The online survey was available in both English and Spanish. Before launching it in January 2024, it was tested with five individuals representing different technical profiles and questions were adjusted following their feedback. The survey was open for three months. Through a non-probability sampling strategy, it targeted climate adaptation researchers and practitioners working in climate-sensitive organisations who had either used or could potentially use climate information for adaptation purposes. We used a dual dissemination strategy targeting EU projects as well as individual users. We selected various technical and applied projects where the authors are involved or in contact with (i.e. Climateurope2,1 ESiWACE,2 ASPECT,3 MIP4Adapt4). In addition, associations and companies in the agriculture (wine), energy, and insurance sectors were mainly targeted, since those were the sectors of interest of the GLORIA project. Channels used for survey dissemination included, depending on the project, the official mailing lists with internal and external stakeholders, the project communication channels (website and social media), newsletters, and some events organised by the projects themselves. A total of 29 responses were received, mainly from individuals belonging to organisations having their activities in Europe. Answers mostly came from users in the public sector and academia, but also from the private and the third sectors. Regarding the sector of activity (following the classification from the United Nations, 2008), respondents could be categorised as professional, scientific, and technical activities (13), agriculture, forestry and fishing (6), education (3), activities of households as employers (3), water (supply, management) (2), administrative and support service activities (1) and human health and social work activities (1). Scientists (14) were the largest group of respondents. From the other stakeholder categories, the distribution of roles was as follows: engineers (3), director (2), senior manager (2), junior manager (2), technicians (2); followed by one planner, a policy-maker, an analyst and an assistant. The survey included questions on respondents’ awareness of the DT, the main benefits and challenges they expected and the preferred forms of output, among others (the complete survey can be found in Section 1 of the Supplementary information). In the Results section, the question number is indicated with Q#, indicating the origin of the answers.

We conducted two rounds of one-on-one interviews with five key users involved in the development of the sectoral applications that made use of DestinE climate DT output. This includes users from the energy, hydrology and water management, urban planning, and wildfires sectors. They were selected based on an already established relationship of collaboration with scientific partners. The first round of interviews was semi-structured and designed to understand users’ background and day-to-day tasks, their acquaintance with climate-related information and their decision-making contexts. During the interviews, the concept of storylines (i.e., climate simulations showing how extreme events may unfold under different future climates) was introduced to the users, who were initially asked to identify a past extreme event that impacted their activities. This allowed us to involve the users in the definition of storylines and, in the second round of interviews, discuss with them the newly generated results and how they could support users to better draft possible adaptation responses or actions to adapt to climate change.

We attended several digital twin-related events gathering scientists, technology developers and potential users where we systematically observed participants’ interactions (outspoken opinions, questions raised, exchanges during debates) through participant observation (Kawulich, 2005). Among others, these events include two DestinE User Exchange workshops (in February 2023 in Frascati, Italy, and in November 2023 in Bonn, Germany) that had the aim to grow a community of stakeholders and listen to users’ requests and feedback, and a Capability Providers meeting in June 2023 in Bonn, Germany, addressed to foster collaboration and communication among those developing the DestinE platform. We also participated in discussions about the Earth Virtualisation Engine, such as the EVE summit in July 2023 in Berlin, Germany, and events organised during the WCRP Open Science Conference in October 2023 in Kigali, Rwanda. These meetings provided an excellent environment for discussion, where the main knowledge gaps and expectations from users could be identified.

In this section, users’ expectations are structured around the more frequently mentioned ‘main added values’ of the digital twin (see 3.1 Expectations of the climate DT, 3.2 Simulation of different climate adaptation options at the levels where impacts are felt, 3.3 Interactivity allowing on-demand simulations, 3.4 Higher resolution for better understanding processes and simulating more complex systems below and Section 1/Q13 in the Supplementary information for more detail). According to survey respondents, the main added value is the simulation of different climate adaptation options at the national, regional and city levels. Likewise, many users mentioned interactivity, which allows to obtain information relevant for decision-making and test different scenarios for various potentially implemented actions. Related to interactivity, the near-real time availability of model outputs was also mentioned as an added value. This refers to the timely availability of data, that is, the ability of the system to assimilate current environmental data, simulate what will happen, and deliver actionable insights in a relatively short time (within minutes to hours). Other user expectations include the provision of data at higher spatial and temporal resolution and the availability of more accurate information regarding processes. The possibility to represent interactions in the different Earth system compartments was also signalled, although less emphasised. Other individual comments included improvements in the understanding and communication of model outputs as well as the ability to simulate more complex systems.

