Advertisement

Regional Environmental Change

, Volume 18, Issue 6, pp 1801–1813 | Cite as

Assessing climate change adaptation strategies—the case of drought and heat wave in the French nuclear sector

  • Jyri Hanski
  • Tony Rosqvist
  • Douglas Crawford-Brown
Original Article

Abstract

Nuclear energy is a very important component of overall power supply in France. If the effects of future extreme weather events or climate shifts are not addressed, energy systems will be highly vulnerable to extreme weather events or shifts in weather patterns, such as changes in precipitation. Because of the deep uncertainties involved in climate projections and response strategies, any strategy implementation should perform adequately regardless of which scenario actually materialises. In this paper, we analyse the effects of drought and heat wave in the French nuclear energy sector using the Strategy Robustness Visualisation Method. The key feature of the method is the modelling of uncertainty of the quantitative indicators by (min, max) values plotted on radar plots such that each strategy option’s performance can be visually inspected for robustness. The method can be utilised as a “module” of its own in different uncertainty management approaches. Based on the case study, the presented adaptation strategies “Maintaining industrial production and final demand” and “Smart grid infrastructure” were more robust than the “No planned or automatic adaptation”.

Keywords

Climate change Adaptation Strategy assessment French Nuclear energy 

Introduction

Reducing GHG emissions from the energy sector has been a fundamental part of the European Union’s climate change strategies (policies), but the need for adaptation strategies is also recognised (European Commission 2013). The importance of adaptation strategies is highlighted because the effects of climate change are likely to occur even if global mitigation targets are met (Swart et al. 2009). If the future extreme weather events or climate shifts are not addressed, energy systems will be highly vulnerable to extreme weather events or shifts in weather patterns, such as changes in precipitation (Aaheim et al. 2013). The efficiency of electricity generation may be significantly reduced by climate change, due to decreased availability of water for cooling power stations during periods of drought, and restrictions on the return of water to rivers that is already at a high temperature due to extended heat waves (Aaheim et al. 2013; European Commission 2013). In addition, extreme weather periods such as heat waves and droughts will lead to increasingly significant demand peaks, potentially causing demand-driven overstress of energy infrastructure (European Commission 2013). The European Commission (2013) assessed the expected impacts on thermal power plants (including nuclear) to be medium negative by 2025 and extremely negative by 2080.

Nuclear energy is a very important component of the power supply of France, and any reduction in the ability of nuclear facilities to withdraw coolant water at sufficiently low temperature from rivers—and then return it to the rivers—will reduce the power available to the production activities of economic sectors. The energy system is a key part of the French economic system, which depends on reliable and affordable energy. Because of the long life cycles of energy infrastructure, investments in a climate-resilient energy system should be made before climate change affects the availability of water, in order to make the investments robust against climate scenarios. According to modelling results (Perrels et al. 2015), significant changes in vulnerability begin under both RCP 4.5 and 8.5 nearer to 2050 than 2100, and therefore, investments in a climate-resilient energy system must be in place soon after 2050. Additionally, the year 2100 was selected as the final year of the analysis, because it is at the outer edge of available climate projections (OFGEM 2011).

Climate change can be considered a change that introduces deep uncertainties. Under such conditions, calculations of expected values of objectives and constraints may not suffice to characterise the state of knowledge (Kasprzyk et al. 2013). In a situation involving deep uncertainty, decision makers do not know or cannot agree upon the full set of risks to a system or their associated probabilities (Lempert and Groves 2010; Kasprzyk et al. 2013). Thissen et al. (2017) introduced three alternative approaches to planning under deep uncertainty: resilience, robustness and exploratory modelling approaches. When considering climate change adaptation, there is a need for robust strategies (Hanski and Rosqvist 2016; Kasprzyk et al. 2013; Berkhout et al. 2013). Lempert et al. (2006) defined a robust strategy as a strategy that, in comparison to the alternatives, performs relatively well across a wide range of plausible futures.

Scenario analysis explores in a logical and internally consistent manner how the future may, could or should evolve from the past and present (van der Heijden 1996). In other words, scenario analysis explores different alternative future states (e.g. Refsgaard et al. 2007). Several different types of scenarios such as qualitative, quantitative, baseline and policy exist (Alcamo 2001; Refsgaard et al. 2007). The role of stakeholder engagement is highlighted in policy making. Stakeholders identify and clarify policy solutions and play a key role in policy learning (McAllister et al. 2014). Pielke and Sarewitz (2005) called for more focus on the integrated and multidisciplinary aspects of climate impacts and the information needs of decision makers.

