Harmonizing the Development of Local Socioeconomic Scenarios: A Participatory Downscaling Approach Applied in Four European Case Studies

: Scenario analysis is a widely employed method for addressing uncertainties when assessing the physical and socio-economic impacts of climate change. Global scenarios have been extensively used in this context. However, these scenarios are in most cases not suitable for supporting local analyses. On the other hand, locally developed scenarios may lack the global context, thus having limited comparability with or transferability to other locations. The Shared Socioeconomic Pathways (SSP), which have been primarily developed for climate impact research, provide the possibility to extend the existing global narratives and adapt them to local characteristics in order to develop locally relevant scenarios. Here, we propose a methodological framework for producing harmonized scenarios across different case studies. This framework was developed in the EVOKED project and combines elements of top-down and bottom-up approaches to develop local scenarios for four regions in northern Europe. We employ the SSP as boundary conditions and, in cooperation with stakeholders from these four regions, develop local scenarios for a range of SSP. The developed sets of scenarios are consistently informed by global developments and are therefore comparable with other downscaled scenarios developed in different regions. At the same time, they have been based on local participatory processes, thus being locally credible and relevant to the needs of stakeholders. The local scenarios constitute a climate service per se as they can raise stakeholder awareness of the processes that will drive risk, exposure, and adaptive capacity in the future and inform discussions on mitigation strategies and adaptation pathways.


Introduction
Scenario analysis is an integral tool in climate change research for exploring uncertainties in the assessment of potential future socioeconomic and physical impacts of climate change-related hazards such as heat stress, sea-level rise, or heavy precipitation [1,2].Socioeconomic scenarios, in particular, can inform mitigation and adaptation decisions as developments in socioeconomic conditions drive society's adaptive capacity [3,4] as well as potential exposure to climate change hazards [5].For developing adaptation strategies at the local scale, socioeconomic scenarios need to reflect the local characteristics of the study area [6][7][8].Local socioeconomic scenarios that are developed in collaboration with local stakeholders can, for instance, challenge perceptions and attitudes, encourage discussion about current issues, raise awareness, and serve as a tool to bridge knowledge gaps [9][10][11][12].
Previous work has proposed a range of approaches for developing locally relevant socioeconomic scenarios, ranging from top-down expert-led approaches to bottom-up stakeholder-led approaches.Top-down approaches are based on the assumption that socioeconomic developments at the local scale are embedded in the developments at regional, national, and global scales [13,14].Therefore, these approaches use existing global and/or regional socioeconomic scenarios as boundary conditions, which are downscaled based on the characteristics of the study area.These characteristics are established by integrating expert knowledge with a literature review [15][16][17][18][19][20].The advantage of top-down approaches is that the downscaled scenarios are consistent with the global scenarios used as boundary conditions, which ensures comparability of studies across different regional contexts.However, scenarios developed exclusively using a top-down approach may not be relevant for local stakeholders, which limits their usability in local decision-making [21].
Bottom-up approaches, on the other hand, do not use existing scenario frameworks to guide the scenario development process, but consult with local stakeholders from the beginning of the process, resulting in an original set of local socioeconomic scenarios [22][23][24][25][26][27].Such stakeholder-led approaches ensure that the developed scenarios are tailored to the local characteristics of the study area, thereby creating a sense of ownership among stakeholders.As a result, the scenarios are more likely to be adopted as a decision tool at the local level, for instance as a basis for the development of local adaptation strategies [9,24,27].
This study has been designed as part of the EVOKED (EVOKED: Enhancing the value of climate data-translating risk and uncertainty utilizing a living-labs approach (http://tinyurl.com/3tjwbwbz,accessed on 1 March 2024) project, which has developed climate services in close collaboration with local stakeholders, with the aim to support adaptation planning in four case-study regions across Europe (see Section 2) [28].The integration of local knowledge is one important precondition for a successful adaptation process [29].The project has employed a living lab (LL) approach, which puts the end-user (i.e., local stakeholders) at the center of the development process in order to "co-create" and co-develop new knowledge, products or services [30][31][32][33][34].The development of locally relevant socioeconomic scenarios is an integral component of such climate services as these scenarios provide the local basis for exploring different adaptation strategies.The project case studies are exposed to a variety of climate change-related hazards, which yield specific adaptation needs.Therefore, the socioeconomic scenarios need to reflect the local characteristics of each case study region.
To develop such locally relevant socioeconomic scenarios, we devise and apply a scenario development framework that combines elements of top-down and bottom-up scenario development approaches [35,36].Applying this framework, we produce local socioeconomic scenarios that integrate local stakeholder knowledge, while at the same time harmonizing the scenario development process across case studies by using the same underlying scenarios as a starting point.Harmonization of the scenario development process ensures that the developed scenarios are internally consistent and coherent with the overarching scenarios used as boundary conditions, i.e., they follow the same scenario logic and the same underlying scenario assumptions, although storylines at the local scale may differ across case studies [21].In this way, scenarios become locally relevant and can be directly employed in local-level decision-making and planning processes.At the same time, more overarching conclusions can be drawn across case studies and results can be compared to studies in other regions that have used the same scenarios as boundary conditions for the development of local scenarios [37,38].
Our scenario development framework comprises a "multi-scale co-production approach" where the developed local socioeconomic scenarios are coherent across multiple spatial scales (i.e., global-regional-local) [11,39] and are co-produced in collaboration with local stakeholders.We have designed this approach specifically for the development of local socioeconomic scenario narratives, which constitute qualitative descriptions of plausible socioeconomic developments in the form of a story [2,12].We focus on the coproduction of local scenario narratives as narratives ease the integration of local stakeholder knowledge into the scenario development process due to their typically non-scientific language [40].The local narratives describe long-term developments, until 2100 (following the timeframe of the SSP), and can constitute the basis for producing local-scale projections of key variables that need to be considered in adaptation planning, such as the spatial distribution of population and infrastructure [13].
In the next sections, we provide a brief introduction into each case study region; present our "multi-scale co-production approach" and explain how we use this approach to develop and implement the scenario framework in each case study region by tailoring it to the local stakeholder needs of the respective study area; compare the scenario outcomes across the case studies; and critically reflect upon the scenario development process, its implications for adaptation planning at local level and its potential use in climate impact research.

