D2.8 A guideline for end-user centred co-creation of Climate Services across Europe and beyond
Abstract
The I-CISK (Innovating Climate Services through Integrating Scientific and Local Knowledge) project, funded bythe European Commission (2020-2025), focused on developing a new generation of human-centred climateservices (CS) that are socially and behaviourally informed. By placing end-users at the centre of the design,development, and evaluation, I-CISK aimed to promote climate change adaptation and disaster risk reductionthrough co-created, viable, accessible, clear, credible, useful, and usable CS. Central to the project was the development, piloting, and refinement of a co-creation framework for CS (I-CISKWork package 2 Task 2.5). A prototype co-creation framework was devised at the beginning of the project andtailored to the composition and preferences of each LL. It was implemented in seven Living Labs (LLs) locatedin the Netherlands, Spain, Italy, Hungary, Greece, Georgia, and Lesotho. These LLs served as testbeds forparticipatory action research. The prototype co-creation phases were tested by Multi-Actor Platforms (MAPs)within the LLs, through iterative and inclusive collaboration between scientists, service providers, and endusers.LL leads were asked to regularly evaluate the product and process and participated in a final focus groupdiscussion for a reflexive monitoring exercise towards the end of the project. Based on this evaluation and other experiences communicated through project deliverables, an updated ICISKco-creation framework is created and presented in this report.
Full text
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101037293 Deliverable D2.8 A guideline for end-user centred co-creation of Climate Services across Europe and beyond September 2025
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101037293 Innovating Climate services through Integrating Scientific and local Knowledge Deliverable Title: A guideline for end-user centred co-creation of CS across Europe and beyond Author(s): Marthe Wens, Marije Schaafsma Contributing Authors(s): Micha Werner, Lucia de Stefano, Megi Gamtkitsulashvili, Ilyas Masih, Sumiran Rastogi, Gyorgyi Bela, Paolo Mazzoli, Schalk Jan van Andel, Debora Dotta Correa, Alexandros Ziogas Date 15/09/2025 Suggested citation: Wens, M., Schaafsma, M., et al., 2025: A guideline for end-user centred cocreation of Climate Services across Europe and Beyond. I-CISK deliverable 2.8 Availability: ☒ PU: This report is public [Please select] ☐ CO: Confidential, only for members of the consortium (including the Commission Services) Document Revisions: Author Revision Date Marthe Wens First Draft 19/08/2025 Lucia De Stefano, Megi Gamtkitsulashvili, Ilyas Masih, Marc van den Homberg Comments 24/08/2025 Micha Werner, Marije Schaafsma Full review 02/09/2025 Marthe Wens, Marije Schaafsma Final Draft 15/09/2025 Marthe Wens, Marije Schaafsma Including reviewer comments 29/10/2025
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond i Executive Summary The I-CISK (Innovating Climate Services through Integrating Scientific and Local Knowledge) project, funded by the European Commission (2020-2025), focused on developing a new generation of human-centred climate services (CS) that are socially and behaviourally informed. By placing end-users at the centre of the design, development, and evaluation, I-CISK aimed to promote climate change adaptation and disaster risk reduction through co-created, viable, accessible, clear, credible, useful, and usable CS. Central to the project was the development, piloting, and refinement of a co-creation framework for CS (I-CISK Work package 2 Task 2.5). A prototype co-creation framework was devised at the beginning of the project and tailored to the composition and preferences of each LL. It was implemented in seven Living Labs (LLs) located in the Netherlands, Spain, Italy, Hungary, Greece, Georgia, and Lesotho. These LLs served as testbeds for participatory action research. The prototype co-creation phases were tested by Multi-Actor Platforms (MAPs) within the LLs, through iterative and inclusive collaboration between scientists, service providers, and endusers. LL leads were asked to regularly evaluate the product and process and participated in a final focus group discussion for a reflexive monitoring exercise towards the end of the project. The co-creation process was widely appreciated for fostering mutual learning and trust, and for facilitating the integration of scientific data with local knowledge (LK) and user needs. However, its success varied across LLs. Key enabling factors included transparent communication, inclusivity fostering the integration of diverse climate risk experiences, and flexibility in adapting to new insights. However, challenges such as limited capacity and uneven responsiveness among stakeholders sometimes undermined legitimacy and ownership. Throughout the phases, co-evolution of knowledge regarding both CS product and co-creation process was observed among project partners, CS developers, CS providers, CS intermediaries, and CS consumers (the endusers). The final co-created pre-operational CSs, were generally scored high for usability, accessibility, and clarity. However, credibility, and particularly the effective communication of uncertainties, remained a concern for the CS in several LLs. Based on this evaluation and other experiences communicated through project deliverables, an updated ICISK co-creation framework is created and presented in this report. It offers a structured but adaptable roadmap for future CS development and comprises five phases: Phase 0: Build common ground and understanding Phase A: Co-ideate how CS can support decision-making and manage expectations Phase B: Co-develop climate product, bringing together local and global climate data and knowledge Phase C: Co-design climate service information system distributing the jointly produced climate information Phase D: Co-deliver and co-exploit the co-created pre-operational climate service These phases are complemented by an iterative evaluation cycle, supporting flexibility and enabling contextual adjustment and refinement. The final I-CISK co-creation framework supports climate change adaptation and disaster risk reduction decision-making by addressing barriers to the use of climate information and can be used by CS co-creation projects to contribute to the next generation of human-centred CS development.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond ii Table of Contents List of figures ....................................................................................................................................................... iv 1 Introduction .................................................................................................................. ............................... 1 1.1 Scope of this report within the context of the I-CISK project .............................................................. 2 1.2 Reading guide ....................................................................................................................................... 2 2 I-CISK‘s understanding of co-creation ......................................................................................................... 3 2.1 The I-CISK prototype framework .......................................................................................................... 5 3 Method used in this deliverable .................................................................................................................. 6 3.1 Data collection methods for the evaluation ........................................................................................ 6 3.2 Ongoing developments in the Living Labs ........................................................................................... 6 3.3 Product evaluation criteria .................................................................................................................. 7 3.4 Process evaluation criteria ................................................................................................................... 8 4 Evaluation of the co-creation steps ........................................................................................................... 10 4.1 Phase 0 – Building continuous engagement in the Living Labs ......................................................... 10 4.2 Phase A - Co-explore climate information needs and climate service desires .................................. 13 4.3 Phase B - Co-identify adaptation pathways and disaster risk reduction strategies to be supported by the CS 15 4.4 Phase C - Co-develop climate product containing tailored local and state-of-the-art regional climate data and knowledge ...................................................................................................................................... 17 4.5 Phase D - Co-design user-centred climate service system providing tailored climate information .. 19 4.6 Phase E - Co-evaluate co-created, user-centred climate service information system ...................... 22 4.7 Phase F – Co-deliver pre-operational climate service information system ....................................... 23 4.8 Co-evolution of knowledge and CS prototypes ................................................................................. 24 5 Evaluation following the co-creation process evaluation criteria ............................................................. 26 5.1 Active engagement (requires capacity, accountability) ..................................................................... 26 5.2 Inclusiveness (requires fairness, equitability) .................................................................................... 27 5.3 Trust (requires transparency and respect) ........................................................................................ 28 5.4 Constructive interaction, coordination, direction (requires relevance) ............................................ 28 5.5 Flexibility (requires openness and adaptability) ................................................................................ 29 5.6 Legitimacy (requires fairness) ............................................................................................................ 29 6 Evaluation following the product evaluation criteria ................................................................................ 30 7 Overall suggestions for improving the co-creation framework ................................................................ 31 8 Final I-CISK co-creation framework ........................................................................................................... 33 8.1 Co-creation framework overview ...................................................................................................... 33
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond iii 8.2 Phase 0 – Build common ground and understanding ........................................................................ 35 Phase objectives and scope ....................................................................................................................... 35 Core activities of this phase ....................................................................................................................... 35 Practical considerations ............................................................................................................................. 36 Relevant process evaluation criteria.......................................................................................................... 37 8.3 Phase A – Co-ideate how CS can support decision-making ............................................................... 38 Phase objectives and scope ....................................................................................................................... 38 Core activities of this phase ....................................................................................................................... 38 Practical considerations ............................................................................................................................. 40 Relevant process evaluation criteria.......................................................................................................... 41 8.4 Phase B - Co-develop climate product, bringing together local and global climate data and knowledge 42 Phase objectives and scope ....................................................................................................................... 42 Core activities of this phase ....................................................................................................................... 42 Practical considerations ............................................................................................................................. 43 Relevant process evaluation criteria.......................................................................................................... 43 8.5 Phase C - Co-design climate service information system distributing the jointly produced climate product .......................................................................................................................................................... 44 Phase objectives and scope ....................................................................................................................... 44 Core activities of this phase ....................................................................................................................... 44 Practical considerations ............................................................................................................................. 44 Relevant process evaluation criteria.......................................................................................................... 45 8.6 Phase D - Co-deliver and co-exploit the co-created pre-operational climate service ....................... 46 Phase objectives and scope ....................................................................................................................... 46 Core activities of this phase ....................................................................................................................... 46 Practical considerations ............................................................................................................................. 46 Relevant process evaluation criteria.......................................................................................................... 47 8.7 Co-evaluate climate product and co-creation process ...................................................................... 48 9 Conclusions ................................................................................................................... ............................. 49 References ........................................................................................................................................................... 1