Additionally, adaptive management is an alternative for robust strategies in response to uncertainty. It aims to enhance the resilience of systems by flexible, learning-oriented and experimental management (Fritsch 2017). In adaptive management, the strategies are based on current knowledge, predicted conditions and objectives, but must be flexible for adaption to future conditions as these emerge, thereby promoting a continual learning process (Hamilton et al. 2013).

There is still a need for methods that consider uncertain futures, deep uncertainty, scenarios, robustness and adaptation perspectives (Maier et al. 2016). There are a wide range of methods available for supporting uncertainty management, such as expert elicitation, Monte Carlo analysis, multiple model simulation, scenario analysis, sensitivity analysis and stakeholder involvement (e.g. Refsgaard et al. 2007). The need for methods for assessing the robustness of adaptation strategies is emphasised in many studies (e.g. Whateley et al. 2014; Lempert et al. 2006).

In order to meet this need, we introduce a method that is capable of visually demonstrating the robustness of adaptation strategies and can include multiple decision criteria. The Strategy Robustness Visualisation Method (SRVM) combines Multi-Criteria Decision Analysis (MCDA) and Robust Decision-Making (RDM) methodologies. SRVM can be adopted as a part of a complex and evolving process of informing governance under uncertainty. It involves uncertainty management elements such as stakeholder participation, scenario writing and adaptive management.

From the scientific perspective, the paper provides a new method for visualising the robustness of adaptation strategies. The method combines quantitative modelling-based information with expert opinion and visualises the results into radar plots. It is capable of visualising the performance levels of a large number of decision criteria. Additionally, the reliability of the method is increased by presenting a case study using the method. The main goal of the case is to validate the method.

From the managerial perspective, the method can be utilised to increase the robustness of adaptation strategies. The case study combines results from the ARIO model with expert opinion for a comprehensive understanding of the complex decision situation related to adaptation in the French energy sector.

In this paper, we test the method by presenting a case study for analysing the effects of drought and heat wave in the French nuclear energy sector and for assessing the alternative adaptation strategies to respond to them. The adaptation strategies are assessed particularly from the perspective of their robustness. The case deals with the possible reduction in nuclear power output during periods of extreme drought and heat in France. This paper extends the decision support methodology presented in Hanski et al. (2015) and Hanski and Rosqvist (2016) with a detailed case study.

Methodology

There are several complementary methods that can be used in supporting climate change-related decision-making (e.g. Hinkel and Bisaro 2014). Cost-Benefit Analysis (CBA) has been the dominant methodology (Scrieciu et al. 2014). However, there are several examples of other approaches, such as Multi-Criteria Decision Analysis (MCDA) or Robust Decision-Making (RDM), which have been successfully used in climate change-related decision-making (e.g. Lempert and Groves 2010; Porthin et al. 2013). The method is in line with the approach of Miller and Belton (2014), which emphasises the importance of iterative planning processes and adjustment of alternative adaptation strategies and policies when conditions change. Additionally, it is one contribution fulfilling the proposal in Berkhout et al. (2014) that different ways of representing climate change and risks should be matched to actor frames and decision contexts.

The case study presented in this paper utilises a Strategy Robustness Visualisation Method (SRVM) to support complex long-term decision situations (Hanski and Rosqvist 2016; Hanski et al. 2015). This is only the second case study using SRVM, and the method should be applied in other case studies in order to increase its reliability. SRVM is based on combined RDM and MCDA methods. The key feature of the SRVM is modelling of uncertainty of the quantitative indicators by (min, max) values plotted on radar plots such that each strategy option’s performance, under each scenario, can be visually inspected for robustness (see Appendix 1).

The SRVM follows MCDA in terms of performance modelling, but limits measurement to single criteria that remain separated, without combining the criteria into a single utility measure by an additive utility function. Similarly to RDM, SRVM does not assign probabilities to scenarios but describes the few scenarios in such detail that it enables the experts to give an opinion about the performance of the strategies under each scenario (Hanski and Rosqvist 2016).

In the method, the decision context, decision criteria and adaptation strategies are developed in cooperation with the stakeholders in an iterative process. The method can be used to visualise and evaluate the vulnerability of alternative adaptation strategies and to show the key trade-offs between the strategies to decision makers. The case study was conducted from early 2015 to late 2015. The method consists of seven iterative phases as shown in Fig. 1.
Fig. 1

SRVM supports complex climate change adaptation decisions under deep uncertainty

  1. 1.

    Decision context: identification of the stakeholders and all the important factors affecting the decisions, and selecting and involving the relevant stakeholders in the process. Learning from the successes and failures of past strategies (policies) (e.g. Schmidt and Radaelli 2004). Selection or creation of qualitative baseline and policy scenarios (scenario combinations) to test the performance of the strategies (e.g. Alcamo 2001).

