Along the process of adapting to climate change, finding and defining appropriate adaptation measures is a complex task. Moreover, it is the key activity to increase the resilience to future climate-change-induced risks (Dogulu and Kentel, 2015). In addition, good practice in selecting adaptation measures is a fundamental task in adjusting water infrastructure to climate change, which is globally needed (Wilby, 2019). Part of this good practice is to analyse the impact of potential adaptation measures, not only in terms of hazard risk reduction, but also in terms of socio-economic effects, social justice or governance needs for implementation. For example, Zhou et al. (2012) combine climate modelling and an economic cost–benefit assessment in analysing climate adaptation measures for pluvial flooding in urban areas. Harrison et al. (2013) combine climate change scenarios with socio-economic scenarios in a digital platform to allow stakeholders to explore adaptation options within the context of varying futures. European research projects such as ECONADAPT and BASE have also focused on the economics of climate adaptation to support adaptation planning (Watkiss et al., 2015; Garotte et al., 2016; Meyer et al., 2015).

Another part of this good practice is to involve stakeholders in selecting and analysing these adaptation measures. Involving local stakeholders in these analyses, not just through consultation, but through co-production, enhances their relevance, usability, legitimacy and credibility (Palutikof et al., 2019). For example, Bhave et al. (2014) combine top-down climate modelling with bottom-up (involving stakeholders) prioritisation of adaptation measures but do not analyse socio-economic effect of measures or governance requirements for implementation. Andersson-Sköld et al. (2015) use focus group interviews with stakeholders to gauge the perceptions of adaptation measures, as part of a broader integrated framework to analyse the impact of climate adaptation measures. On the other hand, Singh et al. (2020) develop and apply a broad framework to assess the feasibility of adaptation measures, including political, economic and social indicators, but not specific to local conditions and not as part of a participatory framework. This study contributes to the literature by integrating three different analyses (governance, socio-economic and social justice) in a participatory framework, where most other studies capture only one or two of the above-mentioned dimensions (Verkerk et al., 2017; Bojovic et al., 2018; O'Sullivan et al., 2020).

This article presents the approach of selecting and analysing adaptation measures to increasing extreme weather events caused by ongoing climate change that was developed and applied in the H2020 project BINGO (Bringing Innovation to Ongoing Water Management). The purpose of this project is (a) to develop an integrated participatory approach for selecting and evaluating adaptation measures, (b) to apply and evaluate the approach across six case-study river basins across Europe, and (c) to support decision-making towards adaptation capturing the diversity, different circumstances and challenges river basins face across Europe. The project was conducted by over 20 project partners at six research sites in Europe: (1) the city of Badalona (Spain), which faces the risk of flash floods and combined sewer overflows (CSOs) due to increased precipitation; (2) the city of Bergen (Norway), also facing the risk of floods and CSOs due to increased precipitation; (3) Veluwe (Netherlands), a Natura 2000 site where long-term drought may affect the groundwater system; (4) the Troodos mountains (Cyprus), where decreasing precipitation causes water shortages for farmers and communities; (5) the Wupper River basin, which is divided into two sub-cases, one about flood risk due extreme weather events and one about decreasing water levels in the main water reservoir due to decreasing precipitation; and (6) the Sorraia Valley (Portugal) where farmers are confronted with water shortages due to decreasing precipitation.

The BINGO project followed a comprehensive approach from decadal predictions of weather events, to hydrological analysis of the impact of the weather events on water systems, to risk analysis and risk treatment. The work presented in this article focuses on the treatment of risks following extreme precipitation or drought. Risk treatment in project BINGO was organised as a collaborative process between scientists and local stakeholders, through communities of practice (CoPs) (Freitas et al., 2018). These CoPs consisted of representatives of local and regional governments, organisations involved in climate adaptation, and research partners. CoPs provide a social context in which researchers and stakeholders can engage in formal and informal interactions and co-analyse and co-produce the contextual knowledge that is necessary for climate change adaptation (Iyalomhe et al., 2013). The CoPs in the BINGO project were locally created and externally supported by the scientific project partners, which is found to be a necessary condition for a sustainable CoP (Vincent et al. 2018)

Based on the risks that were identified and analysed in the risk analysis, the CoPs selected and analysed adaptation measures, with the goal of informing decision makers about the expected efforts and gains from the implementation of these measures. The approach applied in the BINGO project is in line with steps formulated in the Adaptation Support Tool developed as part of the Climate-ADAPT initiative of the European Union (https://climate-adapt.eea.europa.eu/knowledge/tools/adaptation-support-tool, last access: 8 July 2021). More resources from the BINGO project can be found on the project website (http://www.projectbingo.eu, last access: 8 July 2021).

