Policy and Society

ISSN: 1449-4035 (Print) 1839-3373 (Online) Journal homepage: http://www.tandfonline.com/loi/rpas20

Adaptation Pathways and Real Options Analysis:

An approach to deep uncertainty in climate
change adaptation policies

Joost Buurman & Vladan Babovic

To cite this article: Joost Buurman & Vladan Babovic (2016) Adaptation Pathways and Real
Options Analysis: An approach to deep uncertainty in climate change adaptation policies, Policy
and Society, 35:2, 137-150, DOI: 10.1016/j.polsoc.2016.05.002

To link to this article: https://doi.org/10.1016/j.polsoc.2016.05.002

© 2016 Policy and Society Associates (APSS)

Published online: 01 Feb 2017.

Submit your article to this journal

Article views: 658

View related articles

View Crossmark data

Citing articles: 5 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=rpas20
 Available online at www.sciencedirect.com

ScienceDirect
Policy and Society 35 (2016) 137–150
www.elsevier.com/locate/polsoc

Adaptation Pathways and Real Options Analysis: An approach to

deep uncertainty in climate change adaptation policies
Joost Buurman a,*, Vladan Babovic b
a
Institute of Water Policy, Lee Kuan Yew School of Public Policy, National University of Singapore, 469C Bukit Timah Road,
Singapore 259772, Singapore
b
Department of Civil and Environmental Engineering, Faculty of Engineering, National University of Singapore,
1 Engineering Drive 2, E1A 07-03, Singapore 117576, Singapore

Abstract
Governments face the daunting task of developing policies and making investment decisions for climate change adaptation in an
environment that consist of complex, interlinked systems with manifold uncertainties. Instead of responding to surprises and
making decisions on ad hoc basis, a structured approach to deal with complex systems and uncertainties can provide indispensable
support for policy making. This contribution proposes a structured approach for designing climate adaptation policies based on the
concepts of Adaptation Pathways, Adaptive Policy Making, and Real Options Analysis. Such an approach results in incorporation
of flexibility that allows change over time in response to how the future unfolds, what is learned about the system, and changes in
societal preferences. The approach is illustrated by looking at drainage policies and measures to address flooding in Singapore.
# 2016 Policy and Society Associates (APSS). Published by Elsevier Ltd. All rights reserved.

Keywords: Uncertainty; Climate change adaptation; Adaptation Pathways; Real Options Analysis; Adaptive Policy Making; Water policy

1. Introduction

Policymakers face growing complexity in committing to long-term investment decisions for large-scale
engineering and infrastructural systems. Designers and architects of such systems need to consider technical,
socioeconomic and political uncertainties, ambiguities, intricacies, and processes in the context of long-term strategic
deployment and operations of these systems. Under such backdrop water management is nowadays challenged by
climate-associated changes such as sea level rise and increased spatio-temporal variability of precipitation, as well as
by pressures due to population growth and rapid urbanisation. Dealing with these highly uncertain changes requires
different and new approaches to policy making and planning to develop robust systems.
Several studies have estimated the global costs of climate change adaptation (e.g. De Bruin, Dellink, & Agrawala,
2009; Stern, 2007; UNFCCC, 2007; World Bank, 2010). The World Bank (2010), for instance, estimated the global
adaptation costs between US$70 and US$100 billion per year in case of a 2 8C rise in global temperatures by
2050. Although the estimates in these studies have many uncertainties and inaccuracies (Sarkar, Begum, Pereira, &

* Corresponding author.
E-mail addresses: joost@nus.edu.sg (J. Buurman), vladan@nus.edu.sg (V. Babovic).

http://dx.doi.org/10.1016/j.polsoc.2016.05.002
1449-4035/# 2016 Policy and Society Associates (APSS). Published by Elsevier Ltd. All rights reserved.
138 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150

Jaafar, 2012) they indicate the enormous investments required, for a large part in physical infrastructure, but also in
‘‘soft’’ measures, such as change in institutions and policies, capacity building, and strategy development (World
Bank, 2010). Governments are facing the daunting task of making those adaptation investment decisions that
maximise welfare under the constraints of limited budgets and competing demands for resources. In addition, climate
adaptation investments cannot be considered in isolation: in most cases climate adaptation is one component in a quest
for sustainable development and climate adaptation is an additional criterion – not the main objective – to which a
policy or an investment needs to adhere.
Climate adaptation decisions need to be made in an environment characterised by uncertainties. Climate change is a
long-term process with many future unknowns and policies, and investment decisions that address climate change
usually have a long time horizon. IPCC (2014) projects a temperature rise between 0.3 8C and 4.8 8C by the end of the
century as compared to the beginning of last century, with both extremes having very different outcomes. Hence,
governments have to deal with uncertain information on phenomena such as sea level rise along their coasts, changes
in extreme rainfall events, or occurrence of droughts. In addition, as climate change and adaptation cannot be
considered in isolation, the environment and the intrinsically linked social, economic, cultural, political and other
systems form sources of many additional uncertainties and ambiguities that affect climate adaptation decisions.
Scientific research in various disciplines attempts to support policymakers in making their decisions in an uncertain
context. Traditionally, the work focused on reducing uncertainty through collecting and processing additional data –
addressing a lack of information – or by addressing random variation through statistical analysis (Walker, Haasnoot, &
Kwakkel, 2013). However, uncertainties in climate adaptation decisions can for the most part not be addressed by
addressing a lack of information or statistical analysis: data may simply not exist, for instance on rare, extreme events,
and no-one knows how the future will unfold. These ‘‘deep uncertainties’’ are unknowable at present, but can reduce
over time (Haasnoot, Kwakkel, Walker, & Ter Maat, 2013; Walker, Marchau, & Swanson, 2010). In response to deep
uncertainties, approaches have been developed to help policymakers and system designers in making adaptive plans:
plans that are flexible and that can respond when new information appears or when conditions in the environment
change. In the context of climate adaptation policy making, relevant approaches are Adaptive Policy Making (Walker,
Rahman, & Cave, 2001; Walker et al., 2013), Adaptation Pathways (Haasnoot, Middelkoop, Offermans, Beek, &
Deursen, 2012; Reeder & Ranger, 2011) and Real Options Analysis (Gersonius, Morselt, Van Nieuwenhuijzen,
Ashley, & Zevenbergen, 2011; Linquiti & Vonortas, 2011; Park, Kim, & Kim, 2014). In addition, there are many other
methodologies, tools and techniques to deal with uncertainties in general. A few examples are: scenario planning
(Swart, Raskin, & Robinson, 2004), Monte Carlo Analysis (Zhang & Babovic, 2012), Multi-layer Decision Analysis
(Harvey, Hall, & Peppe, 2012) and safety margin strategies (Hallegatte, 2009). Methodologies and tools are often
complementary, used in combination and can be developed for a specific context. The level of uncertainty they address
– lack of data versus deep uncertainty – also varies. In order to develop robust climate adaptation strategies, correct
framing of uncertainty and selection of appropriate approaches, methodologies and tools is of great importance.
In this setting, the objective of this paper is to analyse the levels, types and sources of uncertainty in climate
adaptation policy making (Section 2), develop a conceptual framework to deal with these uncertainties in a structured
manner based on the approaches of Adaptive Policy Making, Adaptation Pathways and Real Options Analysis
(Section 3), and demonstrate the framework to policy making for pluvial flood protection in Singapore (Section 4).