Case Study Regions
To address climate adaptation at local scale and develop climate services, EVOKED focuses on four local-and regional-scale case studies in Norway, Sweden, Germany and the Netherlands.The selected sites (Figure 1) are exposed to a range of natural hazards and local governments seek to initiate, or enhance the value of ongoing, adaptation planning, via climate services.
ditions for the development of local scenarios [37,38].
Our scenario development framework comprises a "multi-scale co-productio proach" where the developed local socioeconomic scenarios are coherent across mu spatial scales (i.e., global-regional-local) [11,39] and are co-produced in collabo with local stakeholders.We have designed this approach specifically for the develop of local socioeconomic scenario narratives, which constitute qualitative descriptio plausible socioeconomic developments in the form of a story [12,40].We focus on t production of local scenario narratives as narratives ease the integration of local holder knowledge into the scenario development process due to their typically nontific language [41].The local narratives describe long-term developments, until 210 lowing the timeframe of the SSP), and can constitute the basis for producing local projections of key variables that need to be considered in adaptation planning, such spatial distribution of population and infrastructure [13].
In the next sections, we provide a brief introduction into each case study region sent our "multi-scale co-production approach" and explain how we use this appro develop and implement the scenario framework in each case study region by tailor to the local stakeholder needs of the respective study area; compare the scenario outc across the case studies; and critically reflect upon the scenario development proce implications for adaptation planning at local level and its potential use in climate im research.

Case Study Regions
To address climate adaptation at local scale and develop climate services, EVO focuses on four local-and regional-scale case studies in Norway, Sweden, German the Netherlands.The selected sites (Figure 1) are exposed to a range of natural ha and local governments seek to initiate, or enhance the value of ongoing, adaptation ning, via climate services.

City of Larvik, Norway
As a coastal city located in southern Norway, Larvik is exposed to hazards and has historically experienced floods, strong winds, storms and storm surges.These events are expected to become more frequent and more intense with potentially increasing damage costs as a result of climate change [41].In particular, the expected increase in extreme precipitation could lead to increases in intensity and frequency of urban flooding, erosion, quick clay slides, rockslides, and river flooding; while increases in storm activity in Skagerrak, in combination with a rising sea level, will increase the intensity and frequency of storm surges, coastal flooding, and erosion in Larvik.
According to the Civil Protection Act, all Norwegian municipalities have the obligation to carry out comprehensive risk and vulnerability assessments (RVA) which should be considered in land-use planning and should map potential impacts of climate change [42].The municipality of Larvik aims to reduce both current and future risks due to climate change, e.g., through actions of its emergency preparedness group, or by incorporating adaptation analysis in all development projects.
As part of the last update of Larvik municipality's RVA, the potential impacts of climate change due to a range of natural hazards including intense rainfall, flooding, erosion, landslides, and storm surges were explored [43].In this context, exposure maps for Larvik were developed and used for the building development of the Martineåsen area within the municipality, also accounting for a range of suggested local adaptation and mitigation measures; while the knowledge needs and perceptions of stakeholders were explored as input to the development of future socioeconomic scenarios that were relevant not only for the local authorities, but also for the community of Larvik.

Värmland, Sweden
Värmland is an inland county located in Sweden that is exposed to flooding and other water-related hazards from Lake Vänern (the largest lake in the European Union) and river flows.The county capital of Karlstad situated on a delta where the Klarälven river meets Lake Vänern and has experienced several floods and landslides.In Värmlands county there is a need for new housing to deal with urbanization, especially in Karlstad.Smaller municipalities are losing population due to urbanization migration to Karlstad and are using the possibility of living near the water as a way to attract people.This leads to more construction occurring close to the water today than previously, even though recent flooding events have put many areas at risk.
The Värmland County Administrative Board (VCAB) coordinates climate adaptation work in the county.It supports municipalities and other regional actors through awareness raising activities and disseminating knowledge about current and future climate change and climate change impacts.Within Värmland county, VCAB suggests measures which could help increase local and regional resilience to climate change; develops the Regional Climate-Adaption Action Plans [44]; and plays an active part in complying to the European Union Flood Directive.It focuses on adapting society to a changing climate to protect the environment, people, and property in a long-term perspective.

Fluvius Region, The Netherlands
For the Netherlands, climate adaptation is often associated with flood risk (e.g., [45]).However, water scarcity, drought and heat stress are also considered as threats [46], particularly in the rural areas in the sandy east and south of the Netherlands, where drought is expected to become a problem [47,48]; while extreme rainfall and flooding are expected to occur more frequently.These hazards may affect both rural and urban areas [46].In response, the Deltaplan Spatial Adaptation (Deltaplan Ruimtelijke Adaptatie) was developed to help guide Dutch efforts [46] to such hazards.In this approach, the initial focus is to gain insights in the potential risks of climate change (stress test), followed by the initiation of dialogues between local governmental bodies and regional stakeholders (e.g., agriculture, public healthcare, drinking water, nature) (risk dialogue) leading eventually to co-developing adaptation plans (regional adaptation strategy).
The Fluvius region is located in the North-Eastern part of The Netherlands.The abundant presence of water has driven the social and economic development in this region.The regional and local governments in the region collaborate for establishing a water robust and climate proof region, with the aim to provide for a water safe area in 2100; and have launched the program "Living with Water in the IJssel-Vechtdelta" in 2014.This program aimed, among others, to improve the awareness of the regional communities regarding the effects of climate change.Within this context the region has developed a range of climate services included in Platform Water Fluvius [49] and has completed the process of consultation with regional stakeholders.This process has resulted in a regional adaptation strategy and regional implementation agenda, which acts as a strategic document for informing more localized climate adaptation plans for the Fluvius-partners [49].

