You are here: Home / Knowledge / Tools / Uncertainty guidance / Uncertainty guidance topic 2

Uncertainty guidance topic 2

When confronted with the uncertainty in climate information, a natural reaction is to ask climate scientists to improve knowledge and understanding, and to provide, as soon as possible, more reliable forecasts of future conditions. Unfortunately, even though knowledge will improve, uncertainty will remain inherent to adaptation decision making. In many cases, decisions, such as the replacement of existing infrastructure, cannot simply be delayed in the hope that more certain information will become available. The purpose of this section is to help users understand how uncertainty can be managed and how it can be reflected in decision-making processes.

Back to uncertainty guidance main page

What are ways to account for uncertainty in decision-making?

Different approaches have been developed to deal with uncertainty in planning. These approaches offer an alternative in situations where there is not enough certainty to unambiguously determine the best solution. Unfortunately no generally agreed framework exists to select a particular planning approach. Useful approaches and principles to employ when making decisions with inherent uncertainty are:

Adaptive Management
Adaptive management involves the selection of a strategy that can be modified to achieve better performance as one learns more about the issues at hand and how the future is unfolding. A key feature of adaptive management is that decision makers seek strategies that can be modified once new insights are gained from experience and research. They make choices based on their best assessment and that of people whose advice they value. Learning, experimenting and evaluation are key in this approach and are actively planned for in decision-making. Adaptive strategies work best in situations in which the decision time scales are such that incremental adaptation is possible and decisions can be updated as new information becomes available.

Scenario Planning
Faced with deep uncertainty, decision makers may choose to consider multiple plausible outcomes. This is the approach taken by scenario analyses. Scenarios present a set of different, plausible future conditions (or ‘states of the world'). Decision analysis is then undertaken to compare how well alternative policy decisions perform under these different future conditions. In addition to providing a useful description of uncertainty, scenarios can also bring clarity regarding the trade-offs made within the decision-making process. This is especially useful where stakeholders hold differing values and priorities

Robust or Resilient Strategies
This approach identifies the range of possible future circumstances that one might face, and then seeks to identify strategies that will work reasonably well across that range. A robust strategy can be defined as one that performs well over a very wide range of alternative futures. A familiar example of a robust strategy is portfolio theory as applied in financial investment, which suggests that risk calls for portfolio diversification. Decision support tools for robust decision-making are under development.

Resources for further reading:

Back to uncertainty guidance main page

What are different types of adaptation options?

Different types of options are available that decision-makers can use in planning adaptation under climate change. The most appropriate option will depend on the nature of the decision being made, the sensitivity of that decision to specific climate impacts and the level of risk which can be tolerated. Options include:

  1. Selecting "low-regret" (or ‘‘no-regret'') options that yield benefits even in absence of climate change and where the costs of the adaptation are relatively low vis-à-vis the benefits of acting;
  2. Selecting "win-win (-win)" options that have the desired result in terms of minimising climate risks or exploiting potential opportunities but also have other social, environmental or economic benefits.
  3. Favouring reversible and flexible options enabling amendments to be made;
  4. Adding ‘‘safety margins'' to new investments to ensure responses are resilient to a range of future climate impacts;
  5. Promoting soft adaptation strategies, which could include building adaptive capacity to ensure an organisation is better able to cope with a range of climate impacts (e.g. through more effective forward planning);
  6. Reducing decision time horizons (e.g. the forestry sector may choose to plant tree species with a shorter rotation time, see Hallegatte, 2009);
  7. Delaying action (which should not be confused with ‘ignoring the future'). This may be appropriate as part of an active long term adaptation strategy where it has been determined that there is no significant benefit in taking a particular action immediately.

Back to uncertainty guidance main page

What are no-regret adaptation measures?

‘‘No-regret'' measures are activities that yield benefits even in the absence of climate change. In many locations, the implementation of these actions constitutes a very efficient first step in a long-term adaptation strategy. For example, controlling leakages in water pipes or maintaining drainage channels is almost always considered a very good investment from a cost–benefit analysis point-of-view, even in absence of climate change. Improving building insulation norms and climate-proofing new buildings is another typical example of a no-regret strategy, since this action increases climate robustness while energy savings can often pay back the additional cost in only a few years. Dependent on the cost, on the other hand, climate proofing of existing buildings may not always be a no-regret measure.