     
  2. 2.

    Decision criteria: selection of the most important decision criteria for the strategies.

     
  3. 3.

    Adaptation strategies: Selection of the most important strategies to be assessed within the decision context.

     
  4. 4.

    Performance of the strategies: assessment of the adaptation strategies against each decision criterion, under all specified scenarios.

     
  5. 5.

    Results and uncertainty: visualisation of the performance of strategies, with uncertainty ranges specified by the (min, max) pairs of pessimistic and optimistic performance, using radar plots.

     
  6. 6.

    Recommendations: identification of the robustness, or non-robustness (vulnerability) and the key uncertainties of the strategies.

     
  7. 7.

    Monitoring and adjusting strategies: reviewing implemented strategies as the decision context changes.

     

SRVM is used to visualise robustness in a multi-criteria assessment situation, especially in long-term decision-making in which several futures can be envisaged. For defining robust strategies, we use the definition presented in Hanski and Rosqvist (2016): “a strategy is robust if its deviation from the current strategy is positively valued across the range of impact criteria under the given scenarios”. A robust strategy can be described as a low-regret strategy that performs adequately in most future conditions envisioned in the scenarios, where low-regret strategies are those strategies that are insensitive to both short- and long-term scenarios. The selection of low-regret strategies is ultimately made by decision-makers based on the visualisations and other available data.

Performance of the strategies cannot always be modelled by formal methods, such as simulation. Furthermore, the subjective opinions of experts are needed for comprehensive assessments, especially as these relate to decision criteria. SRVM uses only minimum and maximum values if value distributions of the performances are available. For the visualisation of robustness of performance with respect to several criteria, the radar plot technique was selected (see Fig. 2).
Fig. 2

Illustration of the results produced by SRVM under two different scenarios

Visually, robustness of adaptation strategies is shown by several line pairs which are close to each other for all scenarios (Hanski and Rosqvist 2016). According to Montibeller and Franco (2010), visual inspection of performance is found to be the most helpful way of supporting the choice of robust strategies. The radar plots above illustrate a hypothetical decision situation in which three strategies are compared using four decision criteria. The uncertainty that has been presented in the form of probability density function (pdf) is reduced to the two samples of pdf in order for the visualisation to be feasible. Percentiles 0.05/0.01 and 0.95/0.99 would be adequate to represent the variance for the purpose of identifying robustness. The solid lines describe the minimum performance received from the modelling or expert opinion, whereas the dotted lines represent the maximum. The minimum represents either the 0.05 percentile of the performance in the case of modelling, or the minimum scores from the valuation of stakeholders, whereas the maximum represents the 0.95 percentile or the maximum scores, respectively. The strategies are compared to a baseline producing a major increase (+ 2), no change (0) or a major decrease (− 2) of performance of the energy system with regard to these criteria. In strategy 1, employing visual inspection, there is a considerable variance in the minimum and maximum performance, but the performance is rather similar in both scenarios. In strategy 2, there is a high variance in the performance in scenario 1, but scenario 2 shows only a low level of variance. In strategy 3, the variance between the scenarios and the minimum and maximum performance values is low. Therefore, strategy 3 could be called robust, at least in comparison to the other strategies.

Effects of drought and heat wave on the French nuclear energy sector

Decision context

The case considers the possible reduction in nuclear power output during periods of extreme drought and heat in France. Since nuclear power represents a high percentage of the overall power supply of France, any reduction in the ability of nuclear facilities to withdraw coolant water at sufficiently low temperature from rivers—or to reject coolant water back into those rivers after use—will reduce the power available to the production activities of economic sectors. The energy system of France is a key component of the economic system; a vibrant economy depends on a reliable and affordable energy system. Therefore, the case examines how curtailment of the energy system due to extreme weather would “ripple” through the economy, producing indirect economic effects that could in some cases be as large as the direct impacts on the energy system.

Analysis of the economic impacts was carried out using the Adaptive Regional Input-Output (ARIO) model of the Cambridge Centre for Climate Change Mitigation Research (Crawford-Brown et al. 2013; Li et al. 2013). The model divides the French economy into 35 economic sectors from the World Input-Output Database, including the energy sector. Changes to both demand and supply of energy cause changes in the production capacity of the non-energy sectors such as manufacturing, in turn influencing the GDP of the overall economy. These changes become progressively smaller as the initial damage to the energy system, caused by a combined drought and heat wave, dissipate over time. In this case study, it is assumed that there is no long-term damage to the nuclear plant, only a temporary curtailment of power production.