The climate DT aims to provide regularly updated, globally consistent information with high granularity at scales where many of the impacts of climate change are felt. Multi-decadal simulations are produced to cover the recent past and possible future evolutions of climate up to 2050. These simulations can be integrated with impact model applications allowing the computation of indicators tailored to users’ decision-making in support of climate change adaptation. Simulation protocols are adapted to km-scale simulations, which complement the already existing climate projections. Projections will be updated at least every year, considerably improving the frequency of current models run only every 7–10 years (i.e. CMIP, CORDEX). In the survey, research participants see the availability of past, present and future climate data simulations as one of the most relevant opportunities provided by the climate DT (Fig. 2), stressing the importance for users to have a quick look at both past and future data. On the other hand, during the interviews with potential sectoral users, the main opportunity raised was rather centred around the integration of Earth system models with impact models.

Impact-sector applications, i.e., a hydrological model or computations of relevant indicators, transform the climate data to actionable information on climate change impacts. One interviewee mentioned that ‘translating climate data to impact determines whether the information provided is relevant or not for the energy sector’, whereas another participant said that this data would be ‘important for running the downscaled urban impact model’ (in reference to a model they are using for urban planning). Although the integration of Earth system models with impact models is seen as a powerful opportunity by potential users (Fig. 2), in practice it might be challenging for some users without the appropriate competences to fully use and interact with the DT.

The possibility to compute tailored indicators to better match user needs and provide climate services that are fit-for-purpose was mentioned during user interviews and acknowledged in the survey, although with moderate relevance (Fig. 2). Indeed, by combining climate models with impact models, the DT is supporting the calculation of indicators like wind energy capacity factors, one of the key indicators for the wind energy sector, as it allows to directly estimate energy production. Another aspect to consider for this indicator is that it needs to be computed from wind speed at 100 m height, where turbines are placed, rather than at the standard 10 m height at which wind speed is typically provided. Through the data streaming of model outputs (i.e., the continuous and real-time flow of data from climate to impact models) and the provision of a wider diversity of climate variables, the climate DT is suited to support the computation of capacity factors and other sectoral indicators (Lacima-Nadolnik et al., 2024).

Due to the ongoing movement in the climate community towards increasingly higher spatio-temporal resolutions of climate data, questions are beginning to arise as to how this data will be managed (Hu et al., 2018, Bauer et al., 2021). Solving this challenge is important for the computation of the aforementioned tailored indicators. The streaming concept has been proposed to deal with data handling, which brings the climate models and impact models within the same workflow (Grayson et al., 2025). Nevertheless, despite the paradigm change that a new setup developed to transfer climate model outputs into users’ applications with lower memory consumption (streaming) represents, this opportunity was not highlighted by users in the survey. The reason could be that streaming was a new concept, illustrations of which were not yet available by the time the survey was conducted, but could also be related to the formulation of the question, which merged two different concepts (i.e., ‘new setup’ and ‘lower memory consumption’), that could have made it difficult for participants to select it if any of the options was not fulfilled (Fig. 2).

Challenges with the retrieval and handling of the climate DT data (Fig. 3) were mentioned during the interviews. A couple of users (from the water management and urban planning sectors) referred to potential challenges for users with a lack of experience in working with high resolution streamed data. Regarding the integration of digital twin outputs into users’ in-house tools (Fig. 3), some interviewees commented that depending on the bandwidth of the network used, it might be desirable for data to be processed before reaching the end-user, or else this task could be outsourced. Another user mentioned the difficulty in computing special variables that normally are not directly computed by climate models (wildfires sector).