City of Flensburg, Germany
The city of Flensburg is located at the German Baltic Sea coast, directly at the national border to Denmark.The city is prone to coastal flooding during storm conditions, when a high-pressure system with winds from the northeast follows a low-pressure system with strong winds from the west, as these wind conditions can potentially result in a seiche over the entire Baltic Sea basin [50].The intensity and frequency of these flooding events are expected to increase over the next century and beyond due to accelerated sea-level rise [29,[51][52][53].Currently, no assessment of vulnerability to coastal flooding exists for the city and no adaptation measures are in place apart from measures taken on an individual level [54,55].These household-level adaptation measures include sandbags and the installation of mobile barriers during flooding events [56,57].Due to these circumstances, the city of Flensburg has recently initiated the process of developing an adaptation agenda, in cooperation with the local community.Furthermore, local stakeholders do not tend to favor conventional protection strategies in the form of structural solutions, and appear more positive to alternative adaptation options for coping with future flood risk.In this context, the city can benefit from support in exploring potential adaptation options.

The Multi-Scale Co-Production Approach for Local Scenario Development
Our method is based on previous work by Rotmans et al., 2000 [12] who developed scenarios for sustainable development in Europe, using a participatory scenario development approach.They proposed five steps to be taken during each stakeholder workshop.We adapted these steps to our approach and introduced an additional one, resulting in the following six steps shown in Figure 2. Beyond the introduction of an additional step, our approach differs from [12] in its implementation.In our study, we employed an existing scenario framework, the SSP, which we downscale to each of the study regions.By defining local scenario elements for each case study, we extended the SSP for the study regions, thus developing new locally relevant scenarios, while at the same time keeping them consistent between all case studies.Our approach is not only workshop-based and requires a longer period for its implementation.This has allowed more interaction with the stakeholders, which is necessary for introducing the SSP framework; and a larger number of iterations before finalizing every step.The six steps that we undertook for the development of local scenarios are described in detail below: Step 1: Select scenarios as boundary conditions When compiling socioeconomic scenarios at a local scale, it is important to not only account for local developments and processes, but also to take into consideration that the region is embedded in developments at different spatial scales, ranging from global to national, regional, and local levels [13,14].Therefore, in order to ensure compatibility between case studies, it is important in the first step to select scenarios for all case study regions that will serve as boundary conditions for the local scenarios.This step needs to be taken jointly (i.e., with all those working on the different case studies) to ensure that the local socioeconomic scenarios are harmonized across case studies.
In EVOKED, we used the global-scale Shared Socioeconomic Pathways (SSP) as a starting point for providing the global context to the local scenario narratives as the SSP are the current state-of-the-art socioeconomic scenarios used in climate change research [58,59].The SSP describe five broad-scale pathways of plausible socioeconomic development until the end of the 21st century at a global scale, based on societal challenges for mitigation and adaptation [59,60].SSP1 "Sustainability" describes a sustainable world with low challenges for both mitigation and adaptation; SSP2 "Middle of the Road" is a scenario with moderate challenges; SSP3 "Regional Rivalry" has high challenges; SSP4 "Inequality" is characterized by high challenges for adaptation, but low challenges for mitigation; SSP5 "Fossil-fueled Development" has low challenges for adaptation and high challenges for mitigation [61].Each SSP consists of a qualitative narrative and projections of several variables that have been produced on a national level, such as population [62], urbanization [63], and GDP [64].
Sustainability 2024, 16, x FOR PEER REVIEW 6 of 19 Step 1: Determine set of scenarios as boundary conditions Step 2: Establish local scenario elements Step 3: Determine plausible future developments of each scenario element Step 4: Draft scenario narratives Step 5: Facilitate feedback and discussion with local stakeholders Step 6: Refine scenario narratives based on stakeholder discussions Step 1: Select scenarios as boundary conditions When compiling socioeconomic scenarios at a local scale, it is important to not only account for local developments and processes, but also to take into consideration that the region is embedded in developments at different spatial scales, ranging from global to national, regional, and local levels [13,14].Therefore, in order to ensure compatibility between case studies, it is important in the first step to select scenarios for all case study regions that will serve as boundary conditions for the local scenarios.This step needs to be taken jointly (i.e., with all those working on the different case studies) to ensure that the local socioeconomic scenarios are harmonized across case studies.
In EVOKED, we used the global-scale Shared Socioeconomic Pathways (SSP) as a starting point for providing the global context to the local scenario narratives as the SSP are the current state-of-the-art socioeconomic scenarios used in climate change research [58,59].The SSP describes five broad-scale pathways of plausible socioeconomic development until the end of the 21st century at a global scale, based on societal challenges for mitigation and adaptation [59,60].SSP1 "Sustainability" describes a sustainable world with low challenges for both mitigation and adaptation; SSP2 "Middle of the Road" is a scenario with moderate challenges; SSP3 "Regional Rivalry" has high challenges; SSP4 "Inequality" is characterized by high challenges for adaptation, but low challenges for mitigation; SSP5 "Fossil-fueled Development" has low challenges for adaptation and high challenges for mitigation [61].Each SSP consists of a qualitative narrative and projections of several variables that have been produced on a national level, such as population [62], urbanization [63], and GDP [64].
The SSP have been intentionally designed in a broad manner to allow for the development of so-called "extended SSP" for sectoral and/or regional to local applications [59,65].More than thirty key elements that have been identified as important drivers of socioeconomic development at the global scale are described in each global SSP narrative [61].As a result, the global narratives are generic enough to provide the overall The SSP have been intentionally designed in a broad manner to allow for the development of so-called "extended SSP" for sectoral and/or regional to local applications [59,65].More than thirty key elements that have been identified as important drivers of socioeconomic development at the global scale are described in each global SSP narrative [61].As a result, the global narratives are generic enough to provide the overall socioeconomic context in each SSP while providing elements that are relevant to sub-global analyses as well [61,66].We use these scenario elements and their characteristics in each SSP as a starting point for the development of extended SSPs for each case study region.We must note that, as the development of our local SSP extensions was completed in 2018, we did not associate the SSP with radiative forcings, as presented in Gidden et al., 2019 [67] and adopted in the IPCC AR6 [68].Nevertheless, as the scope of our study is the compilation of local narratives for socioeconomic development, relevant future forcings can subsequently be associated to each of the developed local scenarios for impact and vulnerability analysis.
Step 2: Establish local scenario elements In step two, we extended the global SSP key elements by local scenario elements that are important drivers of societal development in each case study region.To establish local elements, we reviewed the locally relevant policy documents and literature, analyzed data of the local and regional administrations and statistics offices, and/or consulted with local stakeholders of each case study region [69].In addition, several guiding questions for each case study region have been compiled in order to explore how drivers of socioeconomic development are embedded in the global to regional scale, which is an important aspect of multi-scale scenario development [10,11]: • What is the demographic structure of the population?What are recent population trends?• What are the major characteristics of political and socioeconomic importance and/or issues of concern in the case study region?