The next sections describe the process of selecting and analysing promising adaptation measures in the order as conducted within the BINGO project for all cases: (1) collecting and selecting adaptation measures, (2) governance analysis of selected adaptation measures, (3) analysis of socio-economic implications and (4) social justice analysis. These steps are illustrated with examples from the case study in the city of Badalona as well as from other sites in brief. A summary of the results of the analysis is provided, comparing different types of measures. Finally, conclusions are drawn on the application of the different methods.

Two approaches were applied to collect potential adaptation measures suitable to the climate change risks identified at the six research sites, namely a desk study of previous adaptation research and consultation of stakeholders involved in the local CoPs. For the desk study, the primary sources for adaptation measures were two previous EU research projects CarpathCC (http://www.carpathcc.eu/, last access: 8 July 2021) and PREPARED (https://cordis.europa.eu/project/id/244232, last access: 8 July 2021). From both projects databases were available with adaptation measures, including a brief analysis of their potential impact and risk reduction potential. From these databases the BINGO research partners selected measures that were (a) potentially relevant for the hazards the research sites are facing and (b) relevant for the main characteristics of the research site (e.g. urban area, agricultural area, natural area). At the same time, in each of the six research sites the first CoP meeting was organised. In this meeting, local stakeholders discussed and identified potential future climate hazards for their research site and identified measures that were either already planned or considered suitable.

These measures were collected as part of workshop reports (van Alphen et al., 2017a) and compiled, together with the measures from the desk research that were selected by the research sites. In total, 91 measures were collected. In many cases, research sites reported similar measures with slightly different wording, or very specific measures could be placed in a broader category. Through this reduction, 44 measures were compiled in a portfolio of adaptation measures (van Alphen et al., 2017b). The portfolio of adaptation measures is now available as an online tool (http://beta.tools.watershare.eu/bingo/\$/, last access: 8 July 2021). In the portfolio, different types of measures are distinguished: informational measures (e.g. raising awareness for behavioural change), financial measures (e.g. insurance and subsidies), regulatory measures (e.g. standards and legal bans) and infrastructural measures (e.g. flood control infrastructure). The complete set of measures can be filtered by type of risk, sector or adaptation objective. Since the portfolio was first created to support the work in BINGO, the broad risk categories reflected the risks first identified in the six case studies: (1) decrease in water quantity due to decreased precipitation, (2) decrease in water quality due to decreased precipitation, (3) floods due to increased precipitation and (4) decrease in water quality due to increased precipitation. The sectors reflected the sectors represented in the case studies: (1) agriculture, (2) flood management, (3) public water supply, (4) urban drainage, (5) water governance and (6) water resource management. This design was chosen so project partners and future users can easily find measures suited to their own circumstances. For each measure an analysis of the governance needs for implementation was given, based on the three-layer framework presented below. This analysis was done by research partners and was not based on specific conditions at the research sites, but on desk research.

After compiling this broad portfolio, a more specific assessment of potential risks at the research site was made and discussed with stakeholders. Local stakeholders could make a selection of adaptation measures from the long list provided by the project team and the measures that were developed locally. This first selection of measures was accompanied by a discussion on the following governance aspects related to the measures: (1) responsibility for implementation, (2) participation/division of roles, (3) availability of necessary resources and (4) potential challenges. During the CoP meetings at the six research sites, these aspects were discussed for the different measures, and a selection was made either through scoring or through voting. The measures were selected for the purpose of further analysis. For instance, in the case of Cyprus, measures were first scored on relevance and feasibility and then voted on by the stakeholders. In some cases, stakeholders decided to analyse measures that were not part of the portfolio but came up in the stakeholder process after the portfolio was already compiled. Table 1 shows the selection of measures for each research site.