2. Uncertainty in policy making for adaptation to climate change

2.1. Adaptation to climate change

In general, the aim of investments and policies for adaptation to climate change is to reduce the risks that
individuals and societies face from extreme weather events and other climate and weather related impacts (IPCC,
2014). For a clear understanding of uncertainty in policy making for adaptation to climate change it is important to
clearly define a number of concepts that are often used in the current adaptation literature.
Firstly, adaptation to climate change is a very broad concept that can encompass different types, different attributes
and different applications of adaptation (Smit, Burton, Klein, & Wandel, 2000). For instance, adaptation can be
reactive or proactive, can occur in natural systems or socio-economic systems and can be targeted at different climatic
variables. For this reason one needs to define clearly the system that adapts, to what the system adapts and how the
system adapts (Haasnoot et al., 2013; Smit et al., 2000).
 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150 139

Adaptation to climate change is often associated with dealing with extremes and shocks, such as floods or droughts.
However, climate change is a slow process that happens over many decades: adaptation to sea level rise of a few
centimetres requires rather different measures than adaptation to an increased occurrence probability of a single event,
for example a five metre storm surge. Hence, adaptation to climate change is in a sense similar to improving systems to
deal with already existing climate variability, though because of climate change the variability may not be stationary.
Adaptation to climate change cannot be seen in isolation and should not be a goal by itself (Agrawal & Carmen Lemos,
2015): adaptation to climate change is part of sustainable development of countries and societies. In their quest for
sustainable development, policymakers have to make trade-offs between benefits and costs of adaptation measures,
opinions on how much risk is socially acceptable, and other development objectives.
Resilience is a concept that is frequently mentioned in the context of climate change adaptation; adapting to climate
change is associated with creating resilient systems (e.g. ADB, 2014; Linkov et al., 2014; Restemeyer, Woltjer, & Van
Den Brink, 2015). Although resilience is a widely debated subject and many interpretations exist, it generally means in
the context of climate change adaptation that systems (technical, environmental, social, etc.) are able to cope with and
adapt to disturbances and trends (Linkov et al., 2014). Resilience is often used as a normative concept by policymakers
and practioneers: it is something good. However, it could also be used in a descriptive way, i.e. measuring resilience
without implying that more resilience is better (De Bruijn, 2005). A resilient system needs to be flexible and able to
adapt to anticipated and unanticipated changes. Typically, designed, planned and engineered systems are devised for
an optimal or most likely range of parameters using incomplete and uncertain information. To deal with uncertainties
in systems they should not be based on a single or limited number of scenarios, but incorporate flexibility to adapt to
changing circumstances (Deng et al., 2013). From an engineering and design perspective, flexibility allows for better
performing systems in the face of climate change adaptation because flexibility allows for making changes when new
information emerges or when the circumstances are different than anticipated. From a policy perspective, however,
there is a need to balance flexibility and providing a stable, predictable environment; a high degree of flexibility could
also have negative impacts if it leads to making inefficient and ineffective ad hoc decisions. Yet, in this paper we use
flexibility in the context of designed and planned systems and consider it as a positive property of a system that is not
static or follows a fixed path.

2.2. Uncertainty

Policy making for adaptation to climate change is subject to many uncertainties from a variety of sources. From a
societal perspective, people have impact on the climate system, and the changes in the climate system have impact on
people. This can be depicted as a casual chain, which directly and indirectly drives policy (Smith & Stern, 2011), see
Fig. 1.
Each of the links contains different degrees and sources of uncertainty when it comes to determining (future) risks.
For the amount of future emissions, there is uncertainty within society (e.g. with relation to economic growth and
adoption of low-carbon technologies) and technology (e.g. new inventions and ‘‘technological fixes’’ to reduce
radiative forcing). To determine the changes in atmospheric concentrations, incomplete understanding of fundamental
biological, chemical and physical processes results in uncertainties. Similarly, weather changes are characterised by
physical, biological and chemical uncertainties. Environmental and societal uncertainties obscure detailed evolution
of impacts on systems. Finally, while exposures to risks by individuals change, the uncertainties in preferences and
values, uncertainties in vulnerabilities and uncertainties to new technologies make the exact nature of these changes
uncertain.
In addition to these uncertainties that are reflected in global and regional climate and impact models, policymakers
face a range of other uncertainties while making adaptation policies for climate change, such as political uncertainties
(e.g. if the policy will be supported by the coalition partners), budgetary uncertainties (e.g. competing demands for
government budget), and uncertainties in the external environment that can affect climate adaptation decisions (e.g. a
terrorist attack could move away attention from the climate discussion). This is summarised as the ‘‘policy
environment’’ in Fig. 1.
In order to deal with all these uncertainties in a policy making context, it makes sense to distinguish between
different dimensions and types of uncertainty. The literature provides different classifications and different fields can
use somewhat different terminology (Smith & Stern, 2011; Walker et al., 2003). Walker et al. (2003) distinguish
between the modellers’ and the policymakers’ view on uncertainty. The modellers’ view focuses on uncertainty in a
140 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150

Fig. 1. Casual chain of climate change and the role of policy (based on Smith and Stern, 2011).

