News and information about EUCP, the research project that develops the foundation for a cutting-edge climate prediction system for Europe.

The EUCP project aims to support both scientists and climate information providers to produce better climate information. To do this, EUCP develops innovative approaches on how to use existing climate predictions*, as well as providing new climate simulations. This will enable climate information providers to produce more consistent, authoritative, and actionable climate information in order to better support decision-makers on climate adaptation and mitigation.

May 2022 brought an official end to the EUCP project. Indeed, by the time you read this final edition of our newsletter we may well have already had our final review meeting, presenting our achievements to the European Commission. We’ve seen so many great outputs, products and findings from across the EUCP project over the last four and a half years. In this newsletter, we’ll take a look at some of EUCP’s latest outputs, our recent project meetings, and the legacy that EUCP leaves behind.

EUCP has produced a selection of policy briefs describing our key advancements and innovations, as well as future progress.

Find out more about our new Policy Briefs on our website

Policy Brief 1 describes some of the changes in the climate across Europe already happening and the impacts this has produced in recent years. Through findings from EUCP research, we outline how the future climate of Europe is likely to change and what impacts this could have on key industries and communities.

Policy Brief 2 highlights some of the opportunities to build on what EUCP has learned in future research activities. New EU Horizon programmes are already exploiting and building on what EUCP has learned and applying it to new projects, such as nextGEMS and climateurope2.

Policy Brief 3 presents some of EUCP’s key outputs, products and methodological advances, intended to highlight developments useful to climate service providers. The EUCP project has provided new scientific findings and various different climate service products to bring the very latest approaches to the hands of climate information users.

Finally, EUCP has also produced a briefing on our work providing high-resolution climate simulations covering some of Europe’s outermost regions. This highlights the benefits these simulations can bring to the island of La Réunion, the Madeira and Canary archipelagos and the Caribbean region.

In EUCP, we have recently developed a web application to host collections of (scientific) storyboards. Originally used by movie directors, a storyboard consists of a series of images that, together, form a story. Our storyboards have nothing to do with filmmaking, but the term still seems appropriate to convey their essence.

These storyboards occupy the middle ground between a scientific poster and a traditional slide-deck presentation. The main content is shown prominently, with room for annotations on the side. This makes it a great stand-alone presentation format that works well for showcasing all kinds of project outputs, linking to external resources for more detailed information.

Currently our collection consists of no less then twenty stories ranging from case studies on flash floods and coastal erosion to the most recent insights on convection-permitting modelling and decadal climate prediction. They cover new datasets, methodologies, and applications.

If you want a quick, high-level overview of some of our most important outcomes, the storyboards are a great start. They are accessible at this link.

We believe the storyboard format can facilitate science communication also beyond EUCP, and we’ve made an effort to keep them simple and user-friendly. More details on how the storyboards can be (re)used in other projects may be found in this blog post.

While we are happy with the project achievements and relationships built, we are at the same time sad that the project has come to an end. Still, we hope to keep in touch with the user community that has been developed over the past 4.5 years through other ongoing work and future initiatives and projects. If you ask yourself which scientific project to follow next, we would like to recommend NextGEMS. Maybe you were fascinated by EUCP’s pioneering data from convection-permitting regional climate models? You will love NextGEMS even more. Check this out: NextGEMS develops a next generation of global models, called Storm-Resolving Earth System Models (SR-ESMs). Correctly representing a ‘storm’ requires a very high resolution; NextGEMS’s storm-resolving models thus use a fifty-fold finer grid than conventional Global Earth System Models (3 km instead of 150 km). This allows them to explicitly represent essential climate processes – storms associated with deep precipitating clouds or the effects of the landscape on the atmosphere. This leap in model resolution will allow outputs at the scales on which impacts of severe weather events are felt, addressing the existing information gap at the geographical scales of interest for society. NextGEMS will thus provide seamless and accurate information about the climate system for the next 30 years.