•
How are local politics embedded in and influenced by regional, national, and global politics?

•
In which ways is the local economy embedded in global markets?What are the biggest companies in the case study region?Are they national, regional, or global players?
As pointed out by Kok et al., 2007 [10] and Özkaynak and Rodríguez-Labajos 2010 [11], it is important to include local stakeholders at this stage of the process in order to confirm whether the established elements are relevant at the local scale and to establish potentially missing elements.
Once the local elements were established, we further identified the global SSP elements that are most relevant to each case study region and extended them based on the respective local SSP elements.We focused only on central global elements in order to keep the total number of scenario elements manageable, thereby ensuring that the draft narrative is as short as possible to facilitate the communication with stakeholders [11].
Step 3: Determine plausible future developments of each scenario element Based on current characteristics in socioeconomic development identified through the literature review and analysis of regional socioeconomic data in step two, we proceeded to define for each local scenario the characteristics of the scenario elements, including the newly established local elements as well as the selected global elements.As it is important to concentrate on a small number of scenarios in order to enhance communication and to facilitate discussion with local stakeholders, we propose using a set of three to four scenarios [18, 70,71].This number is sufficient to account for the full range of uncertainty regarding the challenges for mitigation and adaptation, as defined in the global SSP [72].
To maintain consistency with the global SSP, we used the developments of the global SSP as the basis for each respective local scenario; adapted them where necessary to reflect the developments at local scale; and extended them with further socioeconomic context using the information for the local elements.Here, it is particularly important to ensure that the assumptions of the future developments (at local level) in each scenario are plausible, in order to increase the likelihood that local stakeholders will adopt the developed scenario narratives for local decision-making.Plausibility of scenarios is achieved if, based on their current knowledge and understanding of the world, stakeholders agree that the developments described in each scenario "could happen" [73,74].We would generally propose to assemble the results of step three in a table in order to allow for comparing and contrasting the characteristics of all scenarios in a structured and clear manner.
Step 4: Draft scenario narratives Using the table outlining the characteristics of the scenario narrative elements, full-text narratives were then drafted for each scenario, qualitatively describing the developments in the form of a story [2,12].The narratives add further context to the scenario elements and are written in non-scientific language, with the aim to facilitate stakeholders' understanding of each scenario [40].As pointed out by Kok et al., 2007 [10], local stakeholders will be less overwhelmed by global developments, particularly social and economic developments, if they are not presented as facts but rather as underlying assumptions and changes in world views.In this step, we suggest giving each scenario a new name that reflects the overall socioeconomic developments in each narrative in a concise manner, also referring to the case study context.New scenario names are important for stakeholder identification with the local scenarios.Therefore, they should be developed in collaboration with the local stakeholders, aiming to create a sense of ownership among them [75].
Step 5: Facilitate feedback and discussion with local stakeholders In step five, the first drafts of the narratives along with the scenario names for every case study were discussed with the respective stakeholders, to ensure plausibility and relevance of the local scenarios.In this step, it is essential to clearly communicate to the stakeholders that scenarios are explorations of plausible futures rather than predictions of what will happen [70,74,76].Therefore, it is important to note that the developed narratives can also reflect developments that are considered as less likely by the stakeholders [10].Depending on stakeholders' background knowledge and prior experience with scenarios, sufficient time should be set aside for explaining these basic notions [9,39].
Stakeholder interactions can take place in different communication formats, such as interviews, expert workshops, focus group discussions or stakeholder workshops [10].Besides the narrative text, other visualization tools can be employed in order to facilitate understanding of each scenario; such tools include, for instance, pictures, collages, comics, graphs, explanatory videos, or a theatrical play [10,77,78].It is important that the communication formats and visualization tools are selected based on the amount of time available and the previous experience of the stakeholders.Therefore, the project partners select those formats and tools most suitable for their case study region [12,78].
In EVOKED, we aimed at including a broad group of stakeholders in the process as one important component of the living lab approach.An initial list of stakeholders was compiled for each case study, based on web-based search and previous co-operation.Based on suggestions from our key stakeholders, who were already partners in EVOKED, we extended this list using a snowballing method.Each project partner conducted a stakeholder analysis to establish the relevant stakeholders in each case study region and understand their interest and influence in the region, based on Reed et al., 2009 [79].We aimed to include representatives from business, governments, NGOs, science/experts, and citizens of the case study region, following the recommendations of previous work [10,75,80].More detailed information on the selection of stakeholders can be found in Reimann et al., 2021 [73].
Step 6: Refine scenario narratives based on stakeholder discussions Based on the stakeholder feedback in step five, the narrative drafts were revised to include the ideas and local knowledge of the stakeholders.Here, the scenario experts weigh which of the points raised during the discussion to include in the scenario, thereby ensuring coherence across scales (i.e., with the global SSPs) [21] as well as consistency within each scenario narrative [10,12].Subsequently, the revised narrative drafts were discussed with the local stakeholders again to ensure that the revised scenarios were plausible and relevant at the local level.If necessary, this process should be repeated in several iterations until the scenarios are fully approved by the stakeholders [26,80].Depending on the time and resources available, this iterative process can take place in workshops or focus group discussions, but also remotely via email or during conference calls.