rational, objective decision making process, while the policymakers’ view focuses on how to value outcomes in the
context of different goals, priorities and interests.
In the modellers’ view uncertainty can be classified along the dimensions ‘‘level’’, ‘‘nature’’ and ‘‘location’’
(Walker et al., 2003). The level of uncertainty delineates sections of the continuum from determinism to total
ignorance (Refsgaard et al., 2012; Walker et al., 2010), see Fig. 2. An important distinction is between imprecision or
statistical uncertainty on one side and deep uncertainty on the other side. Imprecision can be modelled and
probabilities can be assigned to events and outcomes. In cases of deep uncertainty this is not possible; it is not possible
to place a probability on an outcome, or at most it is known that the probability is imprecise. This is the case when there
is uncertainty about mechanisms and functional relationships. Deep uncertainty also covers ‘‘unknown unknowns’’
and ‘‘black swan’’ events (Taleb, 2007), i.e. events we currently cannot foresee or are completely unexpected.
It is not always possible to make a clear distinction between different cases as the uncertainty space is a continuum:
decisions in real life contain elements that are quite clear and can be modelled as well as elements that unknown. Most
currently available approaches, tools and methods in uncertainty analysis have been developed for cases of statistical
uncertainty (Walker et al., 2010). The models used are simplified representations of reality to make sense of the
complex decision environment and can typically handle imprecision well through probabilistic approaches or scenario
analysis. Dealing with deep uncertainty, however, is more problematic. Scientists prefer to reformulate policy
questions in order to deal with problems that can be dealt with in terms of imprecision and probability (Smith & Stern,
2011) and introduce simplifying assumptions in their models, limiting the real-world applicability.
For the nature dimension of uncertainty Walker et al. (2003) make a distinction between the two extremes of
epistemic uncertainty and variability uncertainty, which is often called aleatory uncertainty (Helton & Burmaster,
1996). Epistemic uncertainty arises from a lack of knowledge of the system, while aleatory uncertainty refers to

Fig. 2. Different types of uncertainty (based on Walker et al., 2010).

 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150 141

random variability in the system (Kiureghian & Ditlevsen, 2009). Aleatory uncertainty can be addressed by collecting
more data about system variables and performing statistical and probability analysis. For epistemic uncertainty this is
not possible as more research may not yield better outcomes.
The location dimension in the modellers’ view of uncertainty refers to where in the model uncertainty arises
(Walker et al., 2003). This includes aspects such as identification of the boundaries of the model, uncertainty in model
structure, inputs or parameters. Refsgaard et al. (2012) call this the ‘‘sources of uncertainty’’ and include uncertainty
due to multiple knowledge frames, which is the different ways of understanding a problem and finding solutions by
different people. Multiple knowledge frames may lead to ambiguity, which Refsgaard et al. (2012) include as an
additional category in the nature dimension of uncertainty. Although ambiguity can be located in Walker et al. (2003)
dimensions of modellers’ uncertainty (Dewulf, Craps, Bouwen, Taillieu, & Pahl-Wostl, 2005), it resonates more with
the policymakers’ view of uncertainty.
Ambiguity is the result of a process that involves different people with different backgrounds, experiences,
societal positions, values and beliefs (Brugnach & Ingram, 2012). Developing climate change adaptation policies
involves policymakers, scientists, engineers, and many other stakeholders with different frames and hence is full
of ambiguities. Also within these categories of stakeholders different individuals operate with different frames
resulting in ambiguity. Scientific and technical experts often tend to analyse problems from their own frame and
include other stakeholders only at specific points in the analysis (Dewulf et al., 2005). They only deal with
epistemic and aleatory uncertainty rather than with ambiguity, which may lead to results that cannot be
implemented. Connecting different frames requires social learning, negotiation and communication between
different people and communities.
Specifically in decision support for climate adaptation policies, communication on uncertainties of model
outcomes on the one hand, and in the climate adaptation decision making process on the other hand, is of utmost
importance for making informative, well-founded decisions (Smith & Stern, 2011). Scientists should understand the
information needs by policymakers, while policymakers should appreciate the possibilities and limitations of climate
and impact models – which have been mostly developed in the space of imprecision and statistical uncertainty – in a
co-production of science and policy (Dilling & Lemos, 2011). Clear, common understanding of uncertainties and
ambiguities is thereby essential in the development of flexible, adaptive policies.
While acknowledging ambiguity, the next two sections approach uncertainty from the modellers’ perspective. In
the concluding section we will reflect the value of the proposed approach from a policy perspective.

3. Approaches for dealing with uncertainty

Although support for policy making and planning under uncertainty has been developed over a long time, the need
for climate change adaptation policies and investments has provided a stimulus for developments in this area (Wise
et al., 2014). A policy making and planning approach can comprise of different methods and tools. For instance, in a
step-wise approach each step can consist of different types of analysis, such as scenario planning, mathematical
modelling and multi-criteria analysis. Below three relatively recent approaches for dealing with uncertainty in policy
making and planning are discussed: Adaptation Pathways, Adaptive Policy Making and Real Options Analysis. These
approaches have been used or developed for climate adaptation policy making.

3.1. Adaptation Pathways and Adaptive Policy Making

The Adaptation Pathways (AP) approach (Haasnoot et al., 2012; Reeder & Ranger, 2011) was specifically
developed to support policymakers with making water management decisions in view of climate change adaptation.
An essential element of AP is the idea of adaptation tipping points (Kwadijk et al., 2010). Rather than taking different
climate scenarios as a starting point and analysing the impacts of changes and possible measures, it assesses current
policies and management strategies and determines when they no longer meet the clearly stated objectives. Once this
happens an alternative policy or management strategy is required (or a ‘‘tipping point’’ is triggered), which in its own
right may also have a, ‘‘expiry date’’ in the future where it would no longer meet the objectives. In many cases there
can be multiple policies and management strategies supporting the same objectives or leading to similar outcomes, but
each of those has different costs and benefits, in addition to the fact that different stakeholders have different views and
hence may prefer different pathways.
142 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150