The high computational and data demands of these new Earth System models also require novel and more inclusive ways of collaborating. NextGEMS will involve a broader stakeholder community in Earth-system modelling, through Knowledge Coproduction Hackathons. By integrating different types of knowledge and experience, the hackathons will both advance model development and make project outputs available immediately to users for testing and improving them. The latter addresses the recognised time lag between data production and the time when the data is available for users.

A guiding inspiration for the hackathons will be solving real-world challenges. For example, NextGEMS model outputs can show how climate change will affect future solar and wind energy resources, or primary marine production. Close collaboration with the renewable energy and fishery sectors will then help us answer their questions, such as how the projected future resources will affect wind or solar energy production in certain regions of Europe, or how they will affect marine ecosystems off West Africa. In this way, and by pooling the knowledge of everyone involved, the project findings can inform seamless planning for energy and food security.

If you are interested in continuing the collaboration that we had at EUCP Multi-User-Forums and through other occasions, you may consider applying to future NextGEMS hackathons (check out latest updates here) or reach out to the project Storms&Society team. You can also follow the project's latest developments through NextGEMS Vlog (video blog) and @nextgems_eu.

Our final meeting in EUCP (4th – 6th May 2022) was a chance to see all our colleagues again and share the last 4 ½ years of progress and advances under the EUCP project. Over 100 delegates joined us to hear the latest highlights from each work package, the most important things we’ve all learned from taking part in the project, and the legacy EUCP will leave behind.

On day one we heard from Work Packages 2 and 5 on their work on regional climate ensemble projections, including different methods of climate model ensemble constraint and model weighting, and climate storylines. The final talk of the day from Peter Kalverla highlighted the many useful products that have come out of EUCP, including the EUCP storyboard platform and the new EUCP data catalogue. These help put our findings and the important datasets and visualisation underpinning them into the hands of more people in an accessible form.

The second day of the meeting began with a look at work on improving decadal forecasting, largely carried out by Work Package 1, including work on forecast quality assessment, recommendations for the future development of decadal prediction systems, and work on the applications of these decadal predictions as part of the C3S_34c project with the Copernicus Climate Change Service.

The next session of the meeting covered the activities of Work Package 5 on merging climate predictions and projections across timescales. Gabi Hegerl described her work on the importance of using consistent observational constraints in both short-term predictions and longer-term projections, before further talks explored different methods of combining decadal predictions and climate projections.

The third and final day of the meeting began with work from several work packages on high-resolution, convection-permitting climate simulations for Europe and how they can be used. We heard about how these models can provide better representation of Alpine climate than lower-resolution models, as well as how they simulate future high-impact weather events. Hylke de Vries also presented work using these models in some outermost European regions, specifically the Indian Ocean island of La Réunion, the Canary and Madeira archipelagos and the Caribbean.

Delegates then heard from Jens Hesselbjerg Christensen on the final workshop of EUCP’s Multi-User Forum, which is described in detail later in this newsletter. A final panel discussion considered “Where next for European climate prediction?”, considering the progress that has been made in this area and the work that still needs to be done to make climate information more useful and actionable for users. The meeting was closed by the Met Office’s Prof. Jason Lowe OBE, who highlighted the fantastic outputs of EUCP, including the huge number of published papers, the data catalogue and the storyboard platform.

To find out more about all the talks and discussion in our final project meeting, please visit our specific web pages for Day 1, Day 2 and Day 3.

Over 100 participants joined us for our final project meeting

On the 3rd of May 2022 the third and final Multi-User Forum workshop took place virtually. Following the main conclusions from the previous workshop, several targeted questions were formulated and distributed amongst the MUF members before the workshop. Several MUF members (e.g. C3S, DMI, UK Met Office, SMHI, Hungarian Met Office) provided valuable answers in advance.

The first question concerned new methods to deal with uncertainty. For example, how does your organization deal with uncertainty in predictions and projections and what would you like to do to improve it further? All members mentioned how important this issue is. For now, they use probabilistic approaches such as multi-model spread (e.g. climate sensitivity), ensemble of projections, or non-parametric methods such as calculating the 10th / 90th percentile of the distribution of indicators to probe model agreement. They all mentioned that they were inspired by several approaches of ensemble selection further developed within EUCP.

