As the impacts of climate change intensify, bringing an increase in the frequency and magnitude of heat waves, the interest around urban heat mitigation strategies is rapidly growing worldwide. Green roofs, defined as roofing systems that incorporate a vegetated layer, have been proved to reduce urban heat, thanks to their evaporative cooling and lower heat storage than conventional roofs. Thus, they are expected to become increasingly important in the future, given their potential to counteract the projected temperature increases associated with climate change.

Numerous studies emphasize the urban heat mitigation potential of green roofs, yet accurate quantifications of their temperature reductions under future climate are currently lacking. For instance, under climate change, higher temperatures and longer dry periods are expected in central Europe, conditions that can negatively affect green roofs. Recently, microclimate models are gaining traction in evaluating the efficacy of heat mitigation strategies, facilitating the quantification of urban heat reductions under various climate conditions. However, despite their increasing use in the literature, microclimate models are rarely combined with climate projections, due to the complexity of downscaling interdependent weather variables such as precipitation, air temperature and global horizontal radiation. Consequently, the heat reduction potential of green roofs under future climates is largely unexplored, particularly in comparison to their observed performance under current climate. Additionally, it is unknown whether specific roof parameters could contribute to further enhancing heat mitigation, such as plant characteristics, irrigation schemes, or substrate depth.

This study aims to investigate the heat mitigation potential under climate change on a green roof in Mendrisio, Switzerland (characterized by hot, dry summers) using an open source microclimate model developed by Meili et al. (2020), Urban Tethys-Chloris (UT&C). This model was selected because of the fully coupled energy and water balance, and the incorporation of plant-specific characteristics. Continuous year-long monitoring of the green roof enabled to collect surface temperature using infrared sensors. These measurements were used to calibrate and validate the microclimate model. To account for climate change, coupled, sub-hourly, future projections of precipitation, air temperature, solar radiation, relative humidity, and wind speed were used as input to the validated microclimate model. These projections were derived from a convection resolving climate model (COSMO forced by MPI-M-MPI-ESM-LR at RCP 8.5, worst-case emissions scenario) run over the European domain at a 2.2-km, 6-minute resolution for a 10-year period that was bias corrected through quantile mapping. Lastly, variations in key parameters like substrate depth, vegetation type, and green roof irrigation schemes were explored to analyze their impact on urban heat mitigation under climate change.

Preliminary, manual calibration of the microclimate model resulted in a good predictive ability (r² = 0.71), which will be further improved with automatic calibration. In a current climate, the green roof was able to reduce maximum surface temperatures in Summer by approximately 15°C, with respect to an adjacent concrete roof. Further expected results will evaluate potential temperatures reductions in a future climate and determine whether green roofs can counteract increasing temperatures by exploring a range of alternative designs.