Implementation of the Approach in Each Case Study Region
In this section, we describe how we applied the six-step multi-scale co-production approach in each case study region and present the outcomes of this process.Table 1 provides an overview of how each step of the approach was tailored to the local particularities of each case study.
Table 1.Overview of the implementation of the six-step multi-scale co-production approach in each case study region.
Step # Norway Sweden Netherlands Germany

Implementation of the Approach in Each Case Study Region
In this section, we describe how we applied the six-step multi-scale co-production approach in each case study region and present the outcomes of this process.Table 1 provides an overview of how each step of the approach was tailored to the local particularities of each case study.

Implementation of the Approach in Each Case Study Region
In this section, we describe how we applied the six-step multi-scale co-production approach in each case study region and present the outcomes of this process.Table 1 provides an overview of how each step of the approach was tailored to the local particularities of each case study.focus group discussions, but also remotely via email or during conference calls.

Implementation of the Approach in Each Case Study Region
In this section, we describe how we applied the six-step multi-scale co-production approach in each case study region and present the outcomes of this process.Table 1 provides an overview of how each step of the approach was tailored to the local particularities of each case study.Table 1.Overview of the implementation of the six-step multi-scale co-production approach in each case study region. Step

Implementation of the Approach in Each Case Study Region
In this section, we describe how we applied the six-step multi-scale co-production approach in each case study region and present the outcomes of this process.Table 1 provides an overview of how each step of the approach was tailored to the local particularities of each case study.

Implementation of the Approach in Each Case Study Region
In this section, we describe how we applied the six-step multi-scale co-production approach in each case study region and present the outcomes of this process.Table 1 provides an overview of how each step of the approach was tailored to the local particularities of each case study.

City of Larvik, Norway
The local scenarios for Larvik were based on the five global-scale SSP.First (steps one and two), policy documents, the literature, and data specific to the region of Larvik, were analyzed in order to explore challenges and potential trends, already observed in Larvik today.Three of the global SSP were selected following discussions with the municipality, to ensure that the selected SSP are considered most relevant for Larvik.The selected global scenarios included SSP1, SSP2 and SSP3, and plausible future developments were determined for each scenario element in the third step.SSP1, "Sustainable Larvik", was primarily based on the municipal plans, such as the land-use plan and energy-and climate plan, among others.This scenario reflects the way that Larvik is currently planning for the future.SSP2, "Business as usual", is based on relevant policy and strategy documents and knowledge bases, depicting the current state of Larvik.SSP3, "Regional rivalry", is based on similar sources, but depicts a scenario that could take place if the existing strategies "failed".Regional cooperation is an important factor for the future of Larvik as all neighboring municipalities in the region are preparing for an anticipated centralization process towards the larger cities, especially in the Oslo

City of Larvik, Norway
The local scenarios for Larvik were based on the five global-scale SSP.First (steps one and two), policy documents, the literature, and data specific to the region of Larvik, were analyzed in order to explore challenges and potential trends, already observed in Larvik today.Three of the global SSP were selected following discussions with the municipality, to ensure that the selected SSP are considered most relevant for Larvik.The selected global scenarios included SSP1, SSP2 and SSP3, and plausible future developments were determined for each scenario element in the third step.SSP1, "Sustainable Larvik", was primarily based on the municipal plans, such as the land-use plan and energy-and climate plan, among others.This scenario reflects the way that Larvik is currently planning for the future.SSP2, "Business as usual", is based on relevant policy and strategy documents and knowledge bases, depicting the current state of Larvik.SSP3, "Regional rivalry", is based on similar sources, but depicts a scenario that could take place if the existing strategies "failed".Regional cooperation is an important factor for the future of Larvik as all neighboring municipalities in the region are preparing for an anticipated centralization process towards the larger cities, especially in the Oslo region.In a situation where regional municipalities were not able to compete at the same level as the nearest urban metropolitan areas, regional rivalry could erupt, with regional municipalities competing amongst them.
Following a dialogue with the municipality, we decided not to develop dystopic scenarios, since Norway is a welfare state with a fairly robust safety net system for its citizens.The municipality believed that dystopic scenarios would not be considered as plausible by the stakeholders involved in the ongoing urban development plans (followed also by the EVOKED project).Here, legitimacy was an important concern for the municipality and developing scenarios that people could relate to was deemed to be the way forward.
In the fourth step, the scenario table was compiled with partners from the municipality of Larvik; in this process missing information or information that was not relevant was identified.The draft table was then revised and extended into narratives.In the fifth step, the narratives were transformed into illustrations for children and the youth (Figure 3).These illustrations were used during a workshop held at a primary school in Larvik.Feedback from the children and youth participants on these illustrations was then used in step six, for further refining the illustrations.The final illustrations were presented to local politicians as part of a larger discussion on the development of the Martineåsen project.They formed the basis for a further dialog on how to use a blue-green factor tool to ensure a high quality of the blue-green infrastructure and how building developers incorporate stormwater runoff and infiltration in their projects to reduce the impact of urban flooding.