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond iv List of figures Figure 1: From a prototype to a tested user-centred co-creation framework for climate services ................... 2 Figure 2: The protype I-CISK co-creation framework. (Source: I-CISK MS10, Wens et al., 2022) ....................... 5 Figure 3: Product criteria for successful co-creation (I-CISK MS10) .................................................................... 7 Figure 4: Process criteria for effective Climate Services (I-CISK MS10) ............................................................... 8 Figure 5: Perception of successful completion of the initiation phase (own survey) ....................................... 10 Figure 6 (left): Stakeholder groups included in the MAP .................................................................................. 11 Figure 7 (right): Frequency of methods used in phase 0 (own survey) ............................................................. 11 Figure 8: Perception of achieving the enablers in the initiation phase (own survey) ....................................... 12 Figure 9: Frequency of methods used in phase A (own survey) (left) .......................................................... 14 Figure 10: Decision processes that require tailored climate information identified in each MAP (own survey) (right) ................................................................................................................................................................. 14 Figure 11: Self-evaluation of phase A activities (own survey) ................................................................. .......... 14 Figure 12: Frequency of methods used in phase B (own survey) ...................................................................... 16 Figure 13: Climate adaptation options and disaster risk reduction actions co-identified in each LL (own survey) ........................................................................................................................................................................... 16 Figure 14: Perception of achieving the enablers in phase B (own survey) ....................................................... 16 Figure 15: Frequency of methods used in phase C (own survey) ...................................................................... 18 Figure 16: Perception of achieving the enablers in phase C (own survey) ....................................................... 18 Figure 17: Frequency of methods used in phase D (own survey) ..................................................................... 20 Figure 18: Perception of achieving the enablers in phase D (own survey) ....................................................... 21 Figure 19: Ways how the consideration of local knowledge has informed the CS development process (own survey) ............................................................................................................................................................... 25 Figure 20: Perception regarding active engagement of MAP members during the co-creation phases (own survey) ............................................................................................................................................................... 26 Figure 21 (left): Perception regarding presence of resources for active engagement of MAP members (own survey) ............................................................................................................................................................... 26 Figure 22 (right): Perception regarding accountability over the co-creation process and product (own survey) ........................................................................................................................................................................... 26 Figure 23 (left): Perception regarding inclusivity of the MAPs (own survey) ................................................... . 27 Figure 24 (right): Perception regarding diversity within the MAPs (own survey) ............................................. 27 Figure 25 (left): Perception regarding information sharing about the I-CISK project within the LL (own survey) ........................................................................................................................................................................... 28 Figure 26 (middle): Perception regarding the network building character of the MAPs (own survey) (legend left) .................................................................................................................................................................... 28 Figure 27 (right): Perception regarding trust building within the MAPs (own survey) (legend left). ............... 28 Figure 28: Perception of LL leads regarding the experiences and atmosphere during constructive interactions (own survey) ...................................................................................................................................................... 28 Figure 29: Perception regarding trust flexibility within the MAPs (own survey) .............................................. 29 Figure 30: Perception regarding the legitimacy of the co-creation process within the MAPs of the LLs (own survey). .............................................................................................................................................................. 29 Figure 31: Overall performance of the co-delivered prototype climate services in the Living Labs (own survey) ........................................................................................................................................................................... 30 Figure 32: Piloted co-creation framework for user-centred climate services, based on I-CISK experiences .... 33 Figure 33: Phases in the piloted (tested, evaluated, updated) I-CISK co-creation framework ......................... 34
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond v Glossary Acronym Definition CCA Climate Change Adaptation CS Climate service(s) CSIS Climate service information system DRR Disaster risk reduction FGD Focus group discussion I-CISK Innovating Climate services through Integrating Scientific and local knowledge LK Local knowledge LL(s) Living Lab(s) MAP(s) Multi-actor platform(s) – the collection of actors in the living lab MEL Monitoring, learning and evaluation MS Milestone – a project milestone as defined in the I-CISK proposal ToC Theory of Change WP Work Package – a combination of tasks as defined in the I-CISK proposal
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 1 1 Introduction Central to the European Commission funded I-CISK (Innovating Climate services through Integrating Scientific and Local Knowledge) project (2020-2025)1 was to develop next-generation climate services (CS), following a social and behaviourally informed approach for co-producing CS that meet the climate information needs of citizens, decision makers and stakeholders, at the spatial and temporal scale relevant to them. I-CISK recognized that to achieve behavioural change, the active use of climate information in informing decisionmaking toward climate change adaptation and mitigation requires that citizens, stakeholders and decisionmakers are at the centre of value creation; of the design, production, delivery and evaluation of CS, and that they co-create the value of this CS for society. A critical starting point for the I-CISK project was to develop a framework for co-creating human-centred CS by co-exploring user needs, incorporating social and behavioural factors and integrating LK with scientific data, tailoring climate information based on the adaptation options. Suhari et al. (2022) note that “the number of conceptual frameworks, methods and practical tools for the joint production of CS are continuously growing. It is a challenge to set up a consistent co-creation approach for a specific research context and being competent about its application as well as its theoretical and methodological underpinnings”. Therefore, in ICISK, we co-produced, piloted, and improved a framework (I-CISK MS10: Wens et al., 2022) on how to set up action research and ensure a holistic participatory process to co-create human-centred CS in and with the users of seven Living Labs (LLs) in diverse climatic and socioeconomic regions in Europe and Africa (Rijnland - Netherlands, Los Pedroches – Andalucia Spain, Italy, Hungary, Crete - Greece, Alazani Basin – Georgia, and Lesotho) (Masih et al., 2022). This deliverable reports on the development and evaluation of the prototype user-centred co-creation framework, and then presents the final I-CISK framework for co-creating next-generation, human-centred CS. It contains a comprehensive assessment of the prototype co-creation framework developed at the start of the research project and builds upon the research practices and experiences with constructive interactions throughout the I-CISK project. The prototype framework supported a tailored co-creation process and was developed by combining best practices found in different strands of literature, including CS literature and design thinking, and building on relevant frameworks including Information and Communication Technology for Development (ICT4D; Heeks, 2017), and Weather and Climate Services for Africa (WISER; Carter et al., 2019). Several steps in the co-creation process were distinguished, all of which were executed by a diversity of stakeholders (scientists, producers, end-users, decision makers) in the seven I-CISK LLs. As such, the prototype served as a guideline to integrate existing physical and socio-economic datasets, global data sources and existing services (e.g. Copernicus, GEOSS, C3S, ESA), in-situ datasets and innovative new data sources identified during the project. The updated co-creation framework, presented in this report, scrutinizes the context-specific best practices and varied experiences from the different LLs, based on iterative evaluations and reflection on the co-creation process of LLs and Multi-Actor Platforms (MAPs) members that were established in each of the LLs. The prototype co-creation framework is slightly revised and extended with tested methods and approaches that fit the contexts of our seven LLs best. The aim of the updated co-creation framework for user-centred CS is to serve as a guideline for the co-design and co-development of CS across Europe and beyond. 1 Https://icisk.eu
