Guidance: Dynamics identification in Building with Nature

Types, sources and levels of uncertainty

What do we mean exactly when we talk about uncertainty in water management infrastructure planning? Do we only mean uncertainty relating to unpredictable future climatic or socioeconomic conditions? Are we referring to uncertainty in the technical performance of proposed solutions stemming from a lack of knowledge? Or do we also consider the uncertainty resulting from contested perspectives and values? In truth, uncertainty exists in multiple shapes and forms in planning problems, and it is important that practitioners are able to recognise its variety and extents in order to ensure that some uncertainties are not — consciously or otherwise — preferred or ignored over others during planning activities.  Three useful ways in which it can help to think about uncertainty in water management problems include its different types, sources and levels.

Types of uncertainty

Three main types of uncertainty can be distinguished (Brugnach et al., 2008).:

  • Unpredictability: i.e. ‘cannot know’: this relates to the uncertainty stemming from aspects such as naturally variable processes or the unpredictable interactions between stakeholders. Examples include: the occurrence of a future heat wave; the impacts of a heat wave on mangroves in a green-inclusive flood protection scheme; uncertainty about future election results that may change flood protection policy priorities. These uncertainties are always impossible to predict, which is a characteristic that we expect to remain so in the foreseeable future.

  • Incomplete knowledge: i.e. ‘do not know (yet)’: this relates to the uncertainty stemming from, e.g. lack of knowledge,  limited data availability or unreliable data. Examples include: the interactions between incoming waves with mangrove branch canopies; the maximum wave intensity that mangroves can resist; the economic impacts of extreme floods for populations located behind mangrove defences. This type of uncertainty can be reduced through additional monitoring and research to gain (more) knowledge or improve data availability/reliability.

  • Ambiguity: i.e. ‘know differently’:  this relates to the presence of different interpretations of the same concept or situation stemming from, e.g. stakeholders having different and sometimes conflicting perspectives of system behaviour (e.g. engineers vs. environmentalists, experts vs. generalists). Examples include: a strategic preference for cheap, single-focus solutions over more extensive integrated approaches; varying levels of trust in existing design standards and norms. This type of uncertainty may be irreducible, with the different perspectives each sharing equal levels of plausibility and/or legitimacy.

Practitioners need to appreciate that solutions must recognize and plan for futures  relating to all three types of uncertainties. In addition, when choosing for either ‘grey’ or ‘green’ solutions, they should be aware that it is unlikely that we will ever be able to completely reduce uncertainties characterised by their unpredictability or ambiguity.

Sources of uncertainty

Uncertainty can arise from different sources within the broader system. We distinguish between three main sources of uncertainty:

  • Natural system: uncertainty relating to the dynamics of natural processes, including both large environmental drivers of change (e.g. climate, water quality), as well as the response of ecosystems and ecological processes to those drivers (e.g. breeding patterns).
  • Technical system: uncertainty relating to the performance of interventions in the natural system, including both infrastructural (e.g. breakwaters) and technological (e.g. early warning systems). Here, the critical question is the extent to which we can be confident a proposed solution will achieve the desired impacts for a given application.
  • Social system: uncertainty relating to all the economic, cultural, legal, political, administrative and organizational aspects surrounding the problem and any proposed solutions. Examples include: the economic impacts of extreme weather events; societal values and attitudes towards alternative interventions; the absence of design, operation or maintenance standards or norms for NBSs.

Levels of uncertainty

Uncertainty can be experienced to varying degrees, which can be broadly described by four levels of uncertainty: near certainty, risk, uncertainty and so-called “deep” uncertainty. Each level of uncertainty benefits from different assessment and decision-making tools and methodologies.

  • Near certain future: refers to a setting in which the future is largely predictable. These are generally situations involving short-term decisions, in which the system of interest is well-defined and can be analysed using, e.g. deterministic system models. The decision-making process is one of ‘predict and act’, dominated by the formulation of ‘optimal’ policies.
  • Probabilistic future: refers to a setting in which the future can be described probabilistically. The impacts of potential interventions can be estimated using, e.g. stochastic models. Preferred interventions are then selected according to levels of acceptable risk. The decision-making process assumes that the future will largely resemble the past and existing trends.
  • Ranges of futures: refers to a setting in which there are a limited set of (equally) plausible futures, none of which can be assigned definitive weights or probabilities. In these cases, traditional scenario analysis-type assessments are required which assume that the future can be predicted well enough with “best-estimate” models. Decision-making tends to favour “robust” solutions that perform well across the range of plausible futures.
  • Unknown future: refers to a setting involving longer-term decisions in which there are many plausible futures or where the future is completely unpredictable and unknown. Understanding of system interactions may also be contested or unknown. Assessing the impact of possible interventions in such contexts is difficult and demands new approaches. These tend to favour solutions that exhibit both “flexibility” (ability to adapt to whichever conditions emerge) and “robustness” (producing favourable outcomes in a majority of futures).

The types of problems for which Building with Nature is proposed as potential solution are typically highly complex and mired in (deep) uncertainty. These are problem contexts in which all three types of uncertainty (unpredictability, incomplete knowledge, ambiguity) are present. The natural, technical and social processes contained therein are often highly variable and largely unknown. Also, the various actors involved cannot agree upon how the system functions or the likely future system state, let alone what interventions should be made to mitigate against the possible impacts.

Considering the different types, sources and levels of uncertainty provides a relatively simple means by which to identify the different uncertainties acting on these systems, as well as indicate the appropriate assessment and decision-making methodologies to apply in their analysis. In highly ambiguous settings, sufficient attention must also be given to stakeholder-inclusive processes of collaborative assessment to ensure that any ambiguities can be respected and/or reduced.