Through interactive and configurable access to data, models and workflows, the climate DT is presented as an exciting opportunity to satisfy users’ curiosity. On-demand simulations enable experimentation, prediction, and scenario analysis to explore the potential consequences of different climate conditions and user interventions. Defined in our survey as the possibility to request simulations based on users’ needs, interactivity was not especially regarded as an important feature by respondents (Fig. 2). However, the concept of ‘interactivity’ is quite broad, and users may hold different understandings of it. Indeed, different interactivity options are, within some limits, provided by the climate DT, either related to configuring runs of the global multi-decadal climate simulations (e.g., different emission scenarios) or on the impact simulation side (e.g., building a new wind farm with specific characteristics in a particular location). The type of on-demand simulations was left open to the user interpretation during the interviews. For instance, an interviewee interpreted interactivity as a way to explore data without triggering new simulations, as suggested by the comment: ‘requesting simulations is not needed as the user may want to test the interventions, not the simulations per se’ (urban planning user). However, this comment still suggests an interest in interacting with the DT through the testing of different interventions, which is at the basis of adaptation practice. Other users signalled that interactivity is not a relevant characteristic for them, since they already use their own in-house models: ‘we have control over our model and are able to set it up to fit our future outlook and the simulations we want to run’ (energy sector user). This denotes this user’s high level of autonomy running energy models in a business as usual way, and a lack of interest or awareness on what the climate DT has to offer. Interactive online interfaces allowing a diversity of users to easily and intuitively apply their expertise were mentioned by survey respondents. Examples highlighted were the IPCC Atlas and the Climate Reanalyzer, which allow for a near-real time availability of model outputs, identified as one of the added values of the climate DT also related to interactivity features.

The exploration of projection scenarios for the next 30–50 years was selected as one of the top opportunities of the climate DT by survey respondents (Fig. 2). Additionally, during the interviews one participant indicated that this information ‘could be useful to see how wildfires spread may change under different scenarios’ (wildfires sector user). Although 30–50 years was the time frame discussed during the first phase of DestinE, and that is why this option was included in the survey, currently this initiative foresees running projections up to year 2050. Planning horizons for the next 10–30 years have also been identified in a large-scale survey as relevant by climate-sensitive organisations in Europe affected by weather and climate (n = 1644 respondents) (Grainger et al., 2025). One interviewee mentioned the need of having information at longer time scales, mentioning post 2100 assessments, which are needed for decisions related to long-term infrastructure investments in the water sector. The aforementioned survey by Grainger et al. (2025) reports a very low use of climate information for decision-making for planning horizons of 50 years or more. Nevertheless, because some users may be familiar with climate projections shown by the IPCC, they may expect to be able to play with different emission scenarios at these time scales in the DT as well (as suggested by the latter comment of the wildfires user). In this regard, another user mentioned that ‘future scenarios will be interesting to map the uncertainty of the projections’ and that ‘multiple realisations of a projection can improve the robustness of the information’ (energy sector users), suggesting that the availability of multiple scenarios and of ensembles is taken for granted. Although fewer emission scenarios may be initially available in the climate DT, this may not be critical for the next 20 years, for which these projections are run, since the differences between scenarios are small (Lee et al., 2021). However, a wider scenario variety will be required beyond the year 2040.

Dealing with the uncertainty in the results of the simulations and the lack of understanding of how to interpret long-term climate change projections were selected as important challenges by survey respondents (Fig. 3). These aspects will especially affect end users with lower technical background and familiarity with long-term climate information. Moreover, from the numerous interactions with potential users of the climate DT, there seems to be a perceived general lack of guidance on how to use the information from the DT (Fig. 3), especially for non-experts. This was exacerbated by potential users not having a clear idea of the type of information available in the DT and how this availability would evolve in the next phases of DestinE. However, it is important to note that at the time when the survey was conducted, the DestinE system was still not publicly open.