In total, 22 measures were selected and analysed using the methods described above (Table 1). A majority of measures are technical or “grey” measures. This may be explained by the familiarity of the stakeholders and end users with this type of measure. The governance analysis shows that the knowledge and administrative resources for implementation of these measures are present at the sites, and implementation generally does not require the involvement of a broad range of stakeholders. Also, the effectiveness of these measures can often be modelled and is less uncertain than for instance behavioural measures. This is in line with Dhakal and Chevalier (2017), who find that, in the case of urban storm water management, technical solutions remain preferable throughout the world. However, the socio-economic analysis shows that these technical infrastructure measures are often expensive, particularly when compared with blue–green solutions or behavioural measures.

In the case of Badalona, the grey infrastructure proposed has the highest level of risk reduction but is also much more expensive than the SUDS and the EWS. In fact, the cost–benefit analysis shows that the investment and operational costs are not compensated for by the socio-economic benefits considered. The proposed SUDSs have a lower potential for flood and CSO risk reduction (also because the measures analysed only covered a small area of the city), but the improvements they bring, for instance to habitat creation, and enhanced aesthetic and recreational value (Locatelli et al., 2020) give them a higher net benefit. The EWS was the most cost-effective measure, significantly reducing flood risk.

When the measures are compared by their governance needs, we see a different picture. In Badalona, the measures that propose an increase in sewer capacity are part of an already existing urban drainage master plan. That means that the knowledge and competencies to implement these measures are readily available. This is labelled by Dhakal and Chevalier (2017) as a “pro-grey arrangement”. The main barriers to implementation are funding, political decision making and disturbance due to construction works. While the SUDSs require a smaller budget, there is limited experience on how to implement them, and technical expertise and standards/guidelines are currently lacking, although relevant knowledge can be obtained from regional examples or local research partners. SUDSs require the collaboration of a broad range of stakeholders, which requires coordination by the Badalona City Council. This makes implementation significantly more complicated than the proposed technical measures. These barriers are all acknowledged in other cases as well (Dhakal and Chevalier, 2017). The governance needs for the EWS are mostly met; the main challenge is to develop and implement the required protocols for the response to the “warnings” that the EWS gives.

In terms of social justice, as stated above, all measures have a positive impact on Badalona's citizens due to the decrease in the risk of floods and CSOs. The SUDSs have an increased benefit, due to their many positive side effects, but some of them are local and depend on where the measures are implemented.

The case of Bergen shows similar results. The sewer separation (a traditional engineering measure) shows the highest potential for risk reduction, particularly in cases of extreme rainfall. However, they are also very expensive in relation to the risk reduction achievable. The proposed SUDS measures are relatively low priced compared to their overall risk reduction potential, but they do not have the potential to reduce the risk of extreme events. When combined with using the roads as safe flood ways (a clever way of repurposing the grey infrastructure), they are able to handle peak flows in urban drainage at lower costs than sewer separation. The combination of blue–green–grey measures has been proven successful in other studies as well (Alves et al., 2020; Depietri and McPhearson, 2017).

In the case of Bergen, all the governance needs for implementation of sewer separation are met. With regard to SUDSs, there is still additional knowledge required on the performance of SUDSs in cold climates. The BINGO project was instrumental in involving the required stakeholders and so meeting the organisational needs. However, there were too few incentives for private property owners to implement the required measures (see also Dhakal and Chevalier, 2017). This can be circumvented by first implementing the SUDS at municipally owned properties. The implementation of safe flood ways is a less traditional technical solution. It adds a new functionality to roads that fall outside the responsibility of the road authorities and thus require coordination between different municipal authorities, an example of fragmented governance (Dhakal and Chevalier, 2017). Also, the broader impact on public safety when running water with high velocity through the streets needs to be assessed (Skrede et al., 2020).