Whether a measure is no-regret depends on the specific circumstances. For example, additional irrigation infrastructure can be a no-regret measure in regions that already face water scarcity. In other regions, considering the high investment costs, it would be beneficial only if climate change decreases precipitation significantly.

Once no-regret measures have been identified, it is important to know why these no-regret actions are not implemented yet. Many obstacles explain the current situation, including (i) financial and technology constraints; (ii) lack of information and transaction costs at the micro-level; and (iii) institutional and legal constraints. These obstacles can then be addressed in adaptation planning as a first step in a long-term adaptation plan. A study of about 100 spatial planning and water management projects in The Netherlands that were at least partially driven by climate change concerns has shown that taking into account climate change in the planning process generally improved the quality of the projects, usually without increasing costs (link, in Dutch).

Back to uncertainty guidance main page

How are uncertainties quantified?

The quantification of uncertainty can play a valuable role in informing decision-making. Quantification cannot eliminate uncertainty, but it can help to understand the levels of uncertainty we are dealing with. The full value of quantification critically depends on in its communication

Statistical methods and models play a key role in the interpretation and synthesis of observed climate data and the projections of numerical climate models. Such methods are especially important in addressing the question, "What changes in climate are occurring and are projected?" Probabilistic information can also be a useful way of explaining the likelihood of possible futures.

Probability
There are two types of probability; subjective and objective:

  • Subjective or inductive probability – an estimate based on the available information and strength of evidence, e.g. horse-racing odds or taking out insurance. This is also used in the development of climate projections (e.g. UKCP09). 
  • Objective or statistical probability – where all outcomes are accounted for, e.g. rolling dice or choosing a playing card at random.

Describing probability
Likelihood or probability of a future outcome is often interpreted and described in different ways. It is important to be consistent in how the terms used and how these equate to the data which underpins them. In the table, the IPCC provide a useful calibration of likelihood against a series of descriptive terms:

Likelihood scale  
Term Likelihood of the outcome
Virtually certain 99 - 100% probability
Very likely 90 - 100% probability
Likely 66 - 100% probability
About as likely as not 33 - 66% probability
Unlikely 0 - 33% probability
Very unlikely 0 - 10% probability
Exceptionally unlikely 0 - 1% probability

 

Back to uncertainty guidance main page

What scenario to use for adaptation planning?

Different scenarios of future climatic conditions are available. Most commonly used are outputs from climate models (see also What emission scenarios are the basis of most climate projections?). In addition, scenarios for potential impacts are required (e.g., run-off, health effects, biodiversity change), as well as scenarios for socio-economic developments that influence vulnerability (e.g. demography, income). The type and time-scale of the adaptation plan will determine the most appropriate scenarios to use. For initial assessments of vulnerability or sensitivity assessments incremental scenarios can provide information across a wide range of climate variations. For decision time horizons of less than 20 years, scenarios will be required representing ‘near future' and possibly ‘present-day' climates. However, for longer-term decisions (time-scales exceeding 20 years), such as decisions with long-lasting consequences and concerning long-lived assets, a range of climate scenarios developed from the output of a number of climate models should ideally be used, taking into account that outcomes often vary more across models than across emissions scenarios (for one model).

To capture the range of possible futures, scenarios are often developed in a set. However, in practice very often decision-makers or their advisors do not have the time or resources to explore all scenarios in a set or explore the outcomes of different models. In such a situation, an alternative would be to use the scenarios associated with the highest and lowest emission scenario, for the model that has the most extreme outcomes for the variables relevant for the adaptation problem at hand. In case time and resources are so limited that only one or two scenarios for one model can be used, which is unfortunately the case in many situations, the results should be interpreted with the utmost caution, from the perspective that future changes can be much different from that specific projection. This practice may be useful for awareness raising but can lead to maladaptation or overinvestments if the outcomes are used to design adaptation measures. For applications informing important policy decisions or major investment decisions, it is recommended that decision-makers should make use of the full range of scenarios and models available. This will:

  • assist in the identification of critical thresholds in the response of the exposure unit to climate change
  • allow decisions to be taken which are robust to the uncertainties in future climate.

Resources for further reading:

Back to uncertainty guidance main page

Document Actions