Current climate scenarios reflect the projected decline in summer/spring precipitation, and the consequent reduced run-off to rivers in catchment areas, potentially reducing volumetric flow rates in rivers and increasing the temperature of water in the river, which in turn will affect the permitted abstraction and release rates of water from power plants. This could result in the partial shutting down of power production during extreme periods of drought leading to low water flow, exacerbated by water temperature increase when the drought is accompanied by a heat wave. The impact of a changing precipitation rate and ambient air temperature—as provided by the climate modelling component of the ToPDAd project—are normalised to historical data on the climatic conditions that led to the 2003 drought and heat wave that affected nuclear power production. Curtailments of power production during the scenarios of the present study were based on the number of days when drought and/or temperature were more severe than in the 2003 case.

Performance evaluations were obtained from representatives of three stakeholder groups: EdF, Eon and National Grid (UK). The three interviewees provided the scores for the separate criteria under the scenarios and adaptation strategies. These scores reflect judgments by the survey participants of the post-2003 changes to the nuclear power system and regulatory requirements that were introduced to improve performance under conditions of drought and heat. The scores were then fed into a Microsoft Excel-based program to generate the results presented in the figures that follow.

For the case, two adaptation strategies are selected and compared to a baseline of “no planned or automatic adaptation” strategy. The adaptation strategies consider two time horizons, 2050 and 2100. The scenario combinations support the decision makers in their evaluations and are used as input data in modelling the performance of the adaptation strategies. They are based on Representative Concentration Pathways (RCPs) and Socio-economic Pathways (SSPs). RCPs describe greenhouse gas concentrations and provide guidelines to emission projections, and SSPs describe global socio-economic development trajectories. The RCPs in turn modify the number of days of low rainfall (drought) and high ambient air temperature (influencing ambient water temperature in rivers used for coolant water—no impact on coastal cooling is assumed). For the case, RCP 2.6 and RCP 8.5, and SSP 1 and SSP 5 are selected, because these combinations represent the plausible extreme ends of climate and socio-economic predictions (Harjanne et al. 2014). In addition to the two extreme scenario combinations, a baseline scenario was selected. The detailed scenario assumptions are as follows:
  • Baseline. This refers to a future climate and socio-economic condition in 2050 that is basically similar to the current conditions. The baseline does not include climate change impacts and does not match any expected climate development. It is used to distinguish the impact of climate change and general socio-economic development from specific impacts due to changes in the future energy system.

  • Low climate change (RCP2.6+SSP1). The low climate change scenario is a sustainability-oriented, open and cooperative world with low adaptation needs. Global GHG emissions peak during the 2010s and decline substantially thereafter. Inequality both between countries and within economies is decreased as low-income areas develop rapidly. Technological development is also rapid. Economies are globalised and open, with strict environmental protection policies. This scenario assumes moderate to small climate effects and worldwide cooperation ensuring electricity supply connections. Energy and resource efficiency are emphasised, leading to lower overall energy demand.

  • High climate change (RCP8.5+SSP5). High climate change scenario represents a growth-oriented world with low regulation. There are high adaptation needs compared to the baseline, and global GHG emissions continue to increase throughout the twenty-first century. The world has chosen conventional fossil-fuel dominated development due to pressures from social and economic factors. This maintains faster economic growth across the world and helps to create resources for adapting to the climate change impacts, but does not lead to ambitious mitigation targets. Compared to the low climate scenario, the high climate change scenario is an extension of business-as-usual, leading to stronger climate impacts. Centralised, mainly fossil-based electricity production remains as the most important production form globally, although nuclear remains dominant in France. Other electricity generation forms are also important, especially in regions where they have become competitive or where there are strong supporting policy measures.

Decision criteria

In this phase, the decision criteria and their scale are selected. The following criteria are thought by the stakeholders and researchers to be the most important:
  • Peak power loss. Average loss of power generation capacity in the entire grid due to the drought/heat wave over the first week of the extreme weather event

  • Meeting final demand. The cumulative gap (percentage of MWh of demand) between supply and demand during the entire extreme weather event

  • Service to critical industries. The cumulative gap (percentage of MWh of demand) met for industries that are critical for GDP production

  • Power available for export. The total amount of power (MWh) provided to the export sector of the economy for sale elsewhere in the EU

  • Water availability for agricultural use. The extent to which water might be redirected from agriculture to power plants in a water basin during the period of the extreme weather event

For each decision criterion, we provide a specification of where the information needed to judge the performance of an adaptation strategy was obtained in the case (from modelling, other published research or expert opinion), and a scale for making the necessary judgement. Modelling results using the ARIO model were the source of information for the first three decision criteria, whereas the performance of the last two criteria was determined by expert opinion. The 2003 and 2006 heat waves/droughts were utilised as input for the modelling. A five-step scale (− 2…+ 2) was chosen due to the limited resolution in the modelling and to make it easier for the stakeholders to evaluate the future performance of the strategies (Appendix 2).