The second question concerned new methods for ensemble sub-selection. Specifically, how does your organization deal with ensemble selection in predictions and projections and what would you like to do to improve it further? For this topic some of the members used the “model democracy” approach where the motto used is “more the merrier“. However, they all mentioned that they were inspired by several approaches of ensemble selection tried out within EUCP.

The third question addressed convection-permitting climate models. Although there is considerable interest toward such an approach, the workload related to utilising this large amount of data represents a barrier to many users. So, the question is to identify the role for this kind of data in their climate service/daily work? Although users do not see that CPM information could replace their current approach (due the very small range of covered uncertainties), they see such information as complementing their current adopted ensemble approach. They also saw great potential to better understand extreme precipitation and local processes such as those related to the urban heat island effect.

Finally, the last question concerned their interest in decadal climate prediction. How does their organization deal with decadal prediction? What data products do they provide around decadal prediction? While some users have already integrated decadal prediction, others, although they see the potential of such a product, do not plan to integrate such information in the near future. A major concern related to most of the elements mentioned is how to combine the growing diversity of results across CMIP generations, emissions scenarios and model resolutions, now also adding models with a convection permitting resolution, into something usable.

This crucial information was used to structure the workshop. The introduction represented mostly an updating of the MUF members from the latest progress on storyboards and storylines since the last workshop. Following the introduction, two breakout sessions inspired by the answers collected were organized. The discussion in the two breakout sessions started in the users’ context, e.g. why the users want that specific climate information and what they are using.

The first breakout session asked “How reliable is our climate change information? And how to cope with it?” Several discussions took place in that first breakout session and multiple suggestions were made in order to provide reliable and usable information. The first suggestion was to investigate multiple lines of evidence to understand seemingly conflicting sources of climate information, which could then be used to build storylines. Following this suggestion, emerging constraints were discussed, developing narratives which could also be focussing over central and worst-case scenarios. Finally, understanding the sensitivity of climate drivers was suggested as an approach to increase the reliability of climate information.

The second breakout session was on “How to manage/reduce the complexity of the available climate information.” The focus of the session was on the HOW, but during the discussion, it became clear that it was not possible to answer that question without a better understanding of the context in which this question is asked. One example was on the selection of models where the selection criteria are highly dependent on the context of the user needs. In order to have a real sense of the user context, the relation of power between the scientist and the user should be reversed. Only by reversing this power relation can a proper co-design/co-development process take form.

The remaining part of the workshop focussed on newly-started and upcoming follow-up EU projects such as NextGEMS and climateurope2, along with the now fully operational C3S engagement with users. NextGEMS has engaged with a well-defined, limited set of users already from the design phase of the project and they want to integrate with more users. Climateurope2 aims to support and standardize climate services in Europe and beyond by networking projects to support the climate service community, provide the basis for standardization in the EU (and beyond), and finally increase the uptake of quality climate services in public organizations/society. They also intend to potentially answer questions resulting from EUCP. Finally, C3S has the objective to supply data (through CDS) and provide applications using the data provided.

Each of these projects recognized that the user requirements have changed in recent years. There is an increased awareness of the existence of climate information and climate services. For example, the contribution of Working Group II to the IPCC Sixth Assessment Report was welcomed by users; it has brought attention to the problem according to the users’ needs. NextGEMS specifically mentioned that climate information from high-resolution models was relevant to renewable energy users and that they will overcome some of the issues encountered by EUCP by having these users involved from the early stage of the project.

The workshop closed by asking four questions to the main EU projects/organizations that took part in the workshop which included not only NextGEMS, C3S and climateurope2 but also DMI, SMHI, the UK Met Office and the Hungarian Meteorological Service.

Table 1: List of questions and related answers resulting from the 3rd Multi-User Forum.

Please also find our report on the meeting on our website here.

Almost 50 delegates joined our 2022 Multi-User Forum meeting

EUCP set out with a very ambitious goal, to improve predictions and projections of future climate on time-scales out to 40+ years, with a focus on Europe, but applicable globally. As the project draws to a close, we reflect on some of the major steps that the project has made towards these core scientific objectives that look towards a European prediction system.