AP is better suited to deal with deep uncertainty in climate adaptation policies than ordinary scenario analysis
because it does not take climate scenarios as starting point. Instead of analysing what happens if a certain climate
scenario materialises (e.g. one metre sea level rise by 2100), it explores under what conditions a certain policy or
strategy will fail (e.g. a flood defence will no longer provide adequate protection if the sea level rises by one metre).
Regardless, AP requires computational scenario approaches to be carried out, such as Monte Carlo Analysis, to
analyse when tipping points may occur, and hence also operates in the space of statistical uncertainty. Although for the
pathways itself the exact timing of a tipping point is not important, the sequence of actions needs to be determined and
economic value of costs and benefits is time-dependent due to discounting of future costs and benefits. Nevertheless,
changing the climate scenarios only changes timing of actions, rather than the actions itself (Walker et al., 2013).
We clarify the AP approach in an example that will also be used in the subsequent sections. Suppose a city wants to
climate-proof its drainage network. Policymakers have decided that the drainage network should be designed to deal
with a one-in-ten year event, which means that flooding is on average only allowed to occur once in ten years (which
corresponds to a ‘‘clearly stated objective’’, which is required to determine tipping points in the approach). Based on
historical rainfall data and hydrological models engineers can assess the performance of the drainage system in
question and develop potential solutions where the drainage network does not meet the stated objective. Assume the
hydrologist expect that the current drainage system soon will not suffice and climate-induced rainfall intensity will
overwhelm the system more frequently. This means that an adaptation tipping point has been reached and new
drainage strategies need to be developed. Four different strategies are proposed: (1) build a diversion canal; (2)
implement a Next Generation Infrastructure (NGI) policy (promote green roofs, use porous pavements, parks and
green spaces for detention, slow down runoff by meandering rivers instead of straight canals, etc.); (3) construct a
detention tank; and (4) widen drains. Each strategy has a different ‘‘expiry date’’. For instance based on the middle
climate change scenario the diversion canal is expected to have sufficient capacity for the next hundred years; the other
strategies require a switch to an additional strategy at an earlier tipping point. The adaptation pathways map, see Fig. 3,
shows sequences of possible strategies, which will all fulfil the stated objective. Engineers may possibly prefer proven,
‘‘hard’’ solutions, such as a drainage tank and canal widening, while the city planning agency may prefer ‘‘green’’
solutions, such as green roofs and more green spaces.
In this respect, AP is basically an approach to help policymakers think about different measures and strategies and
make decisions over time. A flexible, adaptive strategy is being developed with sequencing of actions over time. This
is particularly useful in the context of complex, interacting systems.1 In practice, adaptation pathways are based on a
combination of modelling results and expert opinion; dealing with the future always requires dealing with expectations
and unknowns, which cannot all be modelled. Thus, the approach does not give a single, optimal solution, but rather
provides policymakers support in identifying opportunities, no-regret actions, lock-ins and timing of an action
(Haasnoot et al., 2013).
AP is relatively new and may still develop further. Wise et al. (2014) point out that AP requires clearly identified
policymakers and unambiguous goals for it to work, along with a conducive governance regime. In absence of these,
they propose a broader conceptualization of adaptation pathways of interacting global changes and societal responses,
and hence specifically include the societal change aspects of adaptation. Another development, discussed in Haasnoot
et al. (2013) is the combination of AP with the Adaptive Policy Making (APM) approach (Kwakkel, Walker, &
Marchau, 2010). This is a generic, structured approach for designing policy plans that can adapt to changing
circumstances. It consists of several steps, including an explicit step in which the plan is made more robust through
preparing shaping actions (to reduce failure or enhance success), mitigating actions, hedging actions and seizing
actions (of opportunities). It further consists of a step in which monitoring is set up, through signposts and triggers,
which are not unlike the adaptation tipping points. In the final step of APM trigger responses are prepared, which
provide alternative courses of action.
Haasnoot et al. (2013) combined AP and APM in the Dynamic Adaptive Policy Pathways (DAPP) approach. DAPP
consists of ten steps in which transient scenarios are developed, different actions identified, adaptation pathways
developed, and monitoring and feedback actions set up. The framework developed in this paper is derived from DAPP.
Section 3.3 will elaborate the different steps.

1
Note that the example is simplified and does not include the complexities of the entire drainage system, and interactions with other systems and
policies.
 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150 143

Fig. 3. Adaptation pathways map.

3.2. Real Options Analysis

Evaluation of public investments is often accompanied by cost-benefit analysis (CBA), which is a well-established
policy assessment method that attempts to quantify in monetary terms the value of all consequences of a policy to all
members of society (Boardman, Greenberg, Vining, & Weimer, 2011), although in practice many non-quantifiable and
unknown consequences are excluded from a CBA. Key element of CBA is the discounted cash flow approach to
calculate future costs and benefits to a present value. However, in case of investment decisions under uncertainty,
discounted cash flow may not be appropriate because it cannot capture managerial and policy flexibility (Feinstein &
Lander, 2002), which means that it cannot capture the changes in an investment project, which can be made in the
future. Hence, it will not correctly reflect the value of flexible projects designed under an adaptive policy framework.
Real Options Analysis (ROA) is able to take in into account uncertainty and flexibility, instead of assuming a
deterministic future required for CBA. Real options theory builds on the work of financial options theory (Black &
Scholes, 1973; Merton, 1973). The term ‘‘real options’’ was introduced by Myers (1977) and after initial applications
to uncertain decisions in the oil and gas industry, it now receives wide academic interest and industry application
(Borison, 2003), including application in water management and climate adaptation (e.g. Deng et al., 2013; Dobes,
2010; Gersonius et al., 2011; Jeuland & Whittington, 2014; Kontogianni, Tourkolias, Damigos, & Skourtos, 2014;
Park et al., 2014).
One of the premises of ROA is that flexibility has a value: the fact that a different pathway can be taken or that a
policymaker can adopt a strategy to wait until more information is available, allows one to limit the downside of
making a wrong decision, and capture the upside of new information and opportunities. Two types of real options can
be distinguished: real options ‘‘in’’ a system and real option ‘‘on’’ a system (Geltner & De Neufville, 2012). Real
options in a system are built into the design of a system, for instance making allowance for future expansion, and
require thorough engineering knowledge. The notion of real options on a system lies closer to the financial options
foundation of real options and includes managerial options such as the options to defer or abandon a project, or switch
to another project. Both types of options are relevant in climate adaptation investment decisions.
Monte Carlo Analysis and decision trees are most commonly used to calculate the value of real options (Buurman,
Zhang, & Babovic, 2009; Deng et al., 2013; Geltner & De Neufville, 2012; Gersonius et al., 2011; Zhang & Babovic,
2011; Zhang & Babovic, 2012). Essential for the computation of the value of real options for a cost benefit analysis is
that it assumes that probabilities can be given to possible future events. Hence in this respect ROA deals with
statistical, quantifiable uncertainty, while deep uncertainties are more difficult to assess. However, flexible system
designs and policies that result from incorporating ‘‘in’’ and ‘‘on’’ type options are better at handling situations of deep
uncertainty than traditional static designs and policies.

