Open dataset helps Europe redesign buildings for a vanished climate

Europe’s buildings are being designed with weather data that may already belong to the past. Engineers commonly rely on Typical Meteorological Years, or TMYs, assembled from decades of historical observations to simulate how buildings will consume energy and respond to weather. Yet a structure designed today may still be operating in 2080, when temperatures, heating requirements, cooling loads and seasonal patterns could be substantially different. Researchers from the Norwegian University of Life Sciences and the University of Coimbra have now produced future weather files for more than 5,000 European locations, giving building designers access to hourly climate projections that can be used directly in standard energy-simulation software.

The new dataset addresses a problem that has quietly shaped building design for decades: the mismatch between the climate represented in engineering calculations and the climate a building will actually experience. “The gap between the climate conditions we use in our calculations and the climate the building will experience could exceed half a century,” says Thomas K. Thiis, professor at the Norwegian University of Life Sciences. A building designed using outdated weather assumptions may be perfectly optimized for historical conditions while becoming overheated in summer, overcooled, or less energy-efficient as the climate changes. The researchers’ Future Meteorological Years, or FMYs, are intended to close that gap by translating climate-change projections into practical, hourly weather files for specific places and future periods.

TMYs are widely used in building-performance simulations, from individual homes to hospitals, schools and large commercial campuses. They typically contain hourly values for variables such as air temperature, relative humidity, wind speed, solar radiation and atmospheric pressure. These inputs allow software to estimate heating and cooling demand, indoor comfort, ventilation requirements and the performance of equipment such as heat pumps and chillers. An FMY uses the same simulation-ready structure but modifies the historical baseline according to projected climate changes. This enables engineers to test how a building might perform during the middle or the end of the century rather than assuming that past weather patterns will continue unchanged.

The researchers did not simply select a single climate model and apply its output across Europe. Their study, published in the Journal of Building Performance Simulation, evaluated a large collection of global and regional climate models against observed European conditions. The team examined how well the models reproduced important summer and winter characteristics before selecting those considered sufficiently reliable for generating the future files. This model-selection process is critical because climate models differ in their representation of regional circulation, precipitation, temperature extremes and seasonal variability. A projection that performs well over northern Europe may not reproduce conditions accurately in the Mediterranean, where heat, solar radiation and water scarcity create different design challenges.

The FMYs cover two future periods: mid-century, from 2036 to 2065, and late-century, from 2066 to 2095. They also include several emissions pathways from both the CMIP5 and CMIP6 climate-model frameworks. The scenarios range from relatively low-emissions futures, including RCP4.5 and SSP1-2.6, to high-emissions pathways such as RCP8.5 and SSP5-8.5. These scenarios are not predictions of a single inevitable future. Instead, they represent different possible trajectories shaped by energy systems, land use, technology, policy and global emissions. For building projects, however, the choice has direct practical consequences. A design based on a lower-emissions pathway may require different levels of shading, ventilation and cooling capacity from one intended to remain comfortable under a high-warming scenario.

One of the most important findings concerns the uneven distribution of warming across Europe and throughout the year. In northern locations such as Oslo, projected temperature increases are particularly pronounced during winter, while summer changes may be comparatively smaller. That pattern matters because heating dominates energy demand across much of northern and central Europe. A modest increase in winter temperature can therefore produce a major reduction in heating degree days, a commonly used indicator of how much heating buildings require. The accompanying data visualization shows the largest reductions in heating demand across northern Europe, while Mediterranean regions experience smaller decreases because their historical heating requirements are already limited. “When the warming is concentrated in the season when energy consumption is at its highest, it has a major impact on the calculations,” Thiis says.

The dataset is designed to avoid another source of uncertainty: methodological inconsistency between historical and future weather files. The FMYs were created using the same foundation as the researchers’ recently published TMY dataset, which covers more than 5,000 European locations and is based on the Copernicus CERRA reanalysis. Reanalysis combines observations with numerical weather models to reconstruct past atmospheric conditions at a consistent spatial and temporal scale. By using the same reference period, spatial resolution and processing approach for both present-day and future files, the researchers aim to ensure that differences in simulated building performance reflect climate change rather than differences in data construction. “Engineers can now be sure that the climate change signal of the FMYs is not over- or under-estimated due to methodological mismatch,” says co-author Eugénio Rodrigues of the University of Coimbra.

The implications extend beyond energy bills. Future weather files can help engineers assess overheating risk, indoor air quality, occupant comfort, renewable-energy production and the resilience of heating, ventilation and air-conditioning systems. A building that requires less heating in a warmer winter may simultaneously face greater cooling demand during heat waves. Solar shading, glazing, insulation, thermal mass and natural-ventilation strategies may perform differently under future conditions than they do under historical weather. Hospitals, data centers and other facilities with strict temperature requirements may be especially sensitive to these changes. By allowing designers to run familiar simulations with future climate inputs, FMYs bring climate adaptation into the routine technical decisions that shape buildings long before construction begins.

Researchers in Denmark, Poland, Spain and other European countries are already using the files to investigate building energy performance and climate adaptation. The complete collection of TMYs and FMYs is freely available through climatedataforbuildings.eu in EnergyPlus Weather, or EPW, format. EPW files can be imported into widely used tools including EnergyPlus, IDA ICE, TRNSYS and IES VE, allowing the projections to be incorporated into existing workflows rather than requiring entirely new software. The project’s central message is straightforward but consequential: buildings designed for the decades ahead need weather data from the decades ahead. As Europe grows warmer, the most influential climate signal for a building may not be a dramatic annual temperature change, but a subtle seasonal shift occurring precisely when that building consumes the most energy.

Web References: https://climatedataforbuildings.eu/ ; https://www.tandfonline.com/doi/full/10.1080/19401493.2026.2697216 ; https://www.sciencedirect.com/science/article/pii/S0378778825015440

References: Journal of Building Performance Simulation; Norwegian University of Life Sciences; University of Coimbra; Copernicus CERRA reanalysis.

Keywords

climate change, future meteorological years, building energy simulation, Europe, climate adaptation, heating demand, cooling demand, EPW files, building performance, climate modeling