All measures in Bergen benefit the general public because of the reduction of the risk of CSOs. As in the Badalona case, the SUDSs can provide many side benefits that have additional positive impacts. Negative side effects mostly involve construction and maintenance activities and resulting disturbances. Most measures are financed at the municipal level, reflecting the egalitarian or solidarity principle. SUDS or sewer separation implemented at private properties has to be financed privately, following the “polluter pays” principle (Strehl et al., 2019a). In the Veluwe case, the artificial retention measure, which involves constructing a large water transportation pipe from a nearby lake to the Veluwe, is the most expensive measure. It is also the measure with the highest potential for risk reduction, in this case measured as the additional groundwater recharge in the Veluwe groundwater system (approximately 30 Mm3yr-1). The additional recharge for the green measures (change in vegetation) ranges from 2–20 Mm3yr-1, but at much lower costs. Agricultural water restrictions are less expensive than the other measures. Most expenses go to helping farmers change their farming practices (or buying them out), since water restrictions will force them to change crops. However, the amount of water saved is relatively small (0.2–0.3 Mm3yr-1).

At the Veluwe, the implementation of artificial retention is relatively easy, because the required knowledge is available and the required coordination between actors is limited and can be achieved through existing institutions. The implementation of land use change is much more complicated. It requires the collaboration between stakeholders outside of current arrangements and with diverging interests. The BINGO CoP has already been successful in establishing this cooperation. More importantly, changing land use has a huge impact on public opinion, since the Veluwe is well protected (Natura 2000) and a cherished spot for recreation. Changing its vegetation at the required scale would require a public debate on forest management at the national level (van Alphen et al., 2019). Agricultural water restrictions require a locally embedded stakeholder process to be initiated, involving farmers, municipalities, water authorities and the province. It requires farmers to change their crops and farming practices, which are often considered part of the cultural heritage as well.

Artificial infiltration improves the sustainability of the drinking water supply and helps preserve the groundwater system. These benefits are distributed equally among water users in the region, who, through fees, also bear the costs. The negative side effects are mostly a decrease in attractiveness of the environment due to additional water (and energy) infrastructure. These negative effects disproportionally impact people who live near these infrastructures. Mitigation activities include minimisation of visibility and ecological effects. Land use change has a number of positive effects (sustainable drinking water supply, preservation of groundwater system, a more diverse and robust landscape, increased biodiversity) that impact the general public. Cost for these measures is borne mostly by land owners, who will be compensated by either the province or through water fees. Negative effects mostly have to do with the loss of wildlife and plants specific to pine forests (although overall bio-diversity will increase). During the transition period, tourism entrepreneurs may induce losses due to intensified foresting activities (Strehl et al., 2019a). For the measure agricultural water restrictions, farmers affected by the measure would carry the major burden, but would be compensated by the regional or national government for loss of production capacity. The local groundwater supply and natural environment are positively affected, which directly benefits land owners, local inhabitants and tourists.

In the case of Cyprus, the CEA shows that the most cost-effective measure is the maintenance of groundwater recharge systems (in this case check dams), yielding a 1250 m3 groundwater recharge per euro invested, compared to treated sewage water for irrigation (32.6 m3 recycled water used per euro invested), desalination (1.5 m3 desalinated water consumed per euro invested) and irrigation scheduling technologies (0.90 m3 water savings per euro invested). For the irrigation sector, the irrigation scheduling technologies measure had the highest MCA weighting score (13.5) compared to the treated sewage water option (12.1). For the domestic water supply sector, groundwater recharge systems received the highest final MCA score (14.6) compared to the use of water desalination (13.3) (Strehl et al., 2019a).

According to the governance needs analysis, the implementation of this maintenance scheme mainly requires better coordination between the water authority and the local community councils. Structural, institutional and political rigidities negatively affect the adoption of irrigation scheduling technologies in Cyprus. The lack of political will to charge irrigators with water prices that cover the full costs, i.e. financial, environmental and resource, does not provide an incentive to invest in water-saving technologies (van Alphen et al., 2019). Giannakis et al. (2016) suggest that the low irrigation water price elasticities, the ageing and lower training levels of farming population, the small farm size, and the low level of farm investments also impede the uptake of irrigation scheduling technologies. Support for farm training schemes, including issues such as water conservation and climate change adaptation, could improve the skills of the farmers and foster the adoption of new technologies (Giannakis and Bruggeman, 2015, 2018; van Alphen et al., 2019).