The possibility of testing different intervention scenarios (Fig. 2) was considered a moderately useful opportunity by survey respondents. However, this aspect was more appreciated by interviewees, who indicated that ‘what if’ scenarios would be useful to ‘carry out cross-sectoral risk assessments and prioritise vulnerable regions’ (water management user), and ‘determine the evolution of urban climate and land use in [their city], keeping ventilation corridors in the city open’ (urban planning user). The options to ‘introduce the most suitable forest management strategies to counter wildfire risks’ and ‘test the forest configuration and resources from rescue services’ were also mentioned during the interviews (wildfires sector user). In relation to ‘what-if’ questions, the use of storylines, which was introduced to key users in the first round of interviews, was mentioned by a water sector user as ‘a tool that, going beyond the current projections, could raise awareness on the next chapter for mitigation and adaptation in the German water sector’. This suggests a perceived potential added value of event-based storylines in relation to traditional climate projections.

Higher spatial resolution was indicated by potential users as one of the main opportunities brought by the climate DT (Fig. 2). Although from the scientific community it is clear that the value of higher resolution lies in the representation of physical processes happening at smaller scales in the models (e.g. storms, ocean eddies, etc.), which are crucial for renewable energy, flood prevention and other climate-related impacts, these benefits were not always obvious for users when thinking about their decision-making context. Some respondents hesitated on the possibilities to integrate this new type of data into in-house systems or tools currently in use. Those in need of high-resolution data highlighted their potential benefits, which they described as ‘important for capturing precipitation events and hydrological impacts’ (water sector user, referring to floods) and ‘important to capture the urban heat island effect, and especially relevant in coastal cells that fall at the interface between ocean and land’ (urban sector user). However, some users might not be looking for high-resolution data. As pointed out by a user interested in mapping the energy supply and demand in Europe, ‘climate information should be packaged in such a way that we can directly use it in our model. Otherwise, it takes a lot of effort’. This suggests that the penetration of high-resolution data into users’ decision-making context will very much depend on the impact models already in use.

Not only is the spatial resolution of the climate DT model outputs higher but also their temporal resolution. However, unlike spatial resolution, the usefulness of higher temporal resolution was ranked low in the survey (Fig. 2), showing a lower relevance compared to other opportunities. In contrast, the increase in temporal resolution was more valued during interviews, where climate model outputs were often considered as inputs for impact models. Thus, interviewed users mentioned the need for hourly data (e.g., for determining short-term energy dynamics in the energy sector as well as the real-time monitoring of wildfires) and even sub-hourly data (e.g., for water management during floods and extreme rainfall events). However, this introduces another important aspect to be considered by users of climate DT outputs at high temporal resolution: although outputs at hourly resolution may be directly used as inputs for impact models, their interpretation does not make sense at hourly, daily or even annual time scales, but only contributing to long-term (20–30 years) estimates of the statistical parameters describing the climate variables. This would be equivalent to estimating the “weather” of the climate variables.

Increasing resolution may provide multiple opportunities but it also typically comes at the expense of time and computational resources. In the opinion of one of the survey respondents: ‘I think that we often forget to mention the energy use of the DT compared to its added value for society. I’m not sure if it’s worth it’. This comment implies that despite the investments in technology and model performance, a more holistic approach may be useful when assessing the value of new sources of climate-related information for society.

The availability of climate observations was highly ranked in the survey as a way to better understand the Earth system (Fig. 2). Past observations were found useful for updating statistical databases and relating climate conditions to their potential impact. For instance, during interviews users explained that long time-series from the past are needed for the computation of extreme events, which are rarer (water sector user) as well as for updating the wildfires database from rescue services and looking for patterns in the data (wildfires sector user). Interviewees also mentioned that past long-term datasets would be useful for comparison purposes with other datasets currently in use (energy sector user). This would allow them to assess the quality of data by comparing simulations with observations, and better understand how models work for their area and conditions of interest. Although the availability of quality-assured data was also highly ranked in the survey (Fig. 2), we observed from interactions with potential users at events that many assume the data provided by the climate DT to be quality assured.