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 2 1.1 Scope of this report within the context of the I-CISK project The first of six main objectives of the I-CISK project (I-CISK, 2021) was “to develop a framework for coproducing next-generation, human-centred CS by co-exploring user needs, incorporating social and behavioural factors and integrating LK with scientific data, tailoring climate information based on the adaptation options.” Within the I-CISK project, Work Package (WP) 2 was dedicated to the co-development of human-centred CS through setting up and executing a user-centred co-creation framework, including guidelines on how to identify user needs, integrate LK and behavioural factors. Task 2.5, to which this report pertains, included both the development and evaluation of the prototype co-creation framework and the development of a final, piloted guideline for co-creating CS (Fig. 1). This report thus contains a comprehensive assessment of the prototype developed at the start of the research project. Building upon this evaluation as well as on the research practices and experiences throughout the ICISK project (especially WP2, but also WP1 where the LLs and MAPs were established, and WP3 that focussed on data integration), improvements are suggested and implemented, resulting in the final I-CISK framework for co-creating next-generation, human-centred CS. It should be noted when interpreting the findings, that at the time of data collection for this evaluative analysis (summer 2025), many of the seven LLs were still refining their CS; this report therefore does not provide an assessment of the final I-CISK CS products. Whilst it is mentioned in the project proposal that Task 2.5 would also provide “guidelines for integrating existing physical and socio-economic datasets, global data sources and existing services”, this aspect is described in WP3 Deliverables 3.3 (Pesquer et al., 2024) and 3.5 (Van Andel et al., 2025). The co-creation approach to codeveloping the actual CS solutions takes place in WP5 through so-called “Climate Service Task Forces” and is detailed in D5.2 (Bagli et al., 2024). Figure 1: From a prototype to a tested user-centred co-creation framework for climate services 1.2 Reading guide This deliverable includes an introduction to the concept of co-creation, encompassing the prototype developed in MS10 (Chapter 2), and an elaboration of the methodologies employed to critically evaluate the prototype (Chapter 3). Chapter 4 provides an assessment of the phases of co-creation, while Chapter 5 addresses the evaluation of the associated processes. Chapter 6 includes an evaluation of the pre-operational CS products created in the LLs following the co-creation phases, and Chapter 7 compiles all recommendations for enhancing the prototype framework. Finally, Chapter 8 outlines the revised framework for the co-creation of user-centred CS, and Chapter 9 offers the report's conclusion.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 9 demands and contextual circumstances. The project should ensure fair representation of these stakeholders throughout the process, and their partnership in the process of producing a CS. There should be space for multiple stakeholders to share expertise and challenge each other's views and contributions through negotiations, so aiming for an equitable distribution of power and influence between different stakeholders is critical. This requires logistics, including project communication, to happen in an accessible way and through equitable negotiation and compromises. Given the emphasis in the transdisciplinary co-creation processes of I-CISK, which aims to bring local and scientific knowledge together, it is important for I-CISK researchers to be aware of epistemological approaches and reflect on the role of academic science. It involves changing each other’s perspectives through negotiations and dialogue. When scientists and users try to deliberately co-produce knowledge, they can have very different, if not irreconcilable, ideas about what constitutes credible, relevant, and usable science, which hampers dialogues and real inclusiveness. ●Trust (requires transparency and respect) - The process should ensure that all stakeholders are involved in all parts of the co-creation process, and can understand, share and contribute to details of the process and the outcomes. This means that information is open and shared (when desirable, e.g. sensitive information or Intellectual Property (IP) protected data are excluded from this sharing), and that decisions are clearly explained to parties involved. A common ground (shared understanding) regarding languages and terminologies should be created. Moreover, the process should be respectful towards participants’ divergent values and beliefs, unbiased in its conduct, and fair in its treatment of opposing views and interests. This way, trust can be built. ● Legimacy (required fairness) - The core stakeholders and the final CS should also be seen as legitimate by the end-users, based on their integrity, their performance over time, persistence and effectiveness. For example, knowledge production may be seen as fairer if it takes place independent from political or commercial interests while appropriately considering the values, concerns, and perspectives of the different actors. This criterion also involves whether the process is seen as legitimate, where the core actors have a clear mandate and are committed to delivering a valuable product, and different stakeholder groups feel that their values and interests are represented, making the end-product acceptable.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 10 4 Evaluation of the co-creation steps In this section, we evaluate the different steps (phase 0 to phase E) as defined in the I-CISK prototype cocreation framework (Wens et al., 2022), focusing on the process as it evolved within the LLs and the overall ICISK project. It also presents the evaluation of the co-evolution of knowledge and CS product. 4.1 Phase 0 – Building continuous engagement in the Living Labs Prototype framework description of phase 0 (MS10, Wens et al., 2022) In the initiation phase, different LLs covering a range of relevant sectors and regions are assembled, and the engagement of a variety of end-users in these LLs is identified. This is a collaborative effort supported by snowballing to reach a diverse group of stakeholders. Once a group of engaged stakeholders has been established in a MAP, a roadmap for the process should be established, with a set of envisioned activities, outputs and goals. Common definitions must be agreed upon. Moreover, roles and responsibilities should be shared, and a point of contact should be designated. In I-CISK, LLs with their MAPs form the foundation of successful co-creation. This step focuses on empathy and trust building: who are potential end-users; what is important to them? It cements the relationships and understanding between actors and creates a space where jointly defined issues can emerge. In this phase, the LLs and their MAPs were established, either from scratch or based on existing networks, and initial meetings with the MAP were held to gain insight into the decision-making context, primarily through dialogues and small-scale meetings. On average, eight meetings were devoted to this phase. Most LLs reported successfully completing the inception phase (Fig. 5); they reported successfully clarifying decision-making processes and establishing agreed-upon communication channels. This phase was concluded for all LLs by developing a draft roadmap (with a tailored action plan and contact points) for implementing the co-creation framework (Werner et al., 2022). However, contextual changes in LLs Alazani Basin (Georgia) and Los Pedroches (Andalucia, Spain) hindered follow-up activities. These changes affected both citizen-level translation and regional integration of CS in the LL Alazani Basin and requiring concerted collaborative efforts with MAP members to ensure the long-term sustainability of the CS in LL Los Pedroches. Figure 5: Perception of successful completion of the initiation phase (own survey) Enablers and methods used 0% 50% 100% Is it clear how decisions in the LL are being made to all actors? Is it clear how actors in the LL will communicate with each other? Is it clear to the scientists in the LL what decisions the actors need support for through CS? Do all actors in the LL have the same understanding of the intention of the co-creation process? Do all actors in the LL have the same understanding of the desired outputs and outcomes of the co-creation process? Do all actors in the LL understand the motivations and needs of each other in this co-creation process? No. of LL yes a bit no
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 11 To promote inclusivity, as a factor enabling successful co-creation, LLs invested time in identifying key stakeholders through stakeholder mapping and literature research (Fig. 7), resulting in a diverse range of stakeholder groups involved (Fig. 6). Ensuring that all stakeholders were directly or indirectly involved was achieved within the first year of the project (except in one LL that was still being established at the time of reporting). LL leads reported including a diversity of voices regarding gender, sectors, and hierarchy, although in some MAPs (notably Rijnland and Lesotho), gender balance could not be achieved (Ch 6.4). To foster constructive interaction, coordination, and direction, the LLs also made significant efforts to manage expectations (e.g., regarding the number of CS to be developed). The evaluation revealed that many LLs followed an iterative process, first meeting individually with stakeholders to establish common ground, and then holding general meetings with stakeholders to discuss shared goals and foster mutual understanding of the challenges they faced (Fig. 7). Both approaches contributed to understanding the demand for novel CS in the LLs and demonstrated a clear interest among stakeholders to participate. It also helped in gaining insight into competing priorities and the broader decision-making context. Understanding and agreement on the outcomes of a co-creation process was essential for transparent communication with stakeholders and for adequately managing expectations, which in turn contributes to building trust. The communication methods used consisted of balanced involvement, giving everyone a voice, and providing clarity about the actions taken. A mailing list and periodic informative emails were mentioned as a functional horizontal communication channel, and a Miro board was considered an effective way to summarize expectations and desired outcomes in LL Lesotho. To foster understanding of each other's needs and motivations, LL Rijnland successfully implemented a user story writing workshop with mixed groups of stakeholders. Figure 6 (left): Stakeholder groups included in the MAP Figure 7 (right): Frequency of methods used in phase 0 (own survey) To build trust and promote transparency, efforts were also made to translate and simplify terminology (for example, in the LLs Hungary and Alazani Basin, where the word CS does not exist in the local languages). Time (at least six months of ongoing dialogues) and resources were key factors in the success of this. At the same time, however, it was observed that in most LLs, complete mutual understanding between all MAP partners had not yet been achieved before moving to the next step of the co-creation phase, and that creating a common ground was an ongoing process (Fig. 8). Regarding active engagement, the evaluation suggests that more attention should have been paid to the capacities of actors and their limited resources (Fig. 8: only in half of the MAPs was the available capacity assessed positively, and no LL conducted a full resource analysis). Time and resources remained a constraint, and it was important to acknowledge that experts are busy and have to balance participation in LL MAP 0 1 2 3 4 5 6 7 Policy Makers Business and Industry Civil society organisations Reclamation consortia Environmental Agencies Non-governmental organisations Research and academia 01234567 Literature review Stakeholder mapping Dialogues and small meetings Large meetings with many actors Living Lab resource analysis Questionnaire Capacity analysis User stories No. of LL that applied this method
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 12 meetings with other work, and that there is diversity among actors and therefore diversity in capacity and resources (some have resources available and are more reactive and flexible). A suggestion for future projects would be to develop a table with information on end-users, producers, intermediaries, final end-users and their role in adaptation decision-making processes, as well as the resources of each of these actors. This could be a starting point for a discussion about who should be involved in the MAP and when (and what) each MAP member could contribute during the co-creation process. Such an overview of capacities and resources can be used for accountability in subsequent steps of the process. Outcomes Evaluating the whole inception phase, the results show that the stakeholders involved in the LLs were not yet familiar with co-creation processes at the end of the phase. Figure 8 shows that in most MAPs, stakeholders still did not fully understand the purpose of co-creation, and that this understanding was not achieved, particularly in LL Alazani Basin. There was also some uncertainty among the LL stakeholders regarding the desired outputs and results of the co-creation process, partly due to a further lack of clarity about each other's needs. This emphasized the importance of focusing on direction in subsequent phases to ensure continued active involvement. Figure 8: Perception of achieving the enablers in the initiation phase (own survey) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Do the actors in the LL have enough time, resources and facilities to deliver the quality and be able to actively engage throughout the project? Do the actors in the LL understand each other well enough in terms of definitions, concepts and terminology? Do the actors in the LL have the necessary capacities/skills to engage in co-production? Are all actors in the LL part of decision making with respect to LL activities? No. of LL yes a bit no