Värmland, Sweden
In the first step, we compiled a first list of local and regional scenario elements.SSP2 was based on the official statistical data for the years of 2017-2018 of Arvika municipality and Värmland county.SSP1 was based on policy documents for sustainable development: Arvika's Environmental strategy [81], Strategic plan for Arvika 2019-2020 [82], and Värmlandsstrategin 2014-2020 [83].SSP 3, 4 and 5 were based on assumptions regarding potential developments in case the strategies for sustainable development were only partially (SSP3 and 4) or not at all realized.In step two, the various scenario elements were discussed during an online meeting with representatives from VCAB and Arvika Teknik and Arvika municipality in September 2018 to determine the plausibility of the elements in each scenario and to select the scenarios that were most relevant.
In the third step VCAB and Arvika representatives discussed the possibility of combining SSP3 (regional rivalry) and SPP4 (inequality: A more insular Arvika) into a single (c)

Värmland, Sweden
In the first step, we compiled a first list of local and regional scenario elements.SSP2 was based on the official statistical data for the years of 2017-2018 of Arvika municipality and Värmland county.SSP1 was based on policy documents for sustainable development: Arvika's Environmental strategy [81], Strategic plan for Arvika 2019-2020 [82], and Värmlandsstrategin 2014-2020 [83].SSP 3, 4 and 5 were based on assumptions regarding potential developments in case the strategies for sustainable development were only partially (SSP3 and 4) or not at all realized.In step two, the various scenario elements were discussed during an online meeting with representatives from VCAB and Arvika Teknik and Arvika municipality in September 2018 to determine the plausibility of the elements in each scenario and to select the scenarios that were most relevant.
In the third step VCAB and Arvika representatives discussed the possibility of combining SSP3 (regional rivalry) and SPP4 (inequality: A more insular Arvika) into a single scenario and whether two separate scenarios, one for Arvika and one for the entire region (Värmland), were required.SSP3 was relevant in the debate about larger regions in Sweden, but SSP4 was more relevant for the local conditions in Arvika.The decision was taken to develop two separate scenarios, one for Arvika using SSP4 and one for VCAB using SSP3.The other SSP-based scenarios would remain the same, as well as the global and local elements, but the weighting and description of the elements were different.A first draft of the scenario narratives was compiled in step four; this was subsequently discussed and revised by both VCAB and Arvika, who proposed changes and additions to the narratives to enhance their plausibility and relevance.Eventually, the local and county scenarios were merged into one narrative for the entire county of Värmland and use the Arvika scenarios as local examples.In steps five and six additional stakeholders in VCAB and Arvika were brought in during separate meetings to discuss the potential usefulness of the scenarios.For VCAB the scenarios could be used as input in the municipality's work with a regional climate adaptation strategy.For Arvika they could be used in comprehensive plans or risk and vulnerability assessments, particularly for the building and planning section (Samhällsplanering) and less so for the technical sections (Teknik i väst).

Fluvius-Region, The Netherlands
For the Netherlands, national scenarios with detailed narratives (the "Delta Scenarios") that had been developed in consultation with stakeholders already existed [84].These scenarios describe potential development paths for the Netherlands along two axes: with regard to climate change ("slow and little amount of climate impact vs. fast and large amount of climate impact") and socio-economic trends ("socio-economic shrinkage vs. socio-economic growth").As these scenarios match some of the SSP scenarios, and are at the same time relevant for the local studies in the Netherlands, they were also employed for the local scenario development by adjusting the six-step approach presented in Section 3.
In the first step, the five SSP scenarios were translated into potential developmental storylines for the case study areas.Then, we established local scenario elements based on the literature specific to the case study areas, on the "Delta Scenarios" [84] and on (local) statistical data and trends that are available for both regions [85].We employed the "Delta Scenarios" [84] as a basis for extending the plausible future development of each scenario element and focused on the scenarios "Rest" and "Steam".These correspond with SSP 1 and 5, which we believe best represent the solution space for the future development of the study areas.Based on the local data and the narrative described in the Delta Scenarios, we developed narratives for both selected SSP scenarios.For step five (asking for feedback), we made use of expert judgement from colleagues working at Deltares that are involved in their daily work with scenario development in regard to climate change impacts in the Netherlands.This judgement involved the plausibility of the developments described in the first draft of the local SSP scenarios.Based on the feedback we made final adjustments to the developed scenario drafts for SSP 1 and 5.