How can we improve decadal scale predictions?

Decadal climate predictions have improved considerably in recent years, and EUCP science has shed new light on where those predictions are skillful. Our many examples include studies on the North Atlantic Oscillation, Southern European temperature, and sea surface temperatures in the North Atlantic Subpolar Gyre.

Funding from EUCP has helped establish the WMO Lead Centre for Annual-to-Decadal Climate Prediction, which gathers forecasts from a number of contributing modelling centres and produces a new decadal forecast every year including predictions of various indices such as temperature and rainfall. The availability of new multi-model operational predictions has allowed the development of pioneering new prototype climate services. These cover many different industries, demonstrating the broad applicability of this kind of climate information. Examples include mitigation advice, forecasts of future windspeed for the wind energy sector, and heat stress and drought indices relevant for the wheat farming sector.

Forecasts of Standardised Precipitation Evapotranspiration Index (SPEI) and the Heat Magnitude Day Index (HMDI), two important user-relevant indices associated with wheat yield (Solaraju-Murali et al., 2021).

How can we provide seamless climate prediction information for 1-40+ years?

EUCP has developed pioneering new methods to combine the information from prediction and projection systems in order to provide a single, seamless product for applications which span these timescales from 1-40 years.

One approach explored in EUCP uses the information from the decadal predictions to constrain the longer-term projections (see examples here and here). By sampling projections with the highest agreement with the decadal predictions over their first 10 years, EUCP has found that the selected subset shows improved skill out to around 20 years in the future and could therefore be used to provide climate information across the timescales. An alternative approach is to ‘stitch’ together information from prediction and projection ensembles. EUCP has pioneered this stitching approach using prediction and projection data from 8 climate modelling systems, exploring how observational constraints could be used to minimise inconsistencies at the merging point.

Ongoing research is taking place on these techniques, which could become a valuable part of the future climate information toolkit. The role of observations in skill estimates and constraints needs to be understood in order to use both consistently across the prediction and projection time horizons. The project’s perspective paper discusses the different approaches to constraining predictions and projections and highlights important challenges still to be addressed.

How can we get better regional information from climate projection ensembles?

The CMIP5 and CMIP6 ensembles offer important information about future climate change. EUCP science has shown that methods to constrain these projections offer important added-value over using the raw projection ensemble – importantly lowering the high-end of projected temperature changes, particularly in CMIP6. EUCP has undertaken the first systematic intercomparison of different methods for weighting, constraining and calibrating European projections from CMIP5 and CMIP6, drawing on existing methodologies and new approaches developed within EUCP. We also took this further to produce the first objective assessment of the added value that multiple constraint methodologies offer.

Storylines are another means of exploring the complex uncertainties in climate projections. In EUCP we have explored a range of different approaches including event-based narratives which, for example, ask what an event like the 2018 European drought might look like in a world that is 2 or 3 degrees warmer.

How can we better simulate future local extreme weather?

EUCP has produced a set of new high-resolution CPM projections for domains capturing mainland Europe. For each domain, these high-resolution projections have been made using at least two CPMs developed by different climate modelling centres. EUCP contributed to the important CORDEX FPS-CONV project comparing 12 CPMs, assessing their present-day performance and future projections for a common domain over the Alps. These multi-model CPM projections allow us for the first time to assess the uncertainty in future changes in high-impact weather linked to convection, such as high-intensity, short duration rainfall. EUCP has also applied these new high-resolution simulations to river basin models for the Alps to assess potential future changes in flash flooding.

Convection permitting simulations produced by EUCP have already been used in further collaborations to provide new scientific developments about future high-impact convective weather events such as sting-jets that cause extreme windstorms and slow-moving storms that can lead to high accumulations of rainfall and flooding. Studies on changes in lightning and severe hail incidences are also in preparation.

Read more about how EUCP’s convection-permitting simulations will help Europe plan for extreme weather in our Carbon brief Guest post.

Finally, an important measure of impact for EUCP is its influence and inclusion on the IPCC Sixth Assessment Report (AR6). You can read more about EUCP science in IPCC AR6 here.