3.3. A framework for analysis

APM provides the policymaker with a step-wise approach to develop flexible, resilient plans. An adaptation
pathways map represents a lattice of possible future scenarios: a multitude of pathways is result of the ideation process
during which various design alternatives have been identified and proposed. The very existence of an adaptation
144 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150

pathway implies an incorporated flexibility about future courses of action. In this context, ROA enhances AP by
establishing the economic value created by flexibility, rather than merely comparing cost and benefits from individual
paths, which fails to recognise this economic value. The value (benefits) of flexibility can be compared against costs
required to incorporate the required degree of flexibility. APM provides a structure to develop adaptation pathways and
select promising pathways with the aid of ROA in a policy environment. Hence, combining the three approaches
provides a rigorous framework for guidance in developing flexible policies and justifying required investments.
From the perspective of different types of uncertainty as described in Section 2.2, ROA and AP are complementary
in the sense that ROA requires uncertainty to be quantified, while AP can, to some extent, guide policymakers in
developing flexible policies for deep uncertainty. In ROA to some extent deep uncertainty can be dealt with using a
uniform probability distribution function where every outcome has the same probability, and with scenario analysis.
However, beyond quantifiable uncertainty, systematic, quantitative decision making becomes difficult. In the absence
of modelling, APM is best suited to handle situations of deep uncertainty, though it is not strong in providing guidance
on how exactly the different steps are carried out, e.g. how vulnerabilities are identified or actions selected (Haasnoot
et al., 2013).
Fig. 4 presents the framework for developing adaptive policies using Adaptation Pathways and Real Options
Analysis. The DAPP approach, which combines APM and AP, is taken as a basis. ROA is inserted into several steps.
The framework consists of four main phases: framing and analysis of a problem, design of solutions, selection of
solutions, and execution of the plan. The four phases group eight distinct steps.
The first step (Framing) defines the system that is analysed and the interactions with other systems. Defining a large
system could make analyses in subsequent steps very complex. Hence, if possible, smaller systems with clear
interfaces between the systems should be considered. Furthermore, the current situation should be described, as well as
the objectives against which the performance of the system could be measured. A detailed analysis of uncertainties and
risks, across the entire uncertainty space as depicted in Fig. 2, and both directly related to the system as well as in the
environment, completes the first step. The second step carries out a thorough analysis of possible future situations and
establishes the gaps with the current situation using transient scenarios that describe the risks and uncertainties
identified in step 1. Potential opportunities should be explored in addition to vulnerabilities.

Fig. 4. Framework for developing adaptive policies.

 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150 145

The design phase is an iterative process consisting of three steps: identifying actions and triggers for those actions
(step 3), a first optimisation of actions (step 4), and development of adaptation pathways (step 5). Identifying actions
and options is partly a structured and partly a creative process that can involve many people. For instance, the design of
a ‘‘climate proof’’ drainage system for a new city expansion would follow a structured, iterative engineering design
process with different stages from determining the design criteria, which could include criteria for options, to final
design. In the stages creative input from designers and other stakeholders is needed to find valuable options, such as
phasing of the construction of the system or use of parks as flexible areas for water detention. All possible actions and
options should be considered, which, if combined, could result in a very large set of possible actions over time. The
optimisation step is a first evaluation of all possible actions and options: those that are not technically, practically,
socially or otherwise feasible should not further be considered. In this phase ROA provides a first screening by
identifying those actions and options that have little benefit and high costs. Feasible actions and options are combined
into adaptation pathways.
After the design phase still a large number of adaptation pathways could exist, especially as individual actions and
options can recombine over time, see for instance Fig. 3 in which four strategies combine into twelve possible
pathways. In the selection phase a manageable number of preferred pathways are selected. ROA is used to assess costs,
benefits and value of flexibility for each of the (screened) pathways. Aligning with perspectives of different
stakeholders, several pathways are selected using the results of the ROA.
The selected pathways are used in the execution phase to develop and implement a dynamic adaptive plan.
The dynamic adaptive plan provides information on which actions and decisions to take now, which can be
deferred, and how the preferred pathway can be followed (Haasnoot et al., 2013). Pathways that have not been
selected still remain active as options in case the future turns out to be incompatible with preferred plans. At a
future date, the plan could be updated with newly emerged information, going through another cycle in the
framework.

4. Planning for pluvial flood protection: a case in Singapore

4.1. Background

Singapore is a relatively low-lying, tropical, urbanised island. This poses three challenges in adapting the
drainage system of the country to climate change. Firstly, sea level rise could cause coastal flooding and reduce
the capacity of the drainage network outlets into the sea. Secondly, the drainage network needs to be able to
handle intense, tropical downpours and climate change models do not agree on how the frequency and intensity of
these downpours could change. Thirdly, urbanisation increases the amount of impervious area as more roads and
buildings are being built; less rainwater can percolate into the ground resulting in more water flowing into the
drainage network.
After experiencing widespread flooding in the 1960s and 1970s, the government spent more than two billion
Singapore dollar on drainage and flood alleviation measures, reducing the flood-prone areas by more than 95%
(Tortajada, Joshi, & Biswas, 2013). Floods were relatively minor until in 2010 a heavy downpour of about 100 mm in
2 h caused flash floods in the Orchard Road area, the main shopping belt of Singapore. Although the flood depths were
up 30 cm (Chan et al., 2012) and receded relatively fast, this was considered a major flood for Singapore. Spurred by
subsequent smaller floods the government’s focus on flood management was renewed. An expert panel was appointed
to conduct an in-depth review of the flood protection measures that are planned for Singapore in the next decade (Chan
et al., 2012). The recommendations of the panel included moving towards a risk-based approach for flood
management, improving engagement of stakeholders and the general public, and developing flexible and adaptable
systems to manage future uncertainty. Solutions to make a drainage system flexible and adaptable include making use
of Next Generation Infrastructure and creating diversification in the system. Next Generation Infrastructure
simultaneously aims reduce damage caused by floods and collect and store runoff water, which is an important
aspect in a water-stressed city like Singapore. Diversification implies that the system not only depends on traditional
canals, but other solutions such as retention ponds, detention tanks and alternative routing of runoff water through
diversion canals should be considered. Beyond the public drainage infrastructure, many other measures could be
conceived, such as building codes to avoid basement flooding, to flood insurance and public education to respond to
floods.
146 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150