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 13 4.2 Phase A - Co-explore climate information needs and climate service desires Prototype framework description of phase A (Source: MS10, Wens et al., 2022) This step aims to define the user needs through identifying gaps in climate information to support decisions and barriers to using existing CS. In the co-exploration phase, a set of data and information needs, including CS-related sources, formats and modes of dissemination that are relevant for the local context are collected. This is done to match user requirements with technical possibilities and make decisions on the within-scope and out-of-scope needs. To highlight the interplay between new knowledge added by the CS and existing user knowledge, it is necessary to conduct a detailed context mapping, needs assessment and an evaluation of the existing available products together with all stakeholders. This is an iterative, interactive step, as needs can alter because of the co-creation process itself. This iterative nature makes it possible to change objectives and subsequent aligning expectations throughout the process In this step, all LLs prepared detailed needs and priority assessments through context-specific iterative processes. Seven to ten different types of meetings were held with MAP members per LL to assess the state of knowledge, strengths, and weaknesses of available CS, as well as the context in which they are being implemented. A good description of what end-users currently do and how they do it was part of this phase and aided in this process. This iterative process was incorporated into the I-CISK project in task 2.1 and synthesized end-user desires regarding climate information, CCA and DRR options, and resilience (elaborated in Egan et al., 2025, and Baugh al., 2025). Enablers and methods used This phase required active engagement and took approximately seven to ten collective and bi-lateral meetings in each LL. In one LL, this phase lasted about a year, which some MAP members considered quite long. However, this time also allowed the LLs to ensure that all voices were heard, even in the presence of more involved, dominant stakeholders, and to cautiously go beyond conventional practice. To further support inclusiveness, many LL meetings were held with all MAP members. However, in two LLs, collaboration took only place with actors who were the intended end-users. Supporting trust, all but one LL reported that meetings had an open agenda, which facilitated the creation of mutual understanding and awareness. Considerable effort was made to make the materials accessible (in understandable local language) and to collaboratively develop communication materials, such as fact sheets, to facilitate this phase, especially where MAP members lacked technical expertise. Interactive tools such as Miro boards and other online platforms were used to support collaboration and transparency throughout the process. Communication efforts, including a webpage and Facebook campaigns, were also implemented to encourage broader engagement. To promote mutual learning among diverse stakeholder groups, including scientists, several well-considered approaches were employed (Fig. 9): interactive formats such as plenary sessions, workshops with mixed sectoral breakout groups, and open online forums where stakeholders could exchange views on needs, priorities, and intended deliverables. One LL asked end-users to give short presentations on their needs, presenting a specific set of questions and answers. These discussions allowed for the necessary input and priorities of diverse stakeholders to be further described in their own words. Moreover, most MAPs used user-narratives and interviews to map information needs, detailing variables, time, and spatial scale and resolution. This helped make these needs highly specific. In the I-CISK project, LLs were trained in creating user narratives by one of the project partners during a general meeting and through online training sessions. Field research was also frequently conducted, often to gain insight into the extreme weather challenges faced by the LLs.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 14 Figure 9: Frequency of methods used in phase A (own survey) (left) Figure 10: Decision processes that require tailored climate information identified in each MAP (own survey) (right) Outcomes On average, more than three decision-making processes requiring tailored climate information were identified in each LL (Fig. 10). These identified needs included a water runoff forecasting system, mediumand long-term climate forecasts, and the relationship between phenology and climate. In two of the seven LLs, the needs of the various end-users were not consistent, resulting in multiple use cases. Some LLs reported that it was difficult for MAP members to articulate their needs. A good description of what end-users currently do and how they do it (e.g., through decision timelines) facilitated a deeper understanding of desires and preferences. In the evaluation survey, most LLs indicated that discussions were conducted in an understandable manner (Fig. 11). Ensuring active involvement worked well in half of the MAPs. However, at the end of this phase, differences across LLs could be observed in terms of their ability to define goals. Some LLs already had a prototype CS that could be used as a starting point or were able to define concrete objectives. Other LLs started the I-CISK project with abstract goals, and at the end of this phase, the intended CS remained too vague. This result can in part be attributed to the approach followed in the I-CISK project, where the emphasis in this step was on identifying and articulating needs without having a specific CS in mind. This approach aimed to avoid specifying CS options at this stage, to prevent CS developers from rushing into proposing CS solutions to end-users without carefully listening to the needs and decision-making processes that required CS support. Figure 11: Self-evaluation of phase A activities (own survey) field research participatory appraisal living documents user survey user interviews user stories user workshops sector specific sessions 0246 No. of LL 0 1 2 3 4 5 6 Georgia Spain Italy Lesotho Crete Netherlands Hungary Decision processes identified 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% To what extent did all the stakeholders in the LL actively engage with the project? Did you ensure that discussions around needs and information inputs were done in a manner that the stakeholders would understand? Is it clear in the LL what added value a new CS developed by I-CISK could generate? yes a bit no
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 15 4.3 Phase B - Co-identify adaptation pathways and disaster risk reduction strategies to be supported by the CS Prototype framework description of phase A (MS10, Wens et al., 2022) After the context-specific climate challenges have been identified and building further on the gap analysis of step A, this step provides an opportunity for stakeholders to ideate. Collective brainstorming should result in a multitude of decisions (on policies, actions, and measures) that can be supported by the CS. In the coidentification phase, existing plans for CCA and DRR and potential new strategies or specific measures that can benefit from decision-support through CS, are mapped out. In this phase, these adaptation and risk reduction interventions are assessed, to prioritize and rank them. This is critical to ensure that the final CS supports climate decisions relevant to user objectives. In addition, it is important to establish the time scales of these adaptation or risk reduction objectives and to develop approaches that help achieve these objectives, tailored to the needs of the stakeholders. This step involves collaborative knowledge exchanges with a variety of stakeholders to include local experiences, perceptions and concerns regarding climate and risk management decisions. Importantly, as part of this step, requests regarding timing of information provision, the type of data and the level of spatial and temporal aggregation must be identified. To map the mitigation and adaptation measures, plans, policies, and strategies to be supported by the I-CISK CS, adaptation pathways were jointly defined within the LLs using a variety of methods (Hernández-Mora et al., 2023; De Stefano et al., 2024). It was recognized by multiple LLs, e.g., in LL Italy, that translating the CS into more practical decisions was also a topic in previous and subsequent phases. This ensured the relevance and applicability of the CS to be developed within the local context. Enablers and methods used Involving local expertise in the process provided greater insight into user preferences for CCA and DRR to be supported by the CS and helped identify additional needs of various stakeholders. However, end-users were not always willing to answer questions about specific plans (making this a time-consuming process). By emphasizing understanding institutional mandates and responsibilities (making it clear who is responsible for what), more targeted and active engagement was possible and facilitated. Moreover, MAP participants were open to the flexible aspects of phase B. The iterative approach allowed the co-creation framework to evolve in response to new insights, ensuring strategies remain relevant as circumstances change, fostering constructive interaction. However, this phase showed challenges with respect to transparency, as most LLs reported that not all materials were available in non-technical language (Fig. 14). While literature research was primarily used to review local adaptation studies and map relevant policies and frameworks, this identification phase was informed by both scientific data and LK (Fig. 12). Impact studies, combined with their professional mandates, enabled some stakeholders to easily identify adaptation needs and plans, while for others, further development was required in LL Crete. In some LLs, the concepts of DRR and CCA were insufficiently explored in academic research (E.g. LL Alazani-Lori), making the inclusion of LK crucial to validate climate information at a larger scale. In such cases, a larger data collection effort, including iterations between interviews, field visits, FGD and workshops provided a wealth of information regarding CCA and DRR measures adopted. Through individual user interviews (Fig. 12) with each stakeholder, adaptation goals, plans, and strategies were established, which were used to validate data from the literature and emerged from both individual (bilateral) and collective stakeholder interactions. Interviews were primarily used to gather detailed, individual perspectives and gain a better understanding of specific adaptation options, while FGD were used to explore decision-making processes and collaboratively characterize adaptation pathways, comparing perspectives.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 16 Workshops were used to introduce tools and validate previous findings, but also (like FGD and bilateral meetings) to verify previous findings and refine interpretations. Figure 12: Frequency of methods used in phase B (own survey) Outcomes In all LLs, the phase concluded successfully with the identification of sectorand context-specific adaptation pathways. On average, six different adaptation options were discussed in each LL (Fig. 13), sometimes categorized into overarching clusters by sector or timeframe. Sectors covered included the water sector itself (with water pricing, desalinisation and maintenance plans), agriculture (with crop choices, livestock and feed management, irrigation plans), transport (with heat and de-icing management plans), and tourism (with water (re-)use and coastal protection plans to be supported by prototype CS). Overall, these pathways are judged to be sector-specific, reflecting local and/or sectoral visions, and tailored to the context-specific challenges and risks (Fig. 14) Figure 13: Climate adaptation options and disaster risk reduction actions co-identified in each LL (own survey) Figure 14: Perception of achieving the enablers in phase B (own survey) 02468 User interviews Workshops Literature search Focus groups Decison timelines LL network maps Discussions with users No. of LL 0 2 4 6 8 No. of CCA and DRR actions co-identified 0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100% Was the socio-economic and socio-cultural adaptation decision context well researched? Are the co-identified adaptation context specific, tailored to context-specific challenges and risks? Are the co-identified adaptation pathways sectorspecific, reflecting local, sectoral visions? Are different relevant materials available in local and non-technical language, and understandable for all… Did the process of co-identifying adaptation pathways build on co-creation phase A? No. of LL yes a bit no