Survey respondents preferred visualisations (i.e., maps, data charts, graphics, etc.) as the main channel for the communication of the climate DT outputs. During one of the events, visualisation was described by one of the participants as ‘a useful way to deliver climate data to non-experts’. The second more popular communication channel was the direct connection to the climate DT infrastructure. The access to raw data as well as written and oral communication received less attention.

A limitation of the survey was the low response rate and the presence of certain biases, particularly in the respondents’ country of operation (with one third of the respondents based in Spain) and the sector of their organisations (thirteen from the professional, scientific and technical fields). With regards the country, the dominance of Spanish respondents is likely due to the survey being launched within the framework of the GLORIA project, which focused on the production of climate information for the Iberian Peninsula. The limited number of responses may reflect survey fatigue, especially given that it was mostly conducted online and targeted busy professional groups. In addition, we intentionally chose not to disseminate the survey during DestinE climate DT events to avoid self-selection bias, as we assumed those participants would have a knowledge advantage over the broader target audience.

In the survey, we have used binary questions (yes/no) to assess the awareness of participants about the climate DT, which could introduce random variations and response error, since this format makes data sensitive to non-uniform understanding of the scale (i.e., two respondents that know about the concept but feel that they do not know it in detail, could give opposing answers). In this case, we have tried to overcome this limitation by adding follow-up questions asking users in which context they have heard about the digital twin concept (if they answered ‘yes’) and to guess what it could be (if they answered ‘no’). In the question about the climate adaptation DT, we also asked respondents who have heard about it to describe how they learnt about it.

The phrasing of certain options provided in the survey regarding the opportunities and challenges of the climate DT may have affected how they were interpreted by respondents. To mitigate such a risk, we pilot-tested the survey with five individuals representing different technical profiles, who raised no issues regarding the interpretation of the options. However, lower rankings for some options may reflect a certain ambiguity in the case of double-barrelled items rather than a lack of importance. For this reason, the number of non-responses has also been added to Fig. 2, Fig. 3 for additional information.

Target user groups are non-uniform and include users with different backgrounds and expertise, although expert users and co-developers will be more readily able to interact with the tool. A key aspect that emerged from the discussions with events’ participants was to ensure that users with varying expertise have a clear roadmap of how to interact with the DT. There was also interest in understanding the role of the private sector and new space companies aiming to make profit from the products and services developed. In addition, the younger generations of scientists and professionals were considered by some events’ participants a key target for the uptake of the DT technology, since they are more flexible and open to adopt new tools. There was also a mention to give careful consideration to the co-development with communities, including vulnerable communities (e.g., Indigenous peoples and island states currently affected by sea level rise) and local populations, which are more urgently in need of actionable information to adapt to a changing climate.

The capability of the DT to enhance the temporal and geographical granularity of information for better real-time understanding of climate phenomena and planning was mentioned by events’ participants. The quality of the climate DT model output was also mentioned and was considered a necessary but not sufficient attribute, since in addition to be perceived as of high quality, information also needs to be perceived as useful. Participants in events also stressed the need to clarify how the climate DT data complements and differs from other existing climate data sources such as national reference datasets, CORDEX or CMIP data. On the other hand, the usefulness of the DT data for regulatory and advisory user roles was discussed, including the need to explain how on-demand simulations can inform policy-makers and infrastructure operators (i.e., energy systems and infrastructure construction projects) in the selection of adaptation options at the local and regional scales as well as how they can complement national adaptation plans of EU member states.

Raising awareness and building community support around the climate DT was considered a key issue for events’ participants. During roundtable discussions, participants highlighted the use of innovative communication strategies such as narratives, storytelling, and data visualisation to better engage with general audiences. A mention was made to the communication of the statistical robustness and uncertainty of climate information. Sharing success stories, directly engaging with climate adaptation practitioners, the private sector and educators was suggested to better tailor and use success stories more effectively. For instance, using the concept of 'what if' scenarios was mentioned as a way to stimulate students' curiosity and enhance their understanding of the Earth system, its relationship with human behaviour, and the response to potential interventions for adaptation to climate change. Using amplifiers such as renowned scientists, policy actors, journalists or social media was also proposed as a strategy to engage with a wider community of interest around the climate DT.