City of Flensburg, Germany
The scenario development process in the city of Flensburg closely followed the sixstep approach (see [73]).We established local scenario elements (step two) based on locally relevant literature, such as policy documents, [86][87][88][89][90], as well as quantitative data of the regional and local administrations and statistics offices [91,92].We discussed the preliminary local scenario elements with Flensburg's city administration in a focus group of four participants.During this discussion, we presented the local elements and asked participants to identify missing elements as well as items that were not relevant for the city of Flensburg.Subsequently, we devised the plausible future developments of each scenario element for four selected SSPs, ensuring coherence of the local-scale scenarios with the global SSPs.We chose SSPs 1, 3, 4 and 5 with the aim to span the full range of uncertainty with respect to challenges for mitigation and adaptation.Step three resulted in a table outlining the future characteristics of each scenario element for each selected SSPs.Based on this table, we drafted a one-page narrative for each scenario (step four).We further devised new scenario names that reflect the local developments described in each narrative.
In step five, we discussed the local scenario narrative drafts in a stakeholder workshop, which we co-organized with Flensburg's city administration.We focused on three scenarios (SSP1, SSP3 and SSP5) due to the number of workshop participants (13 stakeholders) and time constraints (half-day workshop).Following Kok and van Vliet 2011 [78], we used a mix of visualization tools and formats of interaction to address a range of perceptions and tastes.At the beginning of the workshop (phase one), a scenario expert provided general information of the SSP framework to the stakeholders via a flip chart presentation.In order to develop a better understanding for the scenarios, the stakeholders assigned the different scenario elements and their characteristics to the respective SSP in a second phase.Similar to a jigsaw puzzle, the task was to use different pieces of graphical elements of the SSP and organize them on a poster (Figure 4).ceptions and tastes.At the beginning of the workshop (phase one), a scenario expert provided general information of the SSP framework to the stakeholders via a flip chart presentation.In order to develop a better understanding for the scenarios, the stakeholders assigned the different scenario elements and their characteristics to the respective SSP in a second phase.Similar to a jigsaw puzzle, the task was to use different pieces of graphical elements of the SSP and organize them on a poster (Figure 4).The narratives were then discussed in smaller groups, with each group focusing on one SSP.As our aim was to motivate a lively discussion of the narrative drafts, we decided upon a group size of four to five participants for each scenario.We asked the workshop participants to provide feedback on the individual elements; report on missing ones; and to provide feedback on the narratives' drafts.After the discussion in smaller groups, we reflected upon all comments with all workshop participants.During this phase, the stakeholders identified the need for a scenario reflecting current development trends in Flensburg.
Last, we integrated the stakeholder comments into each narrative.It was not possible to include all suggestions into the local scenarios as some were not compatible with the developments described in the global narrative of the respective SSP.We sent the revised narratives to the stakeholders via email for a second round of comments.We also attached a justification and explanation document in those cases where we could not include specific comments.As we received further minor comments on SSP1 "Sustainable Flensburg", we changed the narrative accordingly.We additionally replaced the SSP4-based narrative, "Flensburg's elites on the rise" (which was not discussed during the workshop), with the newly devised SSP2-based narrative, "Flensburg as we know it", in order not to exceed a total number of four scenarios.We presented this narrative to representatives of Flensburg's city administration for their comments on its plausibility before finalizing it.The narratives were then discussed in smaller groups, with each group focusing on one SSP.As our aim was to motivate a lively discussion of the narrative drafts, we decided upon a group size of four to five participants for each scenario.We asked the workshop participants to provide feedback on the individual elements; report on missing ones; and to provide feedback on the narratives' drafts.After the discussion in smaller groups, we reflected upon all comments with all workshop participants.During this phase, the stakeholders identified the need for a scenario reflecting current development trends in Flensburg.
Last, we integrated the stakeholder comments into each narrative.It was not possible to include all suggestions into the local scenarios as some were not compatible with the developments described in the global narrative of the respective SSP.We sent the revised narratives to the stakeholders via email for a second round of comments.We also attached a justification and explanation document in those cases where we could not include specific comments.As we received further minor comments on SSP1 "Sustainable Flensburg", we changed the narrative accordingly.We additionally replaced the SSP4-based narrative, "Flensburg's elites on the rise" (which was not discussed during the workshop), with the newly devised SSP2-based narrative, "Flensburg as we know it", in order not to exceed a total number of four scenarios.We presented this narrative to representatives of Flensburg's city administration for their comments on its plausibility before finalizing it.