Adaptation strategies

Two adaptation strategies and a “no planned or automatic adaptation” strategy are considered in the analysis. These strategies were preselected by the authors and were considered the most relevant strategies by the interviewed experts. The used model also set limitations for the selection of adaptation strategies. However, in the future, other adaptation strategies relevant for the decision situation could emerge. Each is designed to reduce the total economic losses due to curtailment of nuclear power production during extreme droughts/heat waves. The following strategies were assessed:
  • No planned or automatic adaptation. Effects of climate and socio-economic change without planned or automatic adaptation

  • Maintaining industrial production and final demand. Residual power during a period of power curtailment is allocated in order to preferentially maintain industrial production and meet final consumer demand (for power) in France, accompanied by a reduction in power available for exports. Example, Residual power is allocated preferentially to both the economic sectors producing the greatest contribution to GDP, and to those firms that could cause bottlenecks in industrial production even if they themselves are not directly significant contributors to GDP.

  • Smart grid infrastructure. Residual power is allocated to maintain industrial production, final demand and exports, with a smart grid and smart buildings introduced to allow for reduction of non-essential energy use during “brown outs”. Example: The national grid is altered to include “smart grid” connections to end users, with power contracts in place to allow the grid operator to re-allocate reduced residual power to operations that (1) cannot be curtailed without significant loss of service to end users and (2) are significant contributors to GDP.

Performance of the strategies

In this phase, each adaptation strategy is assessed against each decision criterion, under all specified scenarios in turn. The method only uses two value pairs from the value distributions, which can be called optimistic (max) and pessimistic (min) values. The value pairs represent the variance in the performance of strategies. In the example below, the average values are also given but they are not used in the assessment of robustness. In the case of a distribution of values, the percentiles 0.05 and 0.95 are associated with pessimistic and optimistic, respectively. Similarly, the variation of expert opinions based on their specified decision criteria is reduced to pessimistic and optimistic values by taking the minimum and maximum scores.

The performance of the adaptation strategies regarding all the decision criteria is depicted in Appendix 3. The distribution and the averages are presented for both scenarios. Based on the scales presented in Appendix 2, the performances of the decision criteria are converted to a − 2…2 scale as depicted in Appendices 4 and 5.

Results and uncertainty

The performance of adaptation strategies is presented in Figs. 3, 4 and 5 using radar plots to present the robustness of the strategies. The uncertainty is shown by plotting two similar colour lines; one linking the minimum values, and the other linking the maximum values of the decision criterion. The higher up the arms of the radar plot the line is, the higher is the positive performance. For example, in case of peak power loss, higher performance means lower peak power loss. As uncertainty increases, the distance between these two lines increases. In this type of visualisation, each scenario line pair is plotted on the same radar plot. Robustness is shown by closeness of the line pairs to each other for all scenarios. Strategies that are insensitive to both short- and long-term scenarios are so-called low-regret strategies. In addition, the plot can also be used for ranking. The higher the score, the higher is the performance.
Fig. 3

Visualisation of the robustness of the strategy “Maintaining industrial production and final demand” in the years 2050 and 2100 compared to the baseline

Fig. 4

Visualisation of the robustness of the strategy “Smart grid infrastructure” in the years 2050 and 2100 compared to the baseline

Fig. 5

Visualisation of the robustness of the strategy “No planned or automatic adaptation” in the years 2050 and 2100 compared to the baseline

In general, all the performance scores in all the strategies are expected either to be worse or to remain unchanged. The adaptation strategies “Maintaining industrial production and final demand” and “Smart grid infrastructure” fare better than the “No planned or automatic adaptation” strategy both in 2050 and in 2100, as the performance scores of the latter strategy are further away from the centre of the radar plot. As expected, the performance of strategies in high climate change scenario is, in general, lower than that in the low climate change scenario. Especially, the minimum performance results of the strategies “Maintaining…” and “Smart…” are higher than that of the “No planned…” strategy, signalling a possibility of low-regret strategies.

