Urban forests or urban savannas? Tailoring cooling strategies to spatial and temporal variability in urban environments

Global warming and urbanization have exacerbated urban heat island issues in cities, affecting the liveability and long-term health and well-being of its citizens. This study explored how green space type (trees versus grasses and shrubs), composition (relative cover) and configuration (aggregation and shape complexity) regulate diurnal air temperature variations. Using a spatially-dense citizen science weather station network in Leuven (Belgium), mean night temperature, mean day temperature, and diurnal temperature range were recorded during summer heat periods of 2022. Urban green space data were collected at two buffer scales (50 m and 250 m) and analysed using composition and configuration metrics. Multiple linear (mixed) models were used to investigate associations between urban green space and air temperature at different temporal scales and for specific locations within the city. Results revealed dynamic effects of urban landscape on air temperature. A 10 % increase in tree cover mitigated daytime warming by 0.11 °C, likely through shading and evapotranspiration. Aggregated trees further improved daytime cooling by 0.42 °C. A 10 % increase in grasses and shrubs offered significant night-time cooling up to 0.47 °C, probably due to high albedo and increased longwave radiation loss. Additionally, increasing the aggregation of grasses and shrubs by 10 % enhanced nocturnal cooling by 1.62 °C, while increasing the shape complexity of trees promoted night-time cooling by 0.44 °C. Further, this study highlighted the need for context-specific strategies. Conserving green locations as urban forests, while designing built-up grey locations as urban savannas, characterised by interconnected grasses and shrubs interspersed with standalone trees, could mitigate urban heat stress and enhance urban resilience.