4.2. Designing a dynamic adaptive plan

Designing a drainage policy for Singapore is hugely complex due to the many uncertainties and path-dependent
decisions. The aim of this section is to provide some insights into this complexity rather than designing a complete
dynamic adaptive plan, and to provide an example of applying the framework for developing adaptive policies. The
principles of some of the calculations for a ROA in the framework will be demonstrated with a simplified example for a
single catchment. Please note that the analysis below is for demonstration of the framework, hence the case is
hypothetical, non-exhaustive and is not based on factual data, though where possible information and numbers from
public sources have been used.
1. Framing. The system of analysis is the flood management system in Singapore. This comprises of the physical
infrastructure (e.g. the drainage network, tidal gates, and flood early warning system), and policies, procedures and
operations (e.g. drainage code of practice, listing flood-prone areas online, and convening an emergency task team
during flood). The flood management system has interactions with many other social, economic and environmental
systems; the most significant interactions would need to be defined in the framing stage. Some examples are the land
use planning system and the economic system, which would suffer from flood impacts. Section 4.1 describes in brief
the current situation; a full analysis would require a detailed description of the drainage system and interacting
systems. Objectives need to be measurable in order to assess the effectiveness of policies once implemented. For flood
protection the objectives are typically stated as levels of protection in terms of a return period. For instance, in
Singapore the design return period (as revised in December 2011) for a catchment with critical installations is 50 years
(Chan et al., 2012); other types of catchments have lower protection levels. Uncertainties and risks related to the flood
management system (and interacting systems) are manifold. Some of the more prominent uncertainties relate to
climate change: the uncertainties surrounding increasing rainfall intensities and frequencies, and sea level rise. These
could be considered statistical uncertainties as climate models can produce probabilities and uncertainty ranges.
However, currently models for Singapore do not yet give good results at the spatial detail required to make confident
rainfall projections and historical data missed the rare rainfall event with two subsequent peaks on 16 June
2010 leading to the Orchard Road flooding, which could in some ways be considered a ‘‘black swan’’. Additional
uncertainties result from modelling of the drainage network and interactions with other systems, such as public
opinion on flooding.
2. Analysis. Climate scenarios form the basis of developing possible future scenarios, though equally important
are socio-economic and spatial planning scenarios that affect potential flooding and flooding damage. Using the
scenarios and drainage models, the future flooding probabilities could be analysed and compared to the required
return periods. Assessment of socio-economic developments could be addressed through scenario analysis as
well. Analysis of extreme scenarios could, to some extent, be used to assess the impact of unexpected events. In
addition to vulnerabilities, opportunities should be explored. Examples are the development of Marina Barrage,
which serves as flood protection measure and created a drinking water reservoir at the same time, and the
opportunity to develop a green growth industry (NCCS, 2012) while developing knowledge in climate change
adaptation measures.
3. Actions. In our simplified example, the trigger for action is if the (modelled) return period for flooding in a
catchment is below 50 years. Fig. 3 shows four possible engineering actions: build a diversion canal, implement green
roofs and porous pavements (next generation infrastructure), build a detention tank, or widen drains. Additional
actions as proposed by Chan et al. (2012) include developing flooding standards using a more integrated risk-based
approach, better management of residual risk using flood risk mapping and improve engagement of stakeholders and
the general public. These will not be taken into account here.
4. Optimisation of actions and options. For the four engineering options a simple cost-benefit model was
developed, based on publicly available information and assumptions by the authors. A diversion canal is an inflexible,
expensive solution, but very effective, with an assumed reduction of the design return period from 50 to 100 years. A
question is, however, when to build the canal: if climate change is slow, it could make sense to postpone the
investment. In contrast, green roofs and porous pavements are a more flexible solution that could be implemented
gradually: the pace of increasing the area of green roofs and porous pavement could be based on the increase in
flooding probability due to intensifying rains and sea level rise. All four engineering options have positive benefits in
our model, hence all will be considered in the adaptation pathways map. However, once a diversion canal is built, no
other measures are required.
 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150 147

5. Development of Adaptation Pathways. Fig. 3 shows the adaptation pathways map for the four options.
Decisions on infrastructure and climate adaptation are typically path-dependent: a future option can only be exercised
if an earlier option was exercised, e.g. widening a diversion canal in the future is only possible if the canal was built in
the first place. This can be represented by a non-recombining tree, which is not unlike the adaptation pathways map in
Fig. 3. The tree basically lists all possible routes through the adaptation pathways map and is useful for calculating
costs and benefits of each route. Solving the tree (i.e. calculating the net present value of flexibility) can be complicated
though, and for some detailed examples we refer to Buurman et al. (2009), Luiz, Dyer, & Hahn (2005), Wang and De
Neufville (2004), De Neufville and Scholtes (2011) and Gersonius et al. (2011).
Fig. 5 shows the decision tree derived from the adaptation pathways map. There are six decision moments as
defined by adaptation trigger points; these are the points where any of a possible chosen strategy no longer fulfils the
stated flood protection objectives. For instance, if a tank was chosen in 2015, it no longer meets the flood protection
standard in 2035 and a new strategy needs to be chosen. The decision years in Fig. 5 are based on a specific climate
change scenario (25 cm sea level rise and 15% more intensive rain events by 2065) though these decision moments
could easily be replaced with the ones resulting from scenarios. Different scenarios could bring forward or backward
decision moments, but will not affect the pathways itself. Note that if a strategy still meets the objectives in a decision
year (e.g. building a diversion canal in pathway 1), the only logical decision is to continue with the same strategy. Also
note that more strategies in a pathways map could quickly lead to very large decision trees, which may be complex to
solve.
6. Selection of preferred pathways. For each of the twelve branches of the decision tree the costs and benefits can
be calculated. A model was set up in which in each year there is a random chance of a flood event. The probability of
flooding in each year increases over time due to climate change, and can be reduced by the proposed measures. The
investment and maintenance costs and the benefits (avoided damage due to a measure) are discounted to the present,
resulting in a net present value for each branch. In our model the branch with the highest benefits was number 9, first
build a detention tank, followed by drain widening in 2035. However, following this pathway could still lead to

Fig. 5. Adaptation pathways represented as non-recombining tree.

148 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150

significant floods events within the stated objective of a 50 years return period. If the government would only look at
the total benefits rather than net benefits, pathway 1 would be preferred, as it has the least cumulative damage costs due
to the significant increase in return period to about 100 years.
Next generation infrastructure is considered a more flexible strategy, as it can be implemented gradually, keeping
up with climate change. In our model, we linked the annual expansion area of green roofs and porous pavements to the
previous years’ sea level rise and rainfall intensification. Hence when there is no observed sea level rise and no rainfall
intensification, no investments are done in that year, while in case of a large change, the investment is above average.
Following this strategy rather than expanding each year with a fixed area, increased net benefits for the Next
Generation Infrastructure strategy by about 15%. Building a diversion canal is an inflexible, one-time strategy.
However, as in our model climate change variability is rather small, this more expensive, a more flexible strategy is not
preferred over this inflexible strategy. In ROA the value of flexibility increases as uncertainty (variability) increases.
Still, the flexibility of combining different strategies (tank + widening) performs better in terms of net benefits than an
inflexible diversion canal.
7. Specify and implement plan. Assuming the government chooses pathway 9 based on the analysis in the
previous steps, the detention tank will be constructed and a plan will be developed for widening of drains. However, if
based on new information in the future rainfall projections change, alternative plans with Next Generation
Infrastructure (pathway 7 and 8) or a diversion canal (pathway 6) could be executed instead.
8. Monitor & execute. Over time monitoring of the system and chosen strategies is required. The defined trigger,
the flood probability of 50 years, should be evaluated periodically, and if the probability is higher, the next strategy in
the pathway should be executed. If conditions change significantly, a new analysis should be carried out and the
process starts again from step 1.

