CLIM4cities - from Citizen Science, Machine Learning and Earth Observation towards Urban Climate Services

As climate change intensifies, urban areas are increasingly exposed to more frequent, severe and longer-lasting temperature extremes, particularly heatwaves. This growing thermal amplitude represents a major challenge for highly urbanised and ageing societies, with direct consequences for public health, energy systems and social equity. Cities are especially vulnerable due to the Urban Heat Island effect, whereby land cover characteristics, urban morphology and reduced vegetation cover amplify thermal stress. Despite this vulnerability, effective local adaptation remains constrained by the limited availability of high-resolution operational air temperature data, to support early warning systems, urban planning, and scenario-based assessments. 

CLIM4cities is a European Space Agency (ESA)-funded project under the Artificial Intelligence Trustworthy Applications for Climate programme that applies Machine Learning (ML) techniques to downscale near-surface air temperature (T2m) and land surface temperature (LST) in urban environments. By integrating numerical weather prediction outputs, Earth Observation data, and quality-controlled crowdsourced observations, CLIM4cities provides sub-kilometric urban temperature information tailored to local decision-making needs. The project constitutes a key step towards the development of cost-effective Urban Climate and Weather components that are interoperable with local Digital Twin systems. 

During its first phase, CLIM4cities developed and evaluated coupled ML-based downscaling models for T2m and LST across four Danish metropolitan areas (e.g. Aalborg, Arhus, Odense and Kobenhavn), demonstrating the feasibility and transferability of the proposed approach. For LST, Sentinel-3 thermal observations and vegetation-related predictors were employed within a scale-invariance downscaling approach, with independent validation using Landsat 8/9 data. Results show that while non-linear ML models can enhance predictive skill at coarser spatial scales, their performance at finer resolutions is limited by the breakdown of scale-invariance assumptions. Incorporating residual correction proved essential to recover fine-scale variability, whereas timestamp-specific linear models often outperformed more complex ML architectures. Model performance exhibits strong seasonal dependence, with the highest score achieved in summer (R² ≈ 0.75), when reduced cloudiness and drier conditions enhance the representation of urban thermal patterns.  

In contrast, T2m downscaling achieved its highest skill using comparatively simpler modelling approaches. Random Forest models consistently performed well across both spatial and temporal evaluation datasets, increased model complexity did not yield substantial gains. Model performance was assessed under average conditions as well as during heatwave and cold-wave events, complemented by sensitivity analyses of key hyperparameters. The results indicate an R² of 0.98 under average conditions, remaining stable during heatwaves and decreasing marginally to 0.97 during cold events. Mean absolute errors below 1K across all subsets confirm the robustness and operational suitability of the approach for monitoring urban-scale atmospheric temperature variability. 

Building on these results, the ongoing CLIM4cities project extension focuses on replicating and validating the T2m ML framework across additional European metropolitan regions spanning diverse climatic and urban contexts. Case studies include Copenhagen, Athens, Seville, and Lisbon, enabling a systematic evaluation of model behaviour across climate zones.