assess heat-related risk in three European cities sampled from different European climate conditions: Turku (Finland), Jena (Germany), and Iai (Romania).A mixed-methods approach combining geospatial analysis and statistics will be employed. Our aim is to create fine-scaled (at least 100 meters resolution) heat risk maps for each city and test whether rental data may be used as a proxy of the vulnerability variables in risk assessment. Rental data can be a valuable proxy for socioeconomic vulnerability in the context of heat stress. Higher rents often correspond to more modern buildings with better insulation and air conditioning, which can offer greater protection from extreme heat. Further we posit a correlation between rental prices and the vicinity of cooling structures such as green spaces and water bodies. Additionally, higher rents may be associated with better access to transport infrastructure, implying a higher connectivity to essential services like healthcare facilities or public buildings with air conditioning. This combined effect suggests that residents in higher-rent areas may have a greater ability to cope with extreme heat compared to those in lower-rent areas.To ensure more relevance for current climate conditions, daily temperature data was acquired from meteorological agencies or research institutions for each city covering the period 2014-2023. We calculated the hazard layer, defined as the sum of temperature days exceeding 20C for each location over the ten-year period. Population density data was obtained from Eurostat in a 1 km2 grid. To achieve a finer scale, this data was down-scaled using a building-volume-based method, providing a more accurate representation of population distribution within each city. Web scraping was used to collect rental data from online platforms, including various attributes. This data will serve as a proxy for socioeconomic vulnerability in the context of heat stress. We acquired data on established socioeconomic vulnerability indicators like income, education, and occupation for each city. Universal kriging is used to interpolate the hazard- and vulnerability layers to account for spatial autocorrelation and generate a continuous surface for each variable. This allows for a more nuanced analysis of the spatial relationships between these factors. The interpolated layers will be integrated into a spatial model to assess the risk of heat-related stress across each city, highlighting areas with high vulnerability. Statistical analysis is conducted to assess the correlation between rental data and established socioeconomic vulnerability indicators. This will validate the use of rental data as a vulnerability proxy and provide further insights into the sociodemographic makeup of vulnerable populations.