Additional concerns stated by participants related to technical and operational aspects of the climate DT. These questions were mainly asked by participants with information technology and modelling background, dealing with data and model integration as well as interface design and access. Some of these questions involved the clarification of federated data spaces and integration of climate DT data with Common EU Data Spaces, the management of authentication and authorisation infrastructure, or the development of local digital twin models and their integration with the global climate DT, among others.

According to Hazeleger et al. (2024), the governance of digital twins must ensure that a diverse range of users and developers are listened to and adequately represented during the development and maintenance of the twins. In the context of the climate DT, only a limited number of expert users, including developers and scientific users, are expected to have access to high-end computational infrastructure and the expertise required to interact directly with the DT. In contrast, other professional users and the general public may lack these capabilities and will need to rely on alternative, non-interactive access to models and data. That said, our analysis suggests that some participants expected the DT to be used beyond the target groups, including vulnerable communities and local populations, for which the tool was not initially intended (see Section 4). This indicates a discrepancy between the actual climate DT target groups and users’ perceptions about these target groups. In case there was an intention to open the climate DT to other audiences, their public-facing components would be crucial and should be co-developed with the communities expected to use it.

Through interviews and participant observation, our work reveals that participants in our research are aware of the climate DT but not of how climate model outputs provide information meaningful to their activities. Besides, not all user groups will be interested in having the same level of interaction with the DT. For instance, downstream users interested in receiving a visualisation of some aspect of the results would require a lower level of awareness of the internal workings of the DT than the co-developer of an application, who will need much more insight. Being transparent and providing appropriate guidance to the different user groups is important to increase the chances of a wider use of the climate DT.

The accessibility to climate DT data should also consider accountability, ensuring that users understand the outcomes of the DT, as well as the responsibility associated with any actions or decisions, before using the information in decision- or policy-making. At present, guidelines on the use and interpretation of climate change projections are being developed by initiatives such as the WMO Climate Services Information System (CSIS), CORDEX, or the Copernicus Climate Change Service (C3S) as a mechanism to support the production and delivery of authoritative climate information. A similar strategy adapted to the climate DT could be implemented to ensure transparency and improve the trustworthiness of the information communicated to users. Guidelines should include information on data quality, acknowledgement of uncertainties, provenance of how the DT produces the information, disclosure of the choices made to design the DT (to identify unconscious biases), and the purposes the data is fit for, among others.

Early communication about the climate DT is crucial to raise awareness among the target groups and develop a community of interest around the DT. Not all interviewed users were aware of the opportunities and challenges of interacting with the DT infrastructure. Therefore, the European Commission devoted considerable effort to early communication actions for DestinE. In addition to press releases, articles, and scientific publications, the DT was introduced to potential users through videos, infographics and animations. Of relevance was the use of case study releases and storylines focusing on the usefulness of DT data for impact sectors. Participation at European and international events was arranged (i.e., Conference of the Parties, meetings of the Environmental Geosciences Union and fora such as the one provided by the Organisation for Economic Cooperation and Development). These communication actions had the aim to raise scientific credibility, enhance stakeholder trust and engagement, and support policy and decision-making uptake by showcasing real-world cases. Additional actions should continue to widen the scope of the DT and the engagement with these user groups as well as new ones, including those that could potentially benefit from the public-facing components of the climate DT. Sustained dissemination and engagement actions are key to involve users in the co-development of the DT (see the ‘Dissemination and engagement’ section below).

The co-development of dissemination activities with empowered users can support engagement with additional stakeholders, including the climate DT community of practice. Some of the activities discussed during climate DT-related events are described below, including success stories, data visualisation, storylines and educational materials for schools (see Section 4, part on ‘Community support around the climate DT’).