Discussion and Conclusions
In the context of the EVOKED project, we developed a set of local scenarios, following the common for all four case studies, stepwise approach described in Section 3. By combining a top-down (steps 1-4) with a bottom-up approach (steps 5-6) we have produced groups of scenarios that are locally relevant and accepted by the potential end-users.Further, the developed scenarios are consistent with the global SSP that were selected as boundary conditions and are therefore comparable with, or complementary to, other local and regional scenarios developed in different regions or in different contexts.
Despite the distinct local context of each case study of EVOKED, the scenarios developed for the four case studies are similar in terms of the SSP that were preferred by the stakeholders of every case study.The same trend was also observed in Kok and van Vliet 2011 [78].In all case studies there was a clear preference for a local SSP1, while three of the case studies developed local extensions for SSP2, SSP3 and SSP5.In particular, the inclusion of SSP2 was, in most cases, included following the demand of the stakeholders.This is possibly due to the nature of the SSP2 which, as a middle of the road scenario, entails less extreme change and its storyline is therefore closer to what stakeholders can envisage and project.The overall choice of scenarios seems to reflect ongoing discussions in climate policy and the general political landscape, with society being torn between sustainable solutions, the use of fossil fuels, and the increasing tendency to protect regional markets.Nevertheless, the participatory approach leads to a shift from "typical" key elements (economic, demographic, technological) to socio-cultural, political and institutional elements [12].
Our approach leads to the development of sets of scenarios that are consistently informed by global developments as well as by local participatory processes, thus being locally credible and relevant to the needs of stakeholders.The methods leave ample room for interpretation in order to develop the scenarios based on the specific context of each case study [93] but are still coherent with the SSPs, thus avoiding a range of very different scenarios that do not allow for cross-case comparisons [13,38].As such, they are characterized by internal consistency and similarity despite possible differences between them [36].Such scenarios allow for adaptation options to be explored across plausible futures, trust to be built in the context of participatory processes, local values to be understood by the developers and policies to be more robust under changing climate conditions [69].The developed scenarios provide the possibility to explore a wide range of societal futures and the resulting climate-change impacts, thus providing a solid basis for informing adaptation planning in each project case study.Importantly, coherence further leads to consistency between spatial scales, which is particularly important in the context of climate change adaptation [78].
The developed narratives also provide the basis for quantifying trends of socioeconomic development and for producing local-scale projections of socioeconomic variables that can directly feed into the analysis of risk and consequently guide the adaptation process.Examples of such variables are spatial population distributions, distribution of income, age groups, and land use changes under each scenario.For example, population projections that can be employed for assessing future exposure and vulnerability of population to potential impacts of climate change were developed for the selected SSP for the case-study of Flensburg, using the respective SSP national population growth projection rates and adjusting them depending on the scenario narrative.Using the same model for all case studies allows for using projection input data from the same baseline scenarios (SSP) [37,94], thus making the quantified projections consistent across the case studies.Further, the narratives constitute a climate service per se as they raise stakeholder awareness of the processes that will drive risk, exposure, and adaptive capacity in the future, and provide context for stakeholder discussions on mitigation and adaptation strategies and pathways [94] by comparing scenarios to determine most desirable pathway and devise policies to achieve the vision [11].As our work has primarily focused on developing narratives for local socioeconomic development and was completed before the year 2021, we have not associated the local SSP to plausible radiative forcings, as presented in IPCC AR6 [68].However, we must point out that in some of the case studies, stakeholders expressed interest in pathways that are deemed not plausible [67] according to the new framework.For example, local sustainable development, corresponding to SSP1, was perceived locally plausible in a world with high-end emissions (e.g., RCP5).Such questions, which may be relevant in a local context but seem implausible in a global context, require the use of the scenario matrix architecture [4] and cannot be fully described in the context of the newer global SSP framework, where socioeconomic development is linked to specific radiative forcings.
The process of co-developing local scenarios with stakeholders also has some drawbacks and challenges to overcome.The quality of the stakeholder feedback largely depends on the background knowledge of the stakeholders and their active participation.Regional long-term scenarios did not generally exist in most of our case studies and only a few stakeholders had previous experience in developing scenarios.In this context, exploring plausible long-term future socioeconomic developments (e.g., until 2100) can seem rather abstract to stakeholders, who require time to fully immerse in the storylines discussed.For this reason, thorough preparation of the stakeholder workshop and the use of visualization techniques and material, such as figures, graphs, animations, and posters, is essential for facilitating comprehension.Further, we have observed that exercises in small groups can be particularly useful as these result in more lively discussions that help participants to engage in the topic, especially in those cases where the group of stakeholders is very diverse in terms of roles and background.
Second, achieving coherence between scenarios from different case studies may lead to loss of variety [78] as the use of common scenario elements can lead to the developed scenarios being too rigid or restricted.Kok et al., 2007 [10] propose that formal downscaling should be attempted only if there is particular scientific or policy value; if the aim is primarily to aid decision making constructing the scenarios for each scale using a broad common framework could suffice.
Third, aligning new scenarios with already existing scenarios at a national level that are accepted and used by stakeholders is a complex process.In EVOKED, for the Dutch case studies, we decided to map the already established national scenarios to the SSP.However, this is not a straightforward task and may introduce uncertainties when comparing case studies.Similarly, Hagemann et al. (2020) [37] note that if stakeholders are free to choose scenarios, the outcomes may eventually look very different, reflecting to a greater extent the local logic based on cultural and geomorphological conditions [37].
Further work to tackle some of the above limitations could involve at improved implementation of the participatory part of the process, in line with the elements of a living lab.Moreover, integrating stakeholders right from the start of the scenario development and increasing the iterations with the stakeholders would allow further integration of local knowledge.However, time availability often constitutes a major constraint in this process.Last, we must note that the development of scenarios has only constituted one aspect of the EVOKED project in the process of the production of climate services.Therefore, the proposed method should be considered in this context and adapted in projects or analysis that, for example, focus entirely on scenario development (e.g., MedAction, Visions).

Figure 1 .
Figure 1.Map of the study region and case study areas.Figure 1. Map of the study region and case study areas.

Figure 1 .
Figure 1.Map of the study region and case study areas.Figure 1. Map of the study region and case study areas.

Figure 2 .
Figure 2. Steps undertaken for the development of the local scenarios.

Figure 2 .
Figure 2. Steps undertaken for the development of the local scenarios.

Figure 4 .
Figure 4. Photo of the completed jigsaw puzzle exercise, showing the pieces of the graphical elements and the completed scenario puzzles.

Figure 4 .
Figure 4. Photo of the completed jigsaw puzzle exercise, showing the pieces of the graphical elements and the completed scenario puzzles.

Table 1 .
Overview of the implementation of the six-step multi-scale co-production approach in each case study region.

Table 1 .
Overview of the implementation of the six-step multi-scale co-production approach in each case study region.

Table 1 .
Overview of the implementation of the six-step multi-scale co-production approach in each case study region.

Table 1 .
Overview of the implementation of the six-step multi-scale co-production approach in each case study region.