Discussion and conclusions

The case study presented in this paper considers the possible reduction in nuclear power output during periods of extreme drought and heat in France. Because of the importance of nuclear energy to the French economy, any reduction in production has significant effects on the French economy. We argue that robustness is a key criterion in analysing the long-term performance of energy sector strategies. Using SRVM, two adaptation strategies “Maintaining industrial production and final demand” and “Smart grid infrastructure” and a “No planned or automatic adaptation” strategy were analysed regarding their robustness. According to the results, both adaptation strategies appeared to be more robust than the “No planned or automatic adaptation” strategy. Therefore, we suggest the implementation of adaptive management (e.g. Fritsch 2017; Hamilton et al. 2013) to develop and maintain the French energy system.

To support the implementation phase, we suggest the identification of implementation actions that are common for both basic strategies. Additionally, we suggest the identification of the options for combining or changing the adaptation strategies as new knowledge of future conditions emerges. More research is needed in combining adaptive management principles to robustness and SRVM.

The selected case was used for demonstration and did not directly lead to decisions, although real stakeholders were involved in the case. However, the authors and stakeholders generally agreed that the methodology improved the understanding of the decision makers in complex climate change adaptation-related decision-making situations.

The main goal of the case study was to validate the method. In the method, the robustness and vulnerability of the selected strategies is assessed using modelling and stakeholder-based information. The case study utilises information from two sources: modelling information from the ARIO model and expert opinion from stakeholders. Additionally, both quantitative and qualitative scenarios are utilised in the study. The ARIO model uses quantitative scenario combinations as an input, whereas the stakeholders use case-specific qualitative descriptions of the different scenario combinations to support their performance assessments. The qualitative case-specific descriptions have been found by the stakeholders to be helpful.

The radar plot was selected as the visualisation technique for this case study. In the case study, the robustness of the adaptation strategies is not as clearly visible as in the illustrative cases presented in Fig. 2. The illustrative cases present ideal cases of robustness, and in practical applications of SRVM, the radar plots will need more extensive interpretation for supporting decisions. However, other visualisation techniques such as bar charts could also be used. Testing the pros and cons of other visualisation techniques needs further research.

Limitations of this research stemmed from case study methodology, expert judgement and the scales used for performance assessment. The presented case was the second case assessed with SRVM methodology (Hanski and Rosqvist 2016; Hanski et al. 2015). Further research is needed to validate the SRVM and assess its strengths and weaknesses.

The SRVM supposes that experts’ judgments do not lead to disagreement on the min and max estimates of the performance criteria. If disagreement exists, more information is needed until the discrepancy is settled. Additionally, in some cases when experts estimate probabilities, they tend to fix on an initial value and then adjusting it (Cooke 1991). In this case, the resulting value may be biased towards the initial value (Skjong and Wentworth 2001).

In this case study, a five-step scale (− 2…+ 2) was chosen due to the limited resolution in the modelling and to make it easier for the stakeholders to evaluate the future performance of the strategies. However, a more extensive scale could be used in other decision contexts.

In conclusion, we argue that the SRVM can be a helpful method for policy learning regarding complex issues where decision-making is conducted under deep uncertainty. Assessing the robustness of adaptation strategies gives the decision makers an idea of low-regret strategies, i.e. strategies that are justified under most, if not all, scenarios. The method highlights the importance of engaging stakeholders and decision makers in the scenario process, and of visualisation of the results. Our key contribution to the identification of robustness by visual inspection can be utilised as a “module” of its own in different uncertainty management approaches, such as described by Refsgaard et al. (2007) and van der Sluijs et al. (2005). As a process, SRVM is closer to decision and scenario analysis than risk analysis, in which the probabilities of the scenarios are assessed, affecting the choice of strategies.

Notes

Funding information

The research presented in this paper was funded by the EU Framework 7 project Tool-support policy-development for regional adaptation (ToPDAd) (www.topdad.eu).