It follows from the governance needs analysis that the use of treated sewage water for irrigation could be implemented relatively easily. However, the total benefit is small, considering that only 6 % of the farmers have access to this source. Also, the long-term effects of possible contaminants are yet unknown. For desalinisation the key governance challenge is financial viability. Local households will pay a higher price for the desalinated water. Yet, as community councils will be responsible for selecting the source of water, there are concerns regarding the prioritisation of a cheaper source (van Alphen et al., 2019).

All four measures proposed in the Cyprus case study are financed at least in part by the sectoral groups/communities that benefit directly and/or indirectly. Irrigation scheduling technologies and the maintenance of groundwater recharge systems have potential side effects which benefit the general public as they increase the qualitative and quantitative state of the groundwater system. Desalination and the use of treated sewage water for irrigation only benefit specific groups of water users, namely the households of the downstream communities of Peristerona watershed (desalination) and the farmers that have access to the treated waste water. Also, they have notable negative side effects (impact of emerging contaminants, carbon emissions and brine discharges) which burden the general public and future generations (Strehl et al., 2019a)

In the first Wupper River basin case (insufficient reservoir storage due to drought), it was found that the technical infrastructure measures are very expensive compared to the behavioural measures, also in relation to the level of risk reduction. The reduction of low water elevation (which effectively reduces the outflow from the reservoir) is by far the most cost-effective measure (EUR 0.001 m-3). The water-saving scheme is not expensive, but the effect on the capacity of the reservoir is relatively low, and so is the cost effectiveness (EUR 0.194 m-3). Water transportation (EUR 0.040 m-3) and the horizontal well (EUR 0.054 m-3) rank in between. It was noted, however, that these infrastructural measures may be needed as a risk buffer in the future, since they provide redundant capacity.

With regard to the governance needs, the reduction of low water elevation is also the easiest measure to implement, with most of the governance needs being met at the research site. The main concern is the effect on the downstream ecology when the outflow of the reservoir is reduced. The technical infrastructural measures in this case are much harder to implement. To build the transfer pipe between the two catchments, water authorities and the environmental agency should be involved, as well as the property owners affected by the route. It also requires setting legal standards and assessing the technical feasibility and environmental impact. The same goes for implementing a new abstraction well (horizontal well). An additional barrier for that measure is the potential change in water quality (harder water) due to mixing of sources.

Reduction of low water elevation enhances social justice by securing the water supply to the general public, without increasing the price of water. The main negative side effect is a reduced flow passing the dam, leading to a decrease in energy production and potential decline of ecological quality. This also affects the general public. The substitution with an alternative water source may increase existing inequalities since it increases the price of water, which disproportionally impacts low-income groups. The same goes for the water transport between catchments, with the addition that it negatively impacts the property owners near the infrastructure. To the extent that these owners will be compensated by the Wupper Association, the cost will be carried by the general public. Finally, the pipe/channel route can have a negative impact on the environment and landscape, which impacts the general public.

For the second Wupper River basin case (flood risk due to increased precipitation) three technical measures were analysed. The retention basin is the most expensive measure (EUR 88 kyr-1), but it also performs best in terms of risk reduction. Alignment protection (EUR 10 kyr-1) and protection measures for property (EUR 3 kyr-1) have a much smaller risk reduction effect, about 10 to 15 times smaller. Since these results are calculated for a specific scenario, it cannot be assumed that just increasing the investment in the latter two protection measures will yield the same risk reduction as the retention basin.

According to the governance analysis, all requirements for implementing the retention basin are in place. With regard to the technical protection measure for property, one of the barriers for implementation is convincing the property owners to take action. Support and funding need to be coordinated between public services and property owners. Flood protection is considered a public service instead of a (partially) private responsibility. When this is the case, it does not encourage private investment (Geaves and Penning-Rowsell, 2016). In the case of alignment protection, this is indeed a matter of public action, where land may be acquired from property owners, but no investment from their side is necessary.