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 17 4.4 Phase C - Co-develop climate product containing tailored local and state-of-the-art regional climate data and knowledge Prototype framework description (MS10, Wens et al., 2022) After the context-specific climate challenges have been and building further on the strategies to be supported of step B, this step too allows stakeholders to ideate. Collective brainstorming should result in a multitude of climatic parameters, thresholds, and climate knowledge, to be integrated into the climate product. In the codevelopment phase, interests (strategies to be supported) are translated into a climate product (with relevant scenarios, time scales, triggers, etc.). It is a process in which providers and end-users work together (often with the help of intermediaries) to combine different knowledge, skills and practices to create new, relevant knowledge that meets the needs of end-users, and that addresses a shared concern. This step thus involves the integration of stakeholder knowledge and experiences (joint combination of the diverse knowledges of the stakeholders) and matching local observations (e.g. through citizen science) with scientific climate data (modelled, projected, e.g. from Copernicus, S2S Prediction project, EMODnet, GEO, ESA Actions) which will enhance the accuracy and acceptance of the CS. End-users within the LLs actively participated in exploring various types of knowledge and experiences related to the current climate system and its associated risks, as elaborated D2.4 (Van den Homberg et al., 2024). The objective is to link expertise from the consortium scientists and LK from the LLs and complement climate data from Copernicus and GEOSS and research with local data, within the particular social, economic, and sectoral contexts of the LLs. Data collection methods, elaborated in D2.2 (Van den Homberg et al., 2022, Ch. 3) were employed to understand the requirements for climate products in relation to decision-making. Based on these insights, climate metrics were co-developed to incorporate local experiences (e.g., habits regarding thresholds) which were consecutively validated (Van den Homberg et al., 2024). Enablers and methods used Decision timelines were employed as a structured framework to understand the sequence and timing of decisions within a community. By mapping out when and why certain decisions were made, insight into the contextual factors that influence choices was gained. The presentation of available climate data and services— sometimes through serious games, other times through dialogues—and the co-testing of forecast reliability and uncertainty visualization (conducted in six LLs) stimulated discussions on useful and potentially useful climate parameters, their limitations, and how uncertainty can be communicated to support decision-making (Fig. 15). This approach was reported to build trust and reduce unrealistic expectations. The decision-time-line exercise was a valuable tool for identifying actual climate data needs between end-users and decision-makers, and for demonstrating potential uses of available climate data. However, maintaining active engagement among MAP members in this phase proved challenging in some cases (e.g., LL Crete). Additionally, a literature review was conducted to explore effective methods for integrating LK into CS, ensuring contextual relevance. This led to a well-defined dataset of integrated information, including its limitations (Fig. 16). While this specificity is a strength, it also poses challenges for mainstreaming the results of the co-creation process beyond the LL area, as they are highly tailored to the LL needs.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 18 Figure 15: Frequency of methods used in phase C (own survey) Outcomes On average, seven meetings were held during this phase. These meetings led to a clear understanding among end-users of how to adapt the CS to their needs. Overall, this phase was considered the easier part of the process, building on the foundations laid in Phases A and B. Collaboration with technically skilled MAP members proceeded smoothly (e.g., LL Italy), but asymmetries in skills, and consequently in decision-making power, posed risks. In LL Alazani Basin, the differences in capacity affected the extent of participation, for example, more technical institutional users led the agenda-setting for the CS development process. However, through the co-creation process, I-CISK LL leads tried to bridge this gap by providing space to other members of the MAP to provide input and feedback. In one case (LL Lesotho), the I-CISK CS was integrated into an existing CCA protocol, meaning that certain data decisions had already been established. LK informed the prototype CS development in multiple ways, from a deeper understanding of climate information needs (building on Phase A), decisions to be supported by prototype CS and local capacities for adaptation (building on phase B), but also provided data and knowledge on local perceptions and knowledge of climate and weather, risks and impacts. The activities focused on LK also helped validate scientific data, but the inclusion of LK was not overall reported to be incorporated to the extent initially aimed for. However, a challenge emerged in some LLs in maintaining the promises made during Phases A and B. Integrating L or data into climate models was not always straightforward due to inaccuracies, lack of standardization, and differing subjective perceptions. For example, in LL Los Pedroches, although local data existed, it often lacked the necessary functional characteristics (e.g., length, reliability) for integration, which lead to some disappointment among MAP members. In LL Hungary, relevant data was generally scarce, particularly because much of it was not open source. As a result (Fig. 16), combining local and global data, models, and knowledge remained difficult. To cope with this, qualitative data was recognized as a critical source of information for developing CS when local quantitative data was unavailable. Figure 16: Perception of achieving the enablers in phase C (own survey) 02468 Questionnaire on the role of weather and climate in decision making Discussions to discuss forecast reliability and limitations Discussions to convey content and uncertainty of climate projections User-driven evaluation of existing & tailored CS Literature review on integrating local knowledge No. of LL 0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100% Have you build on the understanding of users' climate information needs (Step A and B) to produce climate metrics (step C)? Have you explored the limitations for data and information in the stakeholder’s geographic location of interest? Have you co-created the climate knowledge present in the climate service by combining local data/models/knowledge as well as global… No. of LL yes a bit no
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 25 Figure 19: Ways how the consideration of local knowledge has informed the CS development process (own survey) 02468 Identifying complementary needs among the MAP members Identifying key decisions/activities that will be supported by CS Understanding user capacities and preferences Grounded understanding of and risks and impacts Better spatial and/or temporal understanding of information needs Validation of scientific data (through local datasets or local knowledge) Contributing new data that can integrated within CS Understanding local perceptions of short and long term changes in the weather and climate Identification of trustworthy channels for communication and dissemination No. of LLs for which this applies
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 26 5 Evaluation following the co-creation process evaluation criteria In this section, we evaluate the co-creation process as it evolved within the LLs and the overall I-CISK project using the process evaluation criteria (Ch. 3.4). It is important to note that the evaluation reflects a generalisation across LLs of the overall results, but that the LLs showed considerable variation in their development and performance. 5.1 Active engagement (requires capacity, accountability) In the overall I-CISK project, members of the MAP were active in all phases, with over half of the LLs reporting constant or frequent presence of all stakeholders (Fig. 20). Only in the co-delivery phase, this was significantly lower, mainly because this phase had at the time of writing not been concluded in some LLs. It was reported that intermediaries could hamper active engagement of end-users, as it limited more direct contact with these end-users (LL Lesotho). Closer physical distance to the LLs by project partners enabled active engagement. Looking at the availability of resources (Fig. 21), for the capacity to actively engage, time was seen as a large bottleneck that was managed as well as possible (“fairly sufficient”), while resources were not or barely sufficient in two LLs (Hungary, Lesotho). All LLs succeeded in creating clarity on roles and responsibilities. However, this did not always translate into accountability towards the process and end product (Fig. 22): LLs Hungary and Alazani Basin reported that this was not (yet) achieved. It may be, however, that accountability will be created when the CS become operational, for instance in Georgia when the product is transferred to the National Environmental Agency. Figure 20: Perception regarding active engagement of MAP members during the co-creation phases (own survey) Figure 21 (left): Perception regarding presence of resources for active engagement of MAP members (own survey) Figure 22 (right): Perception regarding accountability over the co-creation process and product (own survey) 0% 20% 40% 60% 80% 100% Co-identification Co-exploration Co-development Co-design Co-evaluation Co-delivery No. of LLs Active engagement Always Frequently Sometimes Rarely Never 0% 20% 40% 60% 80% 100% No. of LLs Presence of resources Not sufficient Barely sufficient Fairly sufficient Moslty sufficient Completely sufficient 0% 20% 40% 60% 80% 100% Clarity on roles and responsabilities Feeling accountable Satisfaction with contribution No. of LLs Accountability poor moderate excellent
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 27 5.2 Inclusiveness (requires fairness, equitability) In most LLs, a fair to good diversity in MAP members was achieved (Fig. 23). Creating a balance in gender was not always achieved, while a variety of stakeholder experiences with climate risks was included in all MAP, except in LL Hungary. All reported that the diverse values and interests were considered throughout the project by creating equal opportunities to contribute (only LL Hungary faced some barriers here) (Fig. 24). However, it remained a challenge to ensure an equitable distribution of power and influence, especially in LL Lesotho. Figure 23 (left): Perception regarding inclusivity of the MAPs (own survey) Figure 24 (right): Perception regarding diversity within the MAPs (own survey) While this diversity in representation adhered to the intended inclusiveness of the co-creation approach (see Ch. 3.4. Process evaluation criteria), it was felt that this inclusive representation resulted in concerns around institutional and economic efficiency, thus viability, for some LLs. When only one decision-making level was involved, co-creation could be managed more tightly, potentially accelerating progress toward effective CS. The LL Lesotho included the Lesotho Red Cross Society as the main end-user, which provided focus in the cocreation process focused. In contrast, tailoring services to multiple governance levels proved to introduce complexity, stemming from challenges in aligning diverse stakeholder needs and navigating institutional structures. LLs involving multiple user levels (e.g., LLs Alazani Basin and Los Pedroches) faced greater complexity in developing effective, end-user-centred CS. In LL Los Pedroches, for instance, it remained uncertain whether the CS will reach farmers, considered the most distant end-user level. Similarly, in LL Rijnland, the diversity of stakeholders made it difficult to meet all needs. In LL Alazani Basin, the CS was at the time of evaluation still too technical for citizens but served the National Meteorological Services well. To address this, LLs Los Pedroches and Crete created ‘Umbrella services’ that bring together multiple bespoke indices for different users or sectors. LLs Alazani Basin, Italy, and Rijnland targeted one, or a small number, of main CS users, the co-creation could be managed more tightly potentially accelerating progress towards effective CS. 0% 20% 40% 60% 80% 100% Equal opportunities to contribute Equitable distribution of power and influence Values and interests taken into account No. of LLs Inclusivity very good good fair poor very poor 0% 20% 40% 60% 80% 100% Age Gender Regional representation Vulnerability to the climate risks No. of LLs Diversity very good good fair poor very poor
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 28 5.3 Trust (requires transparency and respect) In I-CISK, information about the project was shared openly with all MAP members (Fig. 25) and the project was able to strengthen existing networks and build new ones (Fig. 26). This led to moderate to excellent trust building in all LLs (Fig. 27), with attention to the full comprehension of the process and product among all MAP members. Overall, this enabler was evaluated very positively. A key learning is to account for cultural differences, for example regarding the right way of approaching partners and how to organise dialogues. Figure 25 (left): Perception regarding information sharing about the I-CISK project within the LL (own survey) Figure 26 (middle): Perception regarding the network building character of the MAPs (own survey) (legend left) Figure 27 (right): Perception regarding trust building within the MAPs (own survey) (legend left). 5.4 Constructive interaction, coordination, direction (requires relevance) In I-CISK, constructive interaction was on average considered excellent or very good (Fig. 28). Only in LL Lesotho, it was mentioned that working towards a joined CS goal, meeting all LL needs, was not fully achieved – the iterative nature can be a challenge and requires efficiency. Moreover, in LL Lesotho, both a pre-defined feasibility study and a delineated Early Action protocol existed, serving as basis for the I-CISK CS. This situation, while ensuring relevancy, might have prevented newer MAP members (such as the Lesotho Metereological Service (LMS) offices) from feeling fully involved, and might have resulted in a challenge of keeping them engaged. Figure 28: Perception of LL leads regarding the experiences and atmosphere during constructive interactions (own survey) 0% 20% 40% 60% 80% 100% Open info Shared info No. of LLs Info about I-CISK poor moderate excellent 0% 20% 40% 60% 80% 100% built new networks strenghten existing networks No. of LLs Network building 0% 20% 40% 60% 80% 100% attention to full comprehension trust built No. of LLs Trust building 0% 20% 40% 60% 80% 100% Goals clearly communicated Working towards joined goal meeting your needs Working towards joined goal meeting MAP needs No. of LLs Constructive interaction excellent moderate poor