5. Conclusions

This study describes a systematic methodology to address uncertainty in climate adaptation policies and
investments. The traditional policy making approach of focussing on best estimates and scenarios of the future may
lead to suboptimal system performance and flaws in the evaluation results. Adopting a strategy that allows for changes
if the future turns out to be different from what was expected leads to better policies and investments, which have a
higher net present value, as demonstrated in the example of drainage management in Singapore.
Incorporating Real Options Analysis in the proposed structured approach for designing climate adaptation policies
allows for monetary valuation of incorporating flexibilities in systems. Flexibility contains value in the context of
physical capital and climate adaptation investments, and this value is greater the more uncertainty or volatility exists in
the value of the underlying assets being developed. In essence, an option (or more generally optionality in adaptation
strategies) provides flexibility. Options allow the government to build later (or sooner), to build more (or less), or to
build different types of structures. Careful consideration is required of the different options and exercising options over
time in view of required policy objectives. The example showed that the pathway with the highest benefits is not the
most obvious one (a diversion canal), but the most obvious pathway does minimise damage costs. Hence, in a policy
context, the merits of using Real Options Analysis is that it provides a manner to objectively – though within the
limitations of quantification of costs, benefits and uncertainty – compare options and provide a valid argument to
incorporate flexibility, which often leads to higher investment costs and increased design complexity. Furthermore,
Real Options Analysis provides a different perspective on uncertainties: uncertainties cannot be avoided, but could be
re-cast to present valuable opportunities. Thus uncertainties and the flexibility offered by innovative solutions to the
design of climate adaptation policies need to be studied integrally.
The proposed approach of combining Real Options Analysis and Adaptation Pathways in an Adaptive Policy
Making process can help policymakers design comprehensive adaptive plans, and evaluate the costs and benefits of
each plan, possibly eliminating pathways that have clearly higher costs than benefits. The approach can handle
uncertainty from a modellers’ perspective, i.e. different levels of uncertainty including statistical and deep
uncertainty. Although it requires that information can be quantified – which sometimes involves making subjective
choices, particularly when dealing with deep uncertainty – the iterative framework (Fig. 4) results in adaptive policies
that are flexible and can be changed based on new information of circumstances, while at the same time provides a
roadmap of well-designed, realistic options. The portfolio of options in an adaptive plan should form the basis of
policymaking.
 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150 149

However, political realities may be different, and ambiguities may lead to different implementations of outcomes.
Flexibility and adaptation may not always be seen as a positive property in different frames and contexts. Neatly
crafted plans may be overridden in crisis situations, such as the immediate construction of flood defences after a
flooding disaster.
The approach can, however, add a significant value in the policy and planning process by providing a structured way
to generate, discuss and analyse options and prepare for different future outcomes, irrespective of simplifications
required to model complex real-world problems. Although policymakers may not follow the approach and its
outcomes, it can identify welfare maximising climate adaptation decisions and provides a way to make uncertainty
explicit.

References

ADB. (2014). Urban climate change resilience: A synopsis. Manila: Asian Development Bank.
Agrawal, A., & Carmen Lemos, M. (2015). Adaptive development. Nature Climate Change, 5, 185–187.
Black, F., & Scholes, M. (1973). The pricing of options and corporate liabilities. Journal of Political Economy, 81, 637–654.
Boardman, A. E., Greenberg, D. H., Vining, A. R., & Weimer, D. L. (2011). Cost–benefit analysis: Concepts and practice. Pearson/Prentice Hall.
Borison, A. (2003). Real options analysis: Where are the emperor’s clothes? Real Options Conference.
Brugnach, M., & Ingram, H. (2012). Ambiguity: The challenge of knowing and deciding together. Environmental Science & Policy, 15, 60–71.
Buurman, J., Zhang, S., & Babovic, V. (2009). Reducing risk through real options in systems design: The case of architecting a maritime domain
protection system. Risk Analysis, 29, 366–379.
Chan, E. S., Balmforth, D., Kan, Y. F., Koike, T., Lampe, L., Lim, K. K., et al. (2012). Report on Key Conclusions and Recommendations of the
Expert Panel on Drainage Design and Flood Protection Measures.
De Bruijn, K. M. (2005). Resilience and flood risk management. A systems approach applied to lowland rivers. Delft University of Technology.
De Bruin, K., Dellink, R., & Agrawala, S. (2009). Economic Aspects of Adaptation to Climate Change: Integrated Assessment Modelling of
Adaptation Costs and Benefits. OECD Environment Working Paper. OECD.
De Neufville, R., & Scholtes, S. (2011). Flexibility in engineering design. MIT Press.
Deng, Y., Cardin, M.-A., Babovic, V., Santhanakrishnan, D., Schmitter, P., & Meshgi, A. (2013). Valuing flexibilities in the design of urban water
management systems. Water Research, 47, 7162–7174.
Dewulf, A., Craps, M., Bouwen, R., Taillieu, T., & Pahl-Wostl, C. (2005). Integrated management of natural resources: Dealing with ambiguous
issues, multiple actors and diverging frames. Water Science and Technology, 52, 115–124.
Dilling, L., & Lemos, M. C. (2011). Creating usable science: Opportunities and constraints for climate knowledge use and their implications for
science policy. Global Environmental Change, 21, 680–689.
Dobes, L. (2010). Notes on Applying ‘Real Options’ to Climate Change Adaptation Measures, with Examples from Vietnam. Crawford School
Research Paper. Crawford School of Economics and Government.
Feinstein, S. P., & Lander, D. M. (2002). A better understanding of why NPV undervalues managerial flexibility. Engineering Economist, 47,
418–435.
Geltner, D., & De Neufville, R. (2012). Uncertainty, Flexibility, Valuation & Design: How 21st Century Information & Knowledge Can Improve 21st
Century Urban Development. ESD Working Paper Series. Massachusetts Institute of Technology.
Gersonius, B., Morselt, T., Van Nieuwenhuijzen, L., Ashley, R., & Zevenbergen, C. (2011). How the failure to account for flexibility in the economic
analysis of flood risk and coastal management strategies can result in maladaptive decisions. Journal of Waterway, Port, Coastal, and Ocean
Engineering, 138, 386–393.
Haasnoot, M., Middelkoop, H., Offermans, A., Beek, E. V., & Deursen, W. P. A. V. (2012). Exploring pathways for sustainable water management in
river deltas in a changing environment. Climatic Change, 115, 795–819.
Haasnoot, M., Kwakkel, J. H., Walker, W. E., & Ter Maat, J. (2013). Dynamic adaptive policy pathways: A method for crafting robust decisions for a
deeply uncertain world. Global Environmental Change, 23, 485–498.
Hallegatte, S. (2009). Strategies to adapt to an uncertain climate change. Global Environmental Change, 19, 240–247.
Harvey, H., Hall, J. W., & Peppe, R. (2012). Computational decision analysis for flood risk management in an uncertain future. Journal of
Hydroinformatics, 14, 537–561.
Helton, J. C., & Burmaster, D. E. (1996). Guest editorial: Treatment of aleatory and epistemic uncertainty in performance assessments for complex
systems. Reliability Engineering & System Safety, 54, 91–94.
IPCC. (2014). In R. K. Pachauri & L. A. Meyer (Eds.), Climate change 2014: Synthesis report. Contribution of working groups I, II and III to the fifth
assessment report of the intergovernmental panel on climate change. Geneva, Switzerland: Intergovernmental Panel on Climate Change.
Jeuland, M., & Whittington, D. (2014). Water resources planning under climate change: Assessing the robustness of real options for the Blue Nile.
Water Resources Research, 50, 2086–2107.
Kiureghian, A. D., & Ditlevsen, O. (2009). Aleatory or epistemic? Does it matter?. Structural Safety, 31, 105–112.
Kontogianni, A., Tourkolias, C. H., Damigos, D., & Skourtos, M. (2014). Assessing sea level rise costs and adaptation benefits under uncertainty in
Greece. Environmental Science & Policy, 37, 61–78.
Kwadijk, J. C. J., Haasnoot, M., Mulder, J. P. M., Hoogvliet, M. M. C., Jeuken, A. B. M., Van Der Krogt, R. A. A., et al. (2010). Using adaptation
tipping points to prepare for climate change and sea level rise: A case study in the Netherlands. Wiley Interdisciplinary Reviews: Climate Change,
1, 729–740.
150 J. Buurman, V. Babovic / Policy and Society 35 (2016) 137–150