Success stories, i.e., selected examples of the use of the climate DT and its data, may be a way of showcasing how users can interact with the DT. In one of the DT-related events, the national climate scenarios for Switzerland (Fischer et al., 2022) were mentioned as an effective outreach practice to communicate the long-term impacts of climate change to society and increase the uptake of climate projections. A key finding was that communication of climate scenarios to the broader public should be targeted toward specific user types, which in that study was done applying a narrative approach with fictitious characters. Visual narratives and graphical storytelling, such as comics, infographics and scrollytelling, are being increasingly applied in the climate science and services fields, since they have shown potential to enhance learning and understanding of different scientific concepts (Tavares et al., 2023, Lc et al., 2022, Seyser and Zeiller, 2018).

Visualisation is also a powerful tool and is often the entry point to climate information that supports stakeholders’ decision-and policy-making processes (Grainger et al., 2016, Calvo et al., 2022, Terrado et al., 2022a). Visualisations were identified as the preferred way for users to receive the model results from the climate DT. Visualisation received also a lot of attention during the climate DT-related events. Increasingly, research insights from other fields such as design, cognition, psychology, behavioural sciences as well as human-centred and participatory design can support the development of more effective visualisations in climate science, which can enhance understanding, empower users and support decision-making (Morelli et al., 2021, Terrado et al., 2022b). The use of the DT to visualise fine-resolution geospatial data, which will appeal to broader audiences, has been highlighted, although visualising uncertainty in this data still remains as a challenge (Casola, 2023).

Storylines are increasingly used in climate science as a powerful approach to communicate and represent uncertainty when simulating recent extreme weather events, like heat waves, floods and droughts under multiple plausible climate futures (Shepherd et al., 2018, Sillmann et al., 2021). One of the approaches applied in the climate DT allows for reconstructing recent extreme events and their potential evolution under different climate conditions while maintaining high-resolution local details (John et al., 2025). From the interviews with key users involved in the development of the climate DT sectoral applications, we found that users are interested in knowing how extreme events may develop under different degrees of future global warming. This suggests that the integration of storylines into the climate DT simulations could help make information more tangible for users and address the challenges for dealing with uncertainty in climate projections, identified as a limitation for the uptake of the DT information. However, additional efforts are needed to welcome the social aspect of storylines, so that they are the result of a true co-production process and can be used as a tool to facilitate knowledge exchange between the climate DT and society (Baulenas et al., 2023).

Educational material for schools was also mentioned as a way to increase the uptake of climate DT data by the younger generations and was regarded as a more effective strategy than trying to change the way in which current users make use of model information. From the reviewed literature, educational materials used in the context of climate change communication encompass various formats that might be useful to adopt to communicate about the climate DT and its opportunities. Some of these formats include climate art and narratives (e.g., storytelling campaigns, comics), games and interactive activities (e.g., serious games, hands-on activities and demonstrations, platforms), discourse strategies (e.g., webinars, context comparison campaigns), as well as lessons and school teaching (Gómez et al., 2025).

Given the high complexity and novelty of the DT technological process, an ambition to apply a genuine co-production from the very beginning is desirable, involving users with a high-stake in testing and applying the latest climate information (e.g., interviewed key users). This allows developers to showcase examples of applications of climate DT outputs already in early stages of their development. These case studies serve to communicate and raise awareness about the climate DT and, hence, could help engage a wider community that may start to build capacity for using the DT and its outcomes.

The adoption of climate information can also be supported by the development of Communities of Practice (CoP) and peer-learning (Page and Dilling, 2019, Vincent et al., 2018). In this regard, research projects working on high-resolution modelling are a good venue to start fostering the creation of such a community, some of which are already underway, but certain institutionalisation and leadership is required to obtain the most from having a committed CoP (Retna and Ng, 2011). Without institutionalisation, there is the risk that these communities do not survive due to projectisation6 (Webber, 2019). On the other hand, the composition of these supporting communities cannot only be formed by researchers -albeit with practical sectoral experience-, but requires heterogeneity to enable continuous detection of weaknesses and needs. Fostering cooperation on the use of modelling data for climate adaptation, climate DTs’ CoPs may be able to ensure a continuous demand for the DT simulations while providing specific feedback and inputs to improve the generation of usable information.