References

  1. Aaheim A, Crawford-Brown D, Axhausen KW, Ciari F, Stahel A, Perrels A, Nurmi V, Pilli-Sihvola K, Meyer B, Prettenthaler F, Richter V, Purwanto AJ, Heyndrickx C, Osborn T, Wallace C, Warren R, Jantunen J, Molarius R, Nokkala M (2013) Sector-level adaptation challenges in the literature. ToPDAd deliverable 1.1. Available: http://www.topdad.eu/upl/files/98434. Accessed December 2015
  2. Alcamo J (2001) Scenarios as tools for international environmental assessments. European Environment Agency, Copenhagen. Environmental Issues Report. Experts Corner Report. Prospects and Scenarios No. 5. Available: http://www.eea.europa.eu/publications/environmental_issue_report_2001_24. Accessed 2 March 2018
  3. Berkhout F, Bouwer L, Bayer J, Bouzid M, Cabeza M, Hanger S, Hof A, Hunter P, Meller L, Patt A, Pfluger B, Rayner T, Reichardt K, van Teeffelen A (2013) EEuropean responses to climate change: deep emissions reductions and main-streaming of mitigation and adaptation. RESPONSES project policy Brief No. D9. Institute for Environmental Studies (IVM), VU University Amsterdam, AmsterdamGoogle Scholar
  4. Berkhout F, van den Hurk B, Bessembinder J, de Boer J, Bregman B, van Drunen M (2014) Framing climate uncertainty: socio-economic and climate scenarios in vulnerability and adaptation assessments. Reg Environ Chang 14:879–893.  https://doi.org/10.1007/s10113-013-0519-2 CrossRefGoogle Scholar
  5. Cooke RM (1991) Experts in uncertainty—opinion and subjective probability in science. Oxford University Press, New York ISBN-10: 0195064658Google Scholar
  6. Crawford-Brown D, Syddall M, Guan D, Hall J, Li J, Jenkins K, Beaven R (2013) Vulnerability of London’s economy to climate change: sensitivity to production loss. J Environ Prot 4:548–563.  https://doi.org/10.4236/jep.2013.46064 CrossRefGoogle Scholar
  7. European Commission (2013) Adapting infrastructure to climate change. Commission Staff Working Document, SWD 137. Available: http://ec.europa.eu/clima/policies/adaptation/what/docs/swd_2013_137_en.pdf. Accessed December 2015
  8. Fritsch O (2017) Integrated and adaptive water resources management: exploring public participation in the UK. Reg Environ Chang 17:1933–1944.  https://doi.org/10.1007/s10113-016-0973-8 CrossRefGoogle Scholar
  9. Hamilton MC, Thekdi SA, Jenicek EM, Harmon RS, Goodsite ME, Case MP, Karvetski CW, Lambert JH (2013) Case studies of scenario analysis for adaptive management of natural resource and infrastructure systems. Environ Syst Decis 33:89–103.  https://doi.org/10.1007/s10669-012-9424-3 CrossRefGoogle Scholar
  10. Hanski J, Rosqvist T (2016) A method for visualisation of uncertainty and robustness in complex longterm decisions. In: Walls L, Revie M, Bedford T (eds) Risk, Reliability and Safety: Innovating Theory and Practice, CRC Press, p 2929–2936.  https://doi.org/10.1201/9781315374987-44 CrossRefGoogle Scholar
  11. Hanski J, Rosqvist T, Crawford-Brown D (2015) ToPDAd deliverable D4.3—demonstration description—visualisation of robust adaptation strategies. Available: http://www.topdad.eu/publications. Accessed Oct 2015
  12. Harjanne A, Nurmi V, Perrels A,Votsis A, Osborn T, Melvin T, Wallace C (2014) ToPDAd deliverable D2.1—climate hazard and impact scenarios. Available: http://www.topdad.eu/publications. Accessed October 2015
  13. Hinkel J, Bisaro A (2014) Methodological choices in solution-oriented adaptation research: a diagnostic framework. Reg Environ Chang.  https://doi.org/10.1007/s10113-014-0682-0
  14. Kasprzyk JR, Nataraj S, Reed PM, Lempert RJ (2013) Many objective robust decision making for complex environmental systems undergoing change. Environ Model Softw 42:55–71.  https://doi.org/10.1016/j.envsoft.2012.12.007 CrossRefGoogle Scholar
  15. Lempert RJ, Groves DG (2010) Identifying and evaluating robust adaptive policy responses to climate change for water management agencies in the American west. Technol Forecast Soc Chang 77(6):960–974.  https://doi.org/10.1016/j.techfore.2010.04.007 CrossRefGoogle Scholar
  16. Lempert RJ, Groves DG, Popper SW, Bankes SC (2006) A general, analytic method for generating robust strategies and narrative scenarios. Manag Sci 52(4):514–528.  https://doi.org/10.1287/mnsc.1050.0472 CrossRefGoogle Scholar
  17. Li J, Crawford-Brown D, Syddall M, Guan D (2013) Modeling imbalanced economic recovery following a natural disaster using input–output analysis. Risk Anal 33:1908–1923.  https://doi.org/10.1111/risa.12040 CrossRefGoogle Scholar
