Periodic Reporting for period 1 - CIRCE (City-oriented Impacts of Regional Climate for Europe)
Reporting period: 2023-07-01 to 2025-06-30
More than half of the world's population lives in urban areas, a figure that rises to 75% for European residents. Due to ongoing global anthropogenic changes, such as climate change and urbanisation, city dwellers are facing increasing pressure and adverse weather and climate effects. In order to adapt to these challenges and build resilient cities, urban planners, stakeholders and decision-makers need high-resolution climate information tailored to their needs (e.g. in terms of spatial resolution, time horizon or impact indicators). Today, most of the climate information produced at the local level comes from regional climate models (RCMs) that are used to refine global climate projections made by global climate models (GCMs) (e.g. CORDEX, Coordinated Regional Climate Downscaling Experiment). One of the problems is that, historically, most regional climate models did not represent cities in great detail due to their still too coarse horizontal resolution (on the order of tens of kilometres) and because they did not activate the physical parameterisations dedicated to urban areas. Thanks to improvements in computing resources, a new generation of high-resolution (a few kilometres) regional climate models (known as convection-permitting regional climate models, or CPMs) has been developed over the last decade. These models enable more complex areas to be studied thanks to their increased resolution, such as mountainous areas, coastlines and cities.
As part of the CIRCE project, we are using the largest ensemble of high-resolution regional climate simulations produced as part of the CORDEX Flagship Pilot Study on convective phenomena (FPS-Convection) of the World Climate Research Programme (WCRP). These simulations have been widely used to study the added value of the new generation of models compared to their predecessors in representing extreme events such as heavy precipitation or heat waves, but they are being used here for the first time to study the urban climate of European cities.
The project is structured around two main areas, with a particular focus on evaluating the different types of models, the uncertainties associated with each type of configuration, and the second part focusing on the expected effects of these choices on climate impact projections.
As part of the CIRCE project, we are using the largest ensemble of high-resolution regional climate simulations produced as part of the CORDEX Flagship Pilot Study on convective phenomena (FPS-Convection) of the World Climate Research Programme (WCRP). These simulations have been widely used to study the added value of the new generation of models compared to their predecessors in representing extreme events such as heavy precipitation or heat waves, but they are being used here for the first time to study the urban climate of European cities.
The project is structured around two main areas, with a particular focus on evaluating the different types of models, the uncertainties associated with each type of configuration, and the second part focusing on the expected effects of these choices on climate impact projections.
The main scientific activities carried out as part of the CIRCE project focused on analysing regional climate simulations at different horizontal resolutions and levels of sophistication for the representation of urban areas. For the first time, two ensembles of high-resolution regional climate simulations were used on several cities to study the combined effects of the urban climate and climate change in Europe. A significant part of the project was devoted to evaluating regional climate models for cities, with a particular focus on the potential added value of higher-resolution models. The results of this evaluation are presented in two scientific publications currently in preparation. The first one presents a review of the different land surface models and urban parameterisations used in regional climate models (RCMs) and global climate models (GCMs) by the European modelling community, followed by a general assessment of the models' ability to represent near-surface air temperature and Urban Heat Islands. The analysis clearly demonstrated the added value of high-resolution models in representing complex heterogeneous areas such as coastlines and, above all, in our case, in representing the urban climate. The added value of high-resolution models for cities was linked to (1) improved horizontal resolution, which allows for better representation of surface heterogeneity, and (2) more sophisticated urban parametrisations that represent more physical processes (e.g. the effect of city geometry on street temperatures or the effect of anthropogenic heat release). A complementary article analysing the relationship between meteorological variables around cities that are known to be potential drivers of Urban Heat Islands (and therefore exacerbated urban climate impacts) concluded that the added value of the latest generation of models can also be observed outside cities and in the relationships between rural and urban areas. For example, a correct representation of the urban climate requires a good urban model, as well as an adequate representation of the surrounding climate (e.g. wind speed, daily temperature range, precipitation, etc.) and their mutual relationship. The second part of the project focused on studying urban climate change (and Urban Heat Islands evolution) in the context of regional warming, without taking urbanisation into account. The results show that changes in urban heat islands will be small compared to regional background warming and are extremely sensitive to the model used. More interestingly, it was found that the changes are not only sensitive to the high-resolution model and urban parameterisation, but also to the entire downscaling chain due to the interdependence between the surrounding rural conditions and the urban climate mentioned above. This highlights a crucial point: the use of high-resolution models, due to their increased complexity, also comes with another layer of uncertainty that must be clearly communicated to end users. For example, switching from a previous generation of regional climate projections to a more recent generation could yield very different results for cities due to differences between model generations.
For the first time, a detailed assessment of an ensemble of high-resolution regional climate simulations has been carried out on several cities. The high level of detail in the evaluation, which is usually limited to weather models (used forecasts), has made it possible to establish guidelines for future simulation exercises. The CIRCE project, conducted in parallel with the World Climate Research Programme’s Flagship Pilot Study on URBan environments and Regional Climate Change (FPS URB-RCC), provided valuable information on methodological choices such as the selection of land cover databases and the approach to be taken to deal with heterogeneous areas (e.g. uncertainties associated with representing a single surface type within a model point versus multiple types), and the importance of considering certain physical processes (e.g. anthropogenic heat release). It also provided a better understanding that certain simple urban parameterisations are not sufficient to study impacts at the local level, which is now being taken into account by certain institutes for future modelling exercises (EURO-CORDEX and the FPS URB-RCC). More importantly for end users of local climate information, the project showed that for certain climate indicators that are highly sensitive to urban heat islands (e.g. the number of tropical nights – where the daily minimum temperature remains above 20°C), the transition to the new generation of models with more detailed urban schemes could have significant impacts on the results. Therefore, users of regional climate information should bear in mind this additional source of uncertainty, and modelling centres should ensure that the information is communicated effectively.