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 29 5.5 Flexibility (requires openness and adaptability) In I-CISK, all LLs succeeded in performing a co-creation process that was open for new perspectives and seemed to have found a moderate balance between time and flexibility (Fig. 29). It was reported that the applied iterative approach allowed the co-creation framework to evolve in response to new insights, ensuring that the strategies remained relevant as conditions changed. However, LLs Hungary and Lesotho experienced a poor responsiveness to feedback, which, compounded with reported lower capacity (Ch. 5.1), lead to lower reported accountability and less constructive interactions (Ch. 5.4). Figure 29: Perception regarding trust flexibility within the MAPs (own survey) 5.6 Legitimacy (requires fairness) Given the reasonable diversity of the MAPs (except in LL Lesotho), a fair representation of end-users throughout the whole I-CISK project was achieved in most LLs; only LL Lesotho reported that fairness was poorly achieved (Fig. 30). Still, all LLs could co-create an end product that was widely deemed relevant by MAP members, as reported by the LL leads. A key lesson for this is the importance of understanding institutional mandates and responsibilities. Identifying roles and responsibilities allows for more targeted engagement, ensuring that specific actors receive the information and support they need. Figure 30: Perception regarding the legitimacy of the co-creation process within the MAPs of the LLs (own survey). 0% 20% 40% 60% 80% 100% Responsive to feedback Balance time with flexibility Open for new actors Open for new perspectives No. of LLs Flexibility excellent moderate poor 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% MAP diversity Representation of endusers in MAP Fairness of representation CS considered relevant No. of LLs Legitimacy excellent moderate poor
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 30 6 Evaluation following the product evaluation criteria In this section, we report the results of the evaluation of the prototype CS themselves. The primary goal of the I-CISK co-creation process was to develop viable, accessible, clear, credible, useful, and usable prototype CS in each of the LLs (see Ch.3.1). As such, we evaluate whether co-creation can indeed contribute to the next generation CS, overcoming the information and domain challenges described in Chapter 2.1. The evaluation suggest that the product goals were largely met in most LLs, with high scores on the indicators applicability (usefulness), affordability for areas and sectors (accessibility), and understandability (comprehensiveness) (Fig. 31). However, the integration of LK and local data (LK in Fig. 31) on climate change and variability (credibility), a key objective of the I-CISK project, was only moderately successful, and identified as an area for improvement in some LLs. Successful integration of LK was achieved in the LL Los Pedroches, where historical data inclusion enabled users to contextualise events based on their own experiences, and in LL Crete where user-defined thresholds and user-suggested data were integrated (Egan et al., 2025). The use of decision timelines was reported to be a successful tool to unpack and understand how local decisions are made and which LK, and local wisdom and experience is used in the triangulation of information supporting CCA and DRR decision (van den Homberg et al., 2024). In LL Rijnland, while the co-delivered prototype CS received positive feedback, some scepticism among MAP members persisted, particularly regarding credibility. Credibility was also flagged as a challenge in LL Crete. Figure 31: Overall performance of the co-delivered prototype climate services in the Living Labs (own survey) Note: based on the evaluative qualitative scores on the indicators for the product evaluation criteria. The red box refers to credibility, the green box to usability, the blue box to usefulness, the purple box to accessibility, the brown box to comprehensiveness, and the orange box to viability. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% LK on climate change, variability included LK on climate risk managementincluded translated into actionable information timeliness uncertainties communicatated applicability for sector and region match with decision-making processes tailored to adaptation decisions accuracy reliability scalability affordability visualisation communcation alligns with preferences understandability comprehensiveness translation to end users economic efficiency Institutional efficiency Evaluation of CS product performance excellent moderate poor
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 31 7 Overall suggestions for improving the co-creation framework Based on lessons learned throughout the I-CISK project, several improvements to the prototype co-creation framework emerged to enhance the co-creation process in all phases. These are summarised here and implemented in the final co-creation framework in Chapter 8. Phase 0: Include resource inventory and monitoring criteria To strengthen co-creation processes, LL leads recommended to include a comprehensive stakeholder inventory. This exercise can map the roles of end-users, producers, and intermediaries in adaptation decisionmaking, as well as their capacities, interests, and other resources. This can support more strategic engagement and ensure that all relevant perspectives are considered. Furthermore, the framework could strengthen its focus on equality strategies by integrating targeted metrics to address gender gaps. While changing inequalities in stakeholders’ institutions is largely beyond the reach of single co-creation projects, targeted approaches can still promote more inclusive participation. Phases A & B: Merge co-exploration and co-design and make this merge step follow a highly iterative process To better reflect the dynamic nature of co-creation, LL Leads recommended to combine phases A and B into a single, iterative phase that supports continuous learning and adaptation. This joint phase should, among other things, clarify the current state of adaptation behaviour and define desired shifts in adaptive practices. A structured timeline for decision-making proved to be very useful for identifying key adaptation decisions, formulating desired changes, and exploring future possibilities. It can also foster more reciprocity in this phase, for example, by providing examples that challenge habitual thinking, while still allowing end-users to identify and articulate their own needs. The evaluation also suggested that expectations should be managed from the outset, e.g., regarding achievable lead times, which is crucial for maintaining trust and momentum. To prevent stakeholder fatigue, activities should be streamlined to ensure that participation remains meaningful, manageable, and sustainable throughout the iterative process. Furthermore, integrating value chain thinking through the development of a draft business case should be added, to help anchor the process in practical and economic reality. The phase should conclude with the joint development of an impact trajectory, such as a Theory of Change (ToC) (Castellana et al., 2025), to create a shared vision and logic for change. Phase E: Integrate evaluation across all co-creation phases To ensure robust monitoring, evaluation, and learning (MEL) throughout the co-creation process, the reflexive FGD at the end of the project showed that it is essential to embed MEL into the workflow by designating explicit moments, such as the end of each phase, to ensure this step is not overlooked. It was recommended to tailor the process MEL criteria for collaboration early in the process, ideally in Phase 0, to the specific objectives of each phase; for example: during the initial co-exploration phase, inclusivity may be more important than active engagement. The criteria should be adopted by the MAP to enable systematic process MEL, ensuring the quality of stakeholder engagement throughout the project. The end-of-project FGD also underscored the importance of developing product evaluation criteria in Phase A to effectively track progress; these can be expressed as Key Performance Indicators (KPIs). Furthermore, it was suggested to incorporate reflection on behavioural changes in CS use throughout the project to ensure the creation of CS that are behaviourally informed as well as to track impact. It is expected that new evaluation approaches will strengthen systemic MEL, fostering continuous improvement and deeper understanding of the effectiveness of co-creation efforts.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 32 Phase F: Add capacity building training and support The main suggestion was to strengthen the focus on training and capacity building to empower stakeholders and increase long-term impact. Cross-cutting Recommendations To increase the effectiveness of co-creation, it was deemed important to move beyond one-way participation and promote multidirectional interaction between project partners and MAP members to fosters richer, twoway exchanges of knowledge and perspectives. This approach supports deeper mutual understanding and more meaningful collaboration. Furthermore, LL leads recommended that the design of the co-creation process should remain iterative and flexible, allowing methods and content to evolve in response to stakeholders' evolving insights. This flexibility, however, must be balanced with a clear path toward a final, collaborative decision to prevent fatigue and ensure closure. Lastly, the final co-creation framework should consider the varied starting levels, requiring additional data collection in some instances or skipping a phase in others.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 33 8 Final I-CISK co-creation framework 8.1 Co-creation framework overview The I-CISK co-creation approach (Fig. 32) combines principles for successful CS with principles for successful co-creation and supports the development of an effective and sustainable end-user centred CS that responds to the knowledge needs of decision-makers. It builds on the prototype (Wens et al., 2022), the evaluation (this deliverable) and other project experiences (all summarized in deliverables, accessible at i-cisk.eu/resources). Figure 32: Piloted co-creation framework for user-centred climate services, based on I-CISK experiences The tested-and-improved framework to co-create CS with stakeholders in focussed LLs consists of one initiation phase (0), four main phases (A-D), and a constant iterative evaluation phase (Fig. 32). Application of the identified phases of the framework, leads to: recognising end-user’s needs, perceptions, knowledge, capacity, and adaptive behaviours, as well as the CCA and DRR strategies they wish to implement removing the barriers of end-users to use climate information effectively to instigate behavioural changes advancing the data science and technology base necessary for the development and implantation of tailored CS that respond to the needs, perceptions, knowledge and capacities of end-users, in collaboration with end-users and other stakeholders jointly fostering a sustainable CS market sector and capacities of CS developers, service providers and purveyors, both in the public and private sector, through tools, platforms and mobile applications that best meet end-user’s needs initiating the co-evolution of knowledge and CS within the LLs enabling end-users to develop more effective CC adaptation strategies based on improved CC knowledge provided by the CS.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 34 Figure 33: Phases in the piloted (tested, evaluated, updated) I-CISK co-creation framework Figure 33 shows how going through each of the co-creation phases, builds the co-evolution of CS and knowledge, especially if each phase is structurally co-evaluated with attention to both the CS and the cocreation process. While Figure 33 shows a rather linear process, one has to imagine it very much iterative (like Fig. 32), and depending on the context, phase 0 and A can be skipped if a LL and MAP already exists, and if a clear demand and feasibility study has already been executed, respectively. The following sections elaborate per co-creation phase which enablers and success factors can be used, and which possible barriers will need to be overcome, to produce such CS. Each step is complemented by an overview of methods and tools that can be applied. For each step, there are: Objectives/scope: these relate to the project objectives Core activities: actions can need to be taken (tailored to the needs of the LL) and tools to execute them Practical considerations: tips to execute the step successfully Relevant process evaluation criteria: critical points to evaluate during each phase. It is important to emphasize the flexibility (rather than prescriptive nature) of these guidelines: the process can be modified, responding to needs or changing circumstances. We recommend searching for an optimal balance between standardization of and customization of the co-creation process (EC 2020 climate resilient Europe), and the roadmap in phase 0 is a first point to discuss tailoring the co-creation process with all stakeholders.