  18. Maier HR, Guillaume JHA, van Delden H, Riddell GA, Haasnoot M, Kwakkel JH (2016) An uncertain future, deep uncertainty, scenarios, robustness and adaptation: how do they fit together? Environ Model Softw 81:154–164.  https://doi.org/10.1016/j.envsoft.2016.03.014 CrossRefGoogle Scholar
  19. McAllister RRJ, McCrea R, Lubell MN (2014) Policy networks, stakeholder interactions and climate adaptation in the region of South East Queensland, Australia. Reg Environ Chang 14:527–539.  https://doi.org/10.1007/s10113-013-0489-4 CrossRefGoogle Scholar
  20. Miller KA, Belton V (2014) Water resource management and climate change adaptation: a holistic and multiple criteria perspective. Mitig Adapt Strateg Glob Chang 19(3):289–308.  https://doi.org/10.1007/s11027-013-9537-0 CrossRefGoogle Scholar
  21. Montibeller G, Franco A (2010) Multi-criteria decision analysis for strategic decision making. Chapter 2. In: Zopounidis C, Pardalos PM (eds) Handbook of multicriteria analysis, applied optimization 103. Springer-Verlag, Berlin.  https://doi.org/10.1007/978-3-540-92828-7_2 CrossRefGoogle Scholar
  22. OFGEM (2011) Adaptation to climate change. Report to DEFRA, Office of the Gas and Electricity Markets. LondonGoogle Scholar
  23. Perrels A, Prettenthaler F, Kortschak D, Heyndrickx C, Ciari F, Boesch P, Kiviluoma J, Azevedo M, Ekholm T, Crawford-Brown D, Thompson A (2015). ToPDAd deliverable D2.4. Available: http://www.topdad.eu/publications. Accessed 13 April 2016
  24. Pielke RA, Sarewitz D (2005) Bringing society back into the climate debate. Popul Environ 26:255–268.  https://doi.org/10.1007/s11111-005-1877-6 CrossRefGoogle Scholar
  25. Porthin M, Rosqvist T, Perrels A, Molarius R (2013) Multi-criteria decision analysis in adaptation decision-making: a flood case study in Finland. Reg Environ Chang 13(6):1171–1180.  https://doi.org/10.1007/s10113-013-0423-9 CrossRefGoogle Scholar
  26. Refsgaard JC, van der Sluijs JP, Højberg AL, Vanrolleghem PA (2007) Uncertainty in the environmental modelling process—a framework and guidance. Environ Model Softw 22(11):1543–1556.  https://doi.org/10.1016/j.envsoft.2007.02.004 CrossRefGoogle Scholar
  27. Schmidt VA, Radaelli CM (2004) Policy change and discourse in Europe: conceptual and methodological issues. West Eur Polit 27:183–210.  https://doi.org/10.1080/0140238042000214874 CrossRefGoogle Scholar
  28. Scrieciu SŞ, Belton V, Chalabi Z, Mechler R, Puig D (2014) Advancing methodological thinking and practice for development-compatible climate policy planning. Mitig Adapt Strateg Glob Chang 19(3):261–288.  https://doi.org/10.1007/s11027-013-9538-z CrossRefGoogle Scholar
  29. Skjong R, Wentworth B (2001) Expert judgement and risk perception. Proceedings of the Eleventh International Offshore and Polar Engineering Conference Stavanger, Norway, June 17-22, 2001. ISBN 1-880653-51-6Google Scholar
  30. Swart R, Biesbroek R, Binnerup S, Carter TR, Cowan C, Henrichs T, Loquen S, Mela H, Morecroft M, Reese M, Rey D (2009) Europe adapts to climate change: comparing national adaptation strategies. Partnership for European Research PEER Report 1Google Scholar
  31. Thissen W, Kwakkel J, Mens M, van der Sluijs J, Stemberger S, Wardekker A, Wildschut D (2017) Dealing with uncertainties in fresh water supply: experiences in the Netherlands. Water Resour Manag 31:703–725.  https://doi.org/10.1007/s11269-015-1198-1 CrossRefGoogle Scholar
  32. van der Heijden K (1996) Scenarios: the art of strategic conversation. John Wiley & Sons, New YorkGoogle Scholar
  33. van der Sluijs JP, Craye M, Funtowicz S, Kloprogge P, Ravetz J, Risbey J (2005) Combining quantitative and qualitative measures of uncertainty in model based environmental assessment: the NUSAP System. Risk Anal 25(2):481e492.  https://doi.org/10.1111/j.1539-6924.2005.00604.x CrossRefGoogle Scholar
  34. Whateley S, Steinschneider S, Brown C (2014) A climate change range-based method for estimating robustness for water resources supply. Water Resour Res 50:8944–8961.  https://doi.org/10.1002/2014WR015956 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.VTT Technical Research Centre of Finland Ltd.TampereFinland
  2. 2.Cambridge Centre for Climate Change Mitigation Research Department of Land EconomyUniversity of CambridgeCambridgeUK

Personalised recommendations