#ModellingMonday – Modelling Urban Heat Stress and Adaptation: 9 Models for a more Liveable Tallinn

Photo: Tallinna pildipank. 

Tallinn, Estonia’s fast-growing capital, is increasingly facing challenges from urban heat and climate extremes. Although cold-related health risks still outweigh heat risks under current conditions, the city is proactively preparing for a hotter future. As part of the Horizon Europe project KNOWING, Tallinn serves as a demonstrator region to explore climate mitigation pathways and assess the impacts of adaptation interventions. A suite of advanced climate and impact models is being employed to simulate urban heat exposure, health effects, and the cascading implications for energy, transport, and infrastructure. 

 High-Resolution Urban Climate Insights with PALM-4U 

Figure 1: Simulations of Tallinn’s max temperatures for 2050 showing temperatures up to 32,6 degrees. Credit: AIT 

 At the core of the heat-related modelling effort lies PALM-4U, a high-resolution urban climate model capable of simulating atmospheric and thermal processes down to a 10-meter grid. In Tallinn, PALM-4U runs four simulations with identical meteorological inputs but different land use configurations, corresponding to the baseline year, and future years 2030, 2040, and 2050. Each scenario incorporates incremental adaptation measures, from building insulation and tree planting to street greening and large-scale renaturation of moorlands. These interventions are drawn from strategic plans like Tallinn’s Sustainable Energy and Climate Adaptation Plan 2030, its Development Strategy 2035, and Bog Renaturalisation Plan. PALM-4U provides spatially resolved outputs such as 2-meter air temperature, heat stress indicators (UTCI – Universal Thermal Climate Index), and urban heat island (UHI) patterns, supporting assessments of cooling effects and thermal comfort. 

Figure 2: The same simulation showing trees. The expansion of green and blue infrastructure is the city administration’s central strategy for a more livable Tallinn. Credit: AIT 

 Linking Urban Heat to Public Health: D-MERF 

 To understand the human health implications of these climatic changes, D-MERF (Delayed-Multivariate Exposure-Response Functions) is used. This epidemiological model estimates temperature-related mortality and the effects of adaptation measures on public health outcomes. Even though the overall mortality in Tallinn is more influenced by cold, increasing heat waves and higher temperatures during summers are expected. The greening measures are directly aimed at reducing the urban heat island effect (UHI effect) and improving thermal comfort for citizens. 

 Regional Climate Input from WRF 

 Providing the broader climatic context, the WRF (Weather Research and Forecasting) model supplies regional climate data for Tallinn at a 5×5 km resolution. These simulations are driven by CMIP6 scenarios and Shared Socioeconomic Pathways (SSPs), offering long-term meteorological data, including temperature and wind, which are essential for both PALM-4U and downstream models like D-MERF and HWLEM. 

 Health and Economic Risks During Heatwaves: HWLEM 

 To bridge the scale between regional climate projections and human health outcomes, HWLEM (Heat Wave Local Effect Model) is applied for Tallinn. This model integrates regional forecasts with local urban morphology to estimate heatwave-driven impacts, such as increased hospitalisation costs and mortality. Using PALM-4U-generated UTCI hazard maps, HWLEM runs four future scenarios aligned with the city’s adaptation roadmap. It provides scenario-specific data on heatwave duration, maximum temperatures, and associated economic costs, offering further insight into the benefits of nature-based solutions.  

 Model Interactions: From Climate to Energy and Behaviour 

 Figure 3: The KNOWING Behaviour Model shows typologies for dealing with climate change and offers ways to involve a wide range of people. Credit: AIT 

These models are not isolated tools, they are interconnected in an offline coupling architecture. WRF provides the climate boundary conditions, which are refined by PALM-4U for detailed urban microclimate simulation. PALM-4U’s outputs are reconciled with the simulations of D-MERF and HWLEM, both of which assess health-related risks and socio-economic consequences. In parallel, PALM-4U results also feed into MAED-City, which models future energy demand shaped by changes in urban structure, insulation levels, and population health. The outputs of MAED-City are used by IESopt, an energy system optimisation model that defines the most efficient renewable energy mix for Tallinn. Land-use changes and transport-related shifts are further assessed by MATSim and PTV Visum, which simulate mobility behaviour and its influence on energy and infrastructure. The KNOWING Behaviour Model captures the likely societal acceptance and behavioural response to the interventions implemented in the city.  

 A Resilient Future for Tallinn 

 Together, these models provide Tallinn with a coherent, data-driven assessment of how targeted urban planning and green infrastructure can reduce heat stress and enhance resilience. By integrating climate, health, energy, mobility, and behaviour dimensions, KNOWING’s modelling framework offers a vision of a healthier, and more adaptive Tallinn, while creating transferable knowledge for other European cities facing similar challenges.