From the social justice analysis, it follows that the retention basin will benefit people downstream of the basin, while property owners above the endangered areas have the basin built on or near their properties. The basin will be financed by the Wupper Association, but property owners may face decreasing value of their properties because of negative environmental impact or decreasing aesthetics. This can be mitigated by an appealing design and environmentally friendly construction of the basin. A positive side effect of the basin that benefits the broader public is the improvement of water quality due to a reduction of direct run-off into natural streams. Protection measures for property are generally paid for by the property owners, who also reap the benefits of reduced flood risk. In the case of municipal buildings, the municipality has the opportunity to embellish public spaces by choosing an appealing design. The alignment protection will most likely be financed by the Wupper Association, as the measure benefits the general public. In the case property owners bear the costs, this will likely lead to increased social inequalities.

In the Sorraia Valley in Portugal, the technical measures involve the rehabilitation and modernisation of existing irrigation networks, which consist of a canal, a transport and distribution system, and a secondary irrigation system. Improving the canal is not the most cost effective in terms of cubic metres of water saved, but it is in terms of impacted area. Improving the transport and distribution system and the secondary irrigation system only affect a small area and need to be replicated in other areas to reach the same impact as improving the canal (Strehl et al., 2019a). The Tagus water resources management model has the potential to be very cost effective, but this is dependent on the level of use the water authority will promote. This is also the most important challenge with regards to the governance needs. Implementing the water resources management model requires a shift from a top-down management approach to a more network-oriented governance model. This requires an integrated approach to water resource management and the participation of a broad range of actors. The rehabilitation and modernisation of the irrigation networks pose no specific governance challenges, apart from acquiring funding for the investments.

From the social justice analysis, it follows that the rehabilitation and modernisation of irrigation networks mainly benefit the farmers, who also pay for the measures. To alleviate the financial burden, they can apply for funding. The assured agricultural sustainability in the region benefits a broader public as well. The water resource management model helps to better plan and manage water resources in the Tagus river basin, which benefits the general public. The costs are borne by the water authorities and then allocated to all water users through a tax or a fee.

The application of the BINGO approach has been successful in generating decision-relevant outcomes for developing adaptation strategies at the research sites. The governance analysis allowed to stakeholders to identify gaps in the governance needs to implement measures and to prepare steps to fill those gaps. The outcomes of the socio-economic analysis allowed stakeholders to prioritise measures by their cost effectiveness, cost–benefit ratio or performance on a broader range of criteria. Sometimes this yielded surprising results, such as the high cost effectiveness of check dams maintenance in the case of Cyprus. Finally, the social justice analysis can help stakeholders choose a proper financing mechanism that fits the desired principle (solidarity, egalitarian, deontological) and gives a first indication of how positive and negative impacts are distributed over different groups. Although the research sites were very different, in both their challenges and their socio-economic and institutional context, the approach presented in this paper yielded useful results in all cases. This supports the transferability of the approach to other cases in Europe.

However, we can identify specific benefits and limitations for each of the analyses (Table 2). The main benefit of the governance analysis is that it provides a systematic overview of the requirements for implementing a certain measure, with attention to a broad range of building blocks for adequate governance. This is not limited to technical and economic aspects, but also includes cultural, communicative and legal aspects. A limitation in the way that the method was applied is that it does not provide specific thresholds for the required level of these indicators, other than reported by the researchers and stakeholders involved.

The socio-economic analysis contributed in structuring decision-relevant information on adaptation measures focusing on potential outcomes of each measure. The methods applied help to quantify and/or rank indicators affecting costs and benefits of the selected measures, from a socio-economic point of view. Moreover, the methods can be integrated in a broader scenario-based approach to assessing adaptation strategies. Limitations of the method primarily deal with the availability of data, which has a strong effect on the validity and reliability of the conclusions drawn from the analysis.

Finally, the social justice analysis gives a broader perspective than the plain focus on the outcomes of adaptation and also considers the distributional effect on different groups in society. This may result in the identification of unbalanced burdens or co-benefits, which leads to better informed decisions and helps to realise climate justice. However, in the way the method was applied, the acquisition of meaningful social justice information and derived interpretations relevant for decision makers highly relies on the interview partners. They need to have a specific knowledge of the local adaptation measures/options planned and the socio-economic environment.