Kwakkel, J. H., Walker, W. E., & Marchau, V. A. W. J. (2010). Adaptive airport strategic planning. European Journal of Transport and Infrastructure
Research, 10, 249–273.
Linkov, I., Bridges, T., Creutzig, F., Decker, J., Fox-Lent, C., Kroger, W., et al. (2014). Changing the resilience paradigm. Nature Climate Change, 4,
407–409.
Linquiti, P., & Vonortas, N. (2011). Real option analysis as a tool for valuing investments in adaptation to climate change. Economics of Adaptation
to Climate Change in Low-income Countries.
Luiz, E. B., Dyer, J. S., & Hahn, W. J. (2005). Using binomial decision trees to solve real-option valuation problems. Decision Analysis, 2, 69–88.
Merton, R. C. (1973). Theory of rational option pricing. Bell Journal of Economics and Management Science, 4, 141–183.
Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5, 147–175.
NCCS. (2012). Climate change & Singapore: Challenges. Opportunities. Partnerships. Singapore: National Climate Change Secretariat, Prime
Minister’s Office.
Park, T., Kim, C., & Kim, H. (2014). Valuation of drainage infrastructure improvement under climate change using real options. Water Resources
Management, 28, 445–457.
Reeder, T., & Ranger, N. (2011). How do you adapt in an uncertain world? Lessons from the Thames Estuary 2100 project. Washington, DC: World
Resources Institute.
Refsgaard, J. C., Arnbjerg-Nielsen, K., Drews, M., Halsnæs, K., Jeppesen, E., Madsen, H., et al. (2012). The role of uncertainty in climate change
adaptation strategies—A Danish water management example. Mitigation and Adaptation Strategies for Global Change, 18, 337–359.
Restemeyer, B., Woltjer, J., & Van Den Brink, M. (2015). A strategy-based framework for assessing the flood resilience of cities – A Hamburg case
study. Planning Theory & Practice, 16, 45–62.
Sarkar, M. S. K., Begum, R. A., Pereira, J. J., & Jaafar, A. H. B. (2012). Review of costs and methods for climate change adaptation. Journal of
Environmental Science and Technology, 5, 397–406.
Smit, B., Burton, I., Klein, R. T., & Wandel, J. (2000). An anatomy of adaptation to climate change and variability. Climatic Change, 45, 223–251.
Smith, L. A., & Stern, N. (2011). Uncertainty in science and its role in climate policy. Philosophical Transactions of the Royal Society A:
Mathematical, Physical and Engineering Sciences, 369, 4818–4841.
Stern, N. (2007). The economics of climate change [Online]. Cambridge University Press. (Accessed).
Swart, R. J., Raskin, P., & Robinson, J. (2004). The problem of the future: Sustainability science and scenario analysis. Global Environmental
Change, 14, 137–146.
Taleb, N. N. (2007). The black swan: The impact of the highly improbable. New York: Random House.
Tortajada, C., Joshi, J., & Biswas, A. K. (2013). The Singapore water story: Sustainable development in an urban city-state. Routledge: Oxford.
UNFCCC. (2007). Investment and financial flows to address climate change. Bonn: United Nations Framework Convention on Climate Change.
Walker, W. E., Rahman, S. A., & Cave, J. (2001). Adaptive policies, policy analysis, and policy-making. European Journal of Operational Research,
128, 282–289.
Walker, W. E., Harremoës, P., Rotmans, J., Van Der Sluijs, J. P., Van Asselt, M. B. A., Janssen, P., et al. (2003). Defining uncertainty: A conceptual
basis for uncertainty management in model-based decision support. Integrated Assessment, 4, 5–17.
Walker, W., Marchau, V., & Swanson, D. (2010). Addressing deep uncertainty using adaptive policies: Introduction to section 2. Technological
Forecasting and Social Change, 77, 917–923.
Walker, W., Haasnoot, M., & Kwakkel, J. (2013). Adapt or perish: A review of planning approaches for adaptation under deep uncertainty.
Sustainability, 5, 955–979.
Wang, T., & De Neufville, R. (2004). Building real options into physical systems with stochastic mixed-integer programming. 8th Real Options
Annual International Conference.
Wise, R. M., Fazey, I., Stafford Smith, M., Park, S. E., Eakin, H. C., Archer Van Garderen, E. R. M., et al. (2014). Reconceptualising adaptation to
climate change as part of pathways of change and response. Global Environmental Change, 28, 325–336.
World Bank. (2010). Economics of adaptation to climate change – Synthesis report. Washington, DC: World Bank.
Zhang, S. X., & Babovic, V. (2011). An evolutionary real options framework for the design and management of projects and systems with complex
real options and exercising conditions. Decision Support Systems, 51, 119–129.
Zhang, S. X., & Babovic, V. (2012). A real options approach to the design and architecture of water supply systems using innovative water
technologies under uncertainty. Journal of Hydroinformatics, 14, 13–29.