Urban heat islands and heat health warning systems in Mediterranean cities. Results from the Life ASTI project

Abstract

BACKGROUND AND AIM: In the LIFE ASTI project a forecasting system for estimating the urban heat island intensity (UHI) and defining high resolution heat health watch warning systems was developed in 3 Mediterranean cities, namely Thessaloniki, Rome and Heraklion. METHODS: Mortality data and temperature data for each city were collected for the period 2013-2018. A city-specific Poisson regression model was fit to define the association between mortality and maximum apparent temperature (Tappmax) considering the following explicative variables: holidays, month (May–August), the interaction between Tappmax and month and the number of consecutive hot days above the cut-off (mean Tappmax value corresponding to all days for which excess mortality was greater than 10%). For each value of Tappmax, the model then estimates the daily excess mortality (difference between observed daily values and a baseline reference) and defines monthly thresholds and warnings levels (1,2,3). Using the LIFE ASTI high-resolution urban forecasting system (spatial resolution 250m and lead time 5 days) a differential warning is issued for each zone within the city. RESULTS:UHI forecast maps and heat warnings are issued for each city together with specific health advice to users through the Life Asti web APP (https://app.lifeasti.eu/). Forecast and observed warnings days for summer 2019-2020 were compared and associated with observed mortality data. Preliminary results on summer 2019-20 showed a slight underestimation of Tappmax values but overall a good accordance of warning days. During heat wave periods increases in mortality were observed in both Rome and Thessaloniki. The model has been adjusted and results for summer 2021 will be presented comparing forecast data with observed data from the LIFE ASTI high resolution monitoring network and mortality data. CONCLUSIONS:High resolution warnings will help activate differential public health response measures within a city and reduce heat-related effects. KEYWORDS: climate, temperature, modeling, exposures