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 41 grounded in local needs and creativity. While sharing examples can help clarify, inspire, a good balance is needed to avoid steering the LL too much towards an example CS. Relevant process evaluation criteria Inclusiveness and legitimacy: The representation of stakeholders throughout the phase is fair, and the process provides a safe space for multiple stakeholders to share expertise and challenge each other's views and contributions. Trust: The activities in this phase are performed recognising participants’ divergent values and beliefs, unbiased in their conduct, and fair in the treatment of opposing views and interests. Constructive interaction: the impact pathway will ensure that collaboration is solutions-focussed and decision-driven, objective and outcome led, with clearly identified roles and responsibilities. It needs to be clear that the process works towards a jointly agreed upon product which has value-added for all involved, meets user needs and produces information of relevance for decisions. Flexibility: The process is agile and responsive to feedback of stakeholders.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 42 8.4 Phase B - Co-develop climate product, bringing together local and global climate data and knowledge After the context-specific climate challenges have been identified and building on the strategies to be supported, this step aims to develop a trusted climate product tailored to end users’ needs regarding decision-making. It consists of global to local climate and weather data, projections and predictions, which are combined with (sectoral) climate knowledge and experiences from conventional scientific sources as well as from local partners into understandable climate metrics. Phase objectives and scope In this co-development phase, the CS needs are translated into an interpreted climate product (if so desired, with relevant scenarios, time scales, triggers, etc). It is a process in which providers and end-users work together (often with the help of intermediaries) to combine different knowledges, skills, and practices and to create new, relevant information that meets the needs of end-users and that addresses a shared concern. Collective brainstorming results in a multitude of climatic parameters, thresholds, and climate knowledge, metrics inclusive of local experiences, data, and knowledge and is integrated into a climate product so that the co-created CS is trusted and addresses users’ needs. This step thus involves the integration of the diverse stakeholder knowledges and experiences, and matching local observations (e.g. through citizen science) with scientific climate data (modelled, projected, e.g., from Copernicus, S2S Prediction project, EMODnet, GEO, ESA Actions), with the aim to enhance the accuracy and acceptance of the CS. Including LK and data may require using qualitative data when quantitative data are unavailable. Core activities of this phase • In-depth (structured and semi-structured) interviews and FGD are useful to jointly explore diverse climate knowledges and experiences. It can also be useful to organise participatory dialogues and quizzes to discover how different stakeholders perceive and experience the current climate system and its threats. This will create an overview of LK, local expertise and citizen science (including data) to build a shared understanding of climate change-related risks and responses. • Collaboratively collect and synthesise climate data. Interpret and combine local, sectoral and local knowledge, and project partners' expertise (conventional science) to jointly analyse climate data from observations and model outputs, both shortand long-term. The regional, modelled, and local datasets can then be adapted to the spatial and temporal scales required by the end-users; data interoperability is organised in the back-end to serve end-user needs in the front-end. Ensure that the climate information and models used are not only scientifically robust but also accessible to stakeholders, considering licensing, language, and technical barriers. • Workshop to co-transform climate data to information, tailoring it to end-user needs. This exercise will result in a set of metrics (including indicators, indices, thresholds, and timing) relevant for CCA and DRR. This includes integrating data across sub seasonal, seasonal, decadal, and long-term climate change timescales, and ensuring that the format and resolution are appropriate for the decision-making context. It is also useful to co-explore accuracy and uncertainty of climate information. • Co-evaluation of the jointly developed climate metrics. Dialogues can be used to verify whether these climate metrics reflect local experiences, such as thresholds and practices, whether the climate information derived from the metrics is understandable, whether different data sources are adequately integrated and differences are clear, and whether this information can adequately support the targeted CCA and DRR policies. • Structured reflection on acceptable levels of performance (uncertainty and skill). This helps define thresholds for usability and informs how CS should communicate confidence levels and what constraints may
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 43 affect implementation. CS performance can be communicated in the form of Relative Operating Characteristics (ROC) curves, heatmaps of the skill metric at different forecast lead times, bar charts summarising the number of hits, misses and false alarms, and plots produced for specific past events. It might be a technical discussion that is only for providers and intermediaries, when end users prefer not using probabilistic information, rather thresholds or ‘worst case-best case’ predictions or projections. In that case, the CS interface should accommodate both. • An elaborate overview of additional participatory methods to support this phase can be found in Van den Homberg et al. (2023). Practical considerations Integrate local and scientific knowledge through mutual jargon understanding, common theoretical understanding, using different models and by joining the diverse interests in the end product. Remember to manage expectations regarding the possibilities of a climate product with respect to its sustainability (economic and institutional viability, options to keep updating it). When selecting datasets for the climate product, consider long-term accessibility and affordability to avoid lock-ins caused by discontinued or prohibitively expensive data sources. Provide enough space in data collection involving LK to participants to share their lived experiences, and, as researchers leading the data collection process, be aware of own biases and respect the knowledge of local communities. Take time (and if necessary, organise capacity building sessions) to clarify concepts of accuracy and uncertainty before evaluating them. For instance, hit rate ant threat score are easier to understand than Mean Absolute Error. Use the possibility to link the climate product and information to existing CCA and DRR protocols. Test user-defined thresholds for action with hindcasting of past events and communicate their consistency. Use a broad framing of LK, including a range of different knowledges derived either through traditional or cultural norms, personal observations, lived or occupational experiences (van den Homberg et al., 2024). Relevant process evaluation criteria Accountability: MAP members feel accountable for the product they co-developed. Inclusivity: collaboration happens in an accessible way and through equitable negotiation and compromises. The distribution of power and influence is equitable between different stakeholders during this phase. Trust, transparency: Information is open and shared (when desirable, e.g. sensitive information or IP-protected data are excluded from this sharing), and decisions are clearly explained to all CS stakeholders involved. Constructive interaction: reflexive discussions regarding the ToC happen targeted and iteratively. Engagement therein should be relevant for all stakeholders: it should be tailored to the context, recognise interests of participants; addressing their changing needs and expectations and ensuring value added for all involved. Flexibility: there is openness to adapt the product and add new perspectives to the MAP and co-creation process, as well as responsiveness to feedback of all stakeholders. Legitimacy: the process of combining local and conventional science and knowledge is seen as legitimate by the end-users.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 44 8.5 Phase C - Co-design climate service information system distributing the jointly produced climate product Once the CS solution has been agreed upon, the climate product developed in Phase B can be combined into summarized, interpreted climate information to add value (a derived synthesis of shortand long-term observations and model outputs on past and future climate and its impact on natural and human systems). The output of this step is a prototype medium to communicate this comprehensive, actionable climate information (the CSIS) to end-users. Phase objectives and scope In this co-design phase, the climate product is visualized in such a way that its information can be easily interpreted and that it efficiently supports decision making. Through a participatory interdisciplinary design approach involving the end-users, a representation of this product and information (interfaces, dashboards etc.) that corresponds with the needs is constructed. This requires the use of existing or complementary, new platforms and tools for the transformation, visualization and communication of the climate information. Providers, intermediaries and end-users work together to design a CS based on a shared understanding of complexity of decision-making, and of individual and institutional capacities. Behavioural factors, drivers, and barriers that influence the uptake and effective use of climate information are assessed and considered so that the CS can be tailored to the local context (sector, region) to optimise it reflectivity. Core activities of this phase • Rapid prototyping dialogues to collaboratively design CS interfaces (including visualisation such as mapbased visualisations, time series graphs and boxplot ranges, icons and infographics, traffic-light alert systems, and summary dashboards), building on previously identified user needs regarding content, timing, access, and communication formats, ensuring the prototypes align with the decision-making context. • Sprint workshops to collaboratively review (define, test and refine) mock-ups of the CS and gather targeted feedback on structure, accessibility, inclusivity, and usability, which feeds into the development of a functional prototype. • Bilateral meetings, FGD and interviews to gather specific information and feedback, especially where user needs are diverse, and to assess behavioural factors, barriers and drivers for uptake. • Climate risk narratives to explore how users interpret uncertainty and probabilistic information. Use visuals, narratives, and scenario-based advice to test the interpretability and relevance of scientific information. These activities support the translation of complex data into accessible formats tailored to user preferences. • Serious games (supplemented with semi-structured interviews) to test how CS use influences adaptation behaviour (Rastogi et al., 2025). This fosters mutual understanding of roles, responsibilities, and perspectives in climate-related decision-making. It also helps identify drivers for adoption and understand feedback loops between adaptation options and decision-making. Practical considerations • Facilitate a longer testing period of mock-ups, including training on their use, for end-users to interact with the CS prototype. This allows trust to be built, feedback to be gathered, and expectations to be adjusted. This strengthens the iterative nature of the co-design process • A LL with many end-users of the CS creates a challenge in targeting the CS visualisation and tailoring the medium. Set clear boundaries and provide a set of options (not fully open possibilities) to let the process go smoothly
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 45 • Provision of definitions e.g. of skill and uncertainty would be beneficial to ensure common understanding when evaluating content and design • Reliability and uncertainty of CS projections are common barriers to adoption. It is crucial to discuss what can be expected from CS information and how to use uncertain data in decision-making • Visual tools also serve an educational role. Through serious games, end-users can be supported in understanding probabilistic forecasts and projections • Long-term sustainability of CS depends on both local engagement and the availability and accessibility of climate information and models and needs considering in this phase. Relevant process evaluation criteria Accountability: All MAP members feel accountable for the CS they co-designed. Inclusivity: The distribution of power and influence between different stakeholders is equitable during this phase. Legitimacy: There is a fair representation of all relevant stakeholder voices in the co-design process. Trust, transparency: information is open and shared (when desirable, e.g. sensitive information or IP-protected data are excluded from this sharing), and decisions are clearly explained to parties involved. Constructive interaction: Reflexive discussions regarding CS development are targeted and iterative. Engagement in this phase should be relevant for all stakeholders: it should be tailored to the context, recognise interests of participants; addressing their changing needs and expectations and ensuring value added for all involved. Flexibility: There is openness to testing and experimentation to better understand what is needed for creating climate information and a climate product tailored to end user needs. There is openness to adapt the product and add new perspectives to the MAP and co-creation process. There is responsiveness to feedback of all stakeholders. Both top-down and bottom-up approaches should be applied for this.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 46 8.6 Phase D - Co-deliver and co-exploit the co-created pre-operational climate service In this phase, the co-created CS is co-exploited and co-disseminated. The participatory implementation of strategies for the appropriate use of the CS includes the creation and update of an exploitation plan, a (train-the-trainer) training, and agreements regarding the procedure of dissemination and sustainability. Phase objectives and scope In the co-delivery phase, a coordinated delivery of data, info, expertise is done. It is crucial to incorporate CS providers as active members of the co-creation process, especially in this phase, as this improves the uptake of CS. Mandates and platform hosts are defined, and business plans are collaboratively developed, ensuring future commercial exploitation. This phase also contains capacity building and policy outreach to strengthen the dissemination of the CS. Different channels relevant to the target audiences are to be used. Audience groups include the CS community, end-users of CS beyond the stakeholders in the MAP, scientific and academic research and education communities, policy makers, public and civil society organisations, and media. Core activities of this phase • Co-evaluation workshops (including scenario-based evaluations) to conduct impact and sustainability assessments. Scenario-based evaluations refer to specific test-cases such as evaluating recent events and the CS information that was given, decision made based on the. The product criteria established in Phase A (e.g. KPIs) can be used to guide the analysis. This reflective practice also supports learning and capacity building regarding CS use, and increases the effectiveness of the CS. • Iterative (e.g., biannual or seasonal) monitoring questionnaires for producers and users to monitor how testbed products (after prototype implementation) are integrated into operational procedures and cultures. These continuous feedback loops provide insight into usability, relevance, and adaptation, and document how services evolve based on user input. It also ensures that the CS remains responsive to changing user needs and scientific developments. • Business model development (building on the discussions in phase A) to assess economic feasibility and support commercial exploitation. These activities will explore financing options, integration into government policies, and potential for replication in similar regions and sectors. • Operational action plans to support the long-term integration of CS into institutional workflows. These will also include formalizations of producer-user relationships to ensure ongoing collaboration. • Training, train-the-trainer activities, and other structured capacity building, guidance, communications, dissemination and exploitation activities embedded in stakeholder-led events. These activities, if involving broad networks and with tailored language and formats, contribute to capabilities development of users and providers of CS and will ensure that CS reach relevant target groups and empower these target groups to effectively use/maintain the CS. Practical considerations Embedding CS in existing institutions ensures that they are used in practice and that mechanisms exist to maintain, evaluate and update the CS, as necessary Continue capacity building and support for providers and users to foster effective use and trust in CS Build on the business plan developed at the start (in Phase A) Establish strong links between CS and policy or action frameworks at local and national levels Take enough time for this phase, preferably one year (but depending on the type of CS).
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 47 Relevant process evaluation criteria Accountability: all MAP members feel accountable for the prototype that is co-delivered. Fairness: negotiation and compromises are made in an equitable way. Constructive interaction: reflexive discussions regarding the prototype happen targeted and iteratively. Flexibility: there is responsiveness to feedback of all stakeholders. Legitimacy: the co-delivered prototype CS is seen as legitimate by end-users.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 48 8.7 Co-evaluate climate product and co-creation process It is important to evaluate the tangible (e.g., new prototypes for CS), intangible (e.g., knowledge, intellectual property), and diversity innovations (e.g., market relevance) generated by the co-creation process, throughout the co-creation process. Ongoing feedback, learning and adjusting based on experiences creates a flexible process that is responsive to changing preferences and needs. The co-creation phases should each have an explicit moment of evaluating the process. This can be done through an iterative dialogue in the MAPs and using the pre-defined product evaluation criteria (Ch. 3.3, tailored during phase 0). Furthermore, a moment of evaluating the co-creation product, collecting data about the KPIs developed during the first phase, should be planned after each consecutive phase. For this, the process evaluation criteria (Ch. 3.4) can be used. The evaluation moments should not be too technical but accessible to all end-users. The MEL process, including how feedback will be considered in upcoming activities, should be always made clear. Evaluation can also be done through system dynamics and behavioural economics modelling to assess the long-term usability and utility of the CS. These methods will help determine how the service influences decision-making and adaptation planning over time and provide insights into how to avoid maladaptation.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 49 9 Conclusions The I-CISK project aimed to develop a framework for co-producing the next-generation human-centred CS based on LK, behavioural insights, and user needs. This report reflects on the co-creation process that took place in MAPs within seven LLs and evaluates the prototype co-creation framework developed at the start of the project, which served as a guide throughout the project. Based on this evaluation, it then presents the final I-CISK guidelines for co-creating viable, accessible, user-centred CS that distribute useful, usable climate information through a comprehensive, credible climate product. Evaluating the I-CISK co-creation process and product Overall, the co-creation process was evaluated very positively and largely deemed to be very useful. The process in the MAPs was characterised by strong active involvement, inclusivity, and constructive interaction, although not uniformly. Trust and legitimacy were obtained through transparent communication and iterative design, while flexibility allowed for adapting to new insights. These co-creation characteristics showed to be closely intertwined: weak performance on one criterium often impacted others. For example, where limited responsiveness and capacity hampered the process, less legitimacy and ownership were also observed. Throughout I-CISK, the co-creation process was widely appreciated by both MAP participants and scientific partners. It supported mutual learning and contributed to the co-evolution of knowledge and services, as evidenced by both scientific and societal actors adapting their insights and practices along the way. Although time-consuming, co-creation proved valuable in building trust, validating LK, and increasing the relevance of climate information. In several LLs, the transition from a consultative approach to a more immersive approach led to richer collaborations and more inclusive and responsive CS products. The effectiveness of the co-created pre-operational CSs in the LLs was generally rated as high. Most of the LLs reported that the services were useful, usable, accessible, and comprehensive. However, credibility, particularly the integration of LK and the communication of uncertainties, remained a challenge in some contexts. Further CS skill and usability validation exercises are recommended. Moreover, the viability of the CS beyond the project lifetime remained uncertain in some LLs due to political changes and/or late attention to business plans. Although broader system changes beyond the LLs were not yet visible, the project laid the foundations for long-term impacts: relationships with policymakers and intermediaries, as well as the integration of the preoperational CS into regional planning, point to the potential for long-term impact. Smaller but meaningful changes, such as increased awareness, improved use of seasonal data, and new connections, were reported in several LLs, suggesting that the co-creation process influenced adaptive thinking and behaviour. It is important to note that this evaluation is based on the MAP leads in a final FGD as well as through selfreporting throughout the co-creation phases. It does not capture the full array of perspectives of all MAP members or LL stakeholders or project partners and might be influenced by the positionality and character of the reporters. An evaluation in two years by all CS value chain actors involved in the I-CISK co-creation process might have strengthened the evaluation presented in this deliverable. A structured end-user evaluation would be valuable to further evaluate the co-creation process and product. Moreover, a reflexive evaluation by the CS developers, providers and purveyors would add new perspectives on the co-creation process and successes of I-CISK.
D2.8 – Co-creation framework for end-user centred co-creation of CS across Europe and beyond 50 A piloted and improved I-CISK co-creation framework Based on insights from the reflexive exercises and evaluations, Chapter 8 of this report presents the piloted and improved I-CISK co-creation framework. This framework combines principles for successful CS development with principles for effective co-creation. It aims to guide future development of sustainable, enduser-focused CS that address the knowledge needs of decision-makers, covering the ‘last mile’ through tailored and actionable information, thus supporting CCA and DRR efforts of end-users. The framework consists of one foundational phase (Phase 0), four main phases (A–D), and a continuous evaluation phase. Each phase follows a structured, yet flexible process for collaboration, knowledge integration, and product development. Clear objectives, key activities, practical considerations, and evaluation criteria are developed for all phases. Phase 0 focuses on the establishment of a LL and the formation of a diverse and inclusive MAP. This phase builds trust, identifies end-users, and develops a joint roadmap for collaboration. Phase A focuses on co-ideation. MAP members identify relevant CCA and DRR strategies, clarify information needs and preferences, and agree on goals and realistic expectations. These can be summarized in an impact pathway, in the form of a ToC or other format, to guide the further development of CS. Phase B focuses on the co-development of the climate product. Local knowledge and data, global scientific data, and user experiences are integrated into a customised climate product that meets the identified needs. Phase C focuses on the co-design of a user-centred CSIS. Through participatory design, the climate product and information are translated into usable and accessible information tailored to the LL's decision-making contexts and presented in a way and through a medium that optimally supports the end-users' adaptation behaviour. Phase D focuses on co-delivery and implementation. The CS is embedded in decision-making processes, and stakeholders are supported to use climate information effectively. Capacity building and market development are encouraged to ensure that the final CS has a sustainable life. A continuous evaluation phase supports monitoring, evaluation and learning throughout the project iteratively, guided by product and process evaluation criteria. These criteria are adopted by the MAP and adjusted to the LL context. The I-CISK project has demonstrated that co-creation is not a linear or uniform process, but a dynamic and evolving practice. By embracing complexity and fostering true collaboration, CS can be developed that are scientifically robust, socially relevant, and usable. The final framework provides a structured yet adaptable roadmap for future projects to strengthen climate resilience through user-driven, inclusive CS.