Online Training for Water Utilities: Plan
Deep uncertainty occurs when decision makers and stakeholders do not know or cannot agree on how likely different future scenarios are.
- If there’s not an agreement or knowledge or confidence in these future scenarios.
- When decision makers or stakeholders do not agree or do not know what consequences could result from their decisions.
These are called tools and methods for decision making under deep uncertainty (DMDU). These methods and tools can help water managers.
Water managers already conduct risk management. DMDU methods are intended to augment risk management, to pull in the impact of changing climate.
- Planning should consider multiple futures
- Robust plans perform well over multiple futures
- Plans that are flexible and adaptive perform better
- Computer analytics can explore multiple futures
The traditional model is titled here Predict and Act (Figure X): make a prediction of what the future conditions will be; based on that prediction, make a decision. These types of methods work well when uncertainties are small, when decisions are insensitive to changes, and when you’re confident about the future.
But when uncertainties are deep, “predict, then act” methods can break down as decision makers and analysts develop their own form of “tunnel vision” in which they focus on particular concerns and motivations without seeing the big picture. As shortcomings are identified, this can lead to decision-making gridlock or unexpected outcomes.
Under conditions of deep uncertainty
- Uncertainties are often underestimated
- Competing analyses can contribute to gridlock
- Misplaced concreteness can blind decision makers to surprise
Decision Making Under Deep Uncertainty (DMDU) begins with a proposed strategy and continues with stress tests of that strategy using multiple model runs to understand how that one strategy would perform under a range of plausible future conditions. Stress tests identify conditions under which a proposed strategy performs well and conditions under which it performs poorly. Rather than seek confidence in a specific model, one is seeking confidence in a decision.
- Can a robust and flexible strategy perform well under a range of future conditions?
- What uncertainties are most important?
- What actions do we need to take now?
- What actions can we postpone to the future?
“Backwards” analysis can help focus on important questions under deep uncertainty.
Methods In DMDU
Scenario planning is a widely used method of envisioning possible future conditions involving four components: (1) Identify a decision challenge; (2) chose key driving forces that are most important and uncertain; (3) flesh out scenario narratives; and (4) use scenarios to develop a robust adaptive plan.
Walking decision makers through developing contingency plans or adaptive plans as conditions change in the future.
- Which options to deploy first?
- What options to deploy next?
- How do we make our choices less vulnerable to uncertainties about the SLR scenario?
Robust Decision Making
- Stress test strategies over many plausible paths into the future,
- Use the resulting database to identify conditions where strategies fail, and
- Use this information to identify more robust strategies
Robust Decision-Making (RDM) examines a proposed strategy over a range of plausible futures. One may use a structured “XLRM” (Lempert et al. 2003) process to frame the decision-making process. First define the uncertainties (“X”); policy or management levers (“L”); relationships (“R”) among uncertainties, levers, and outcomes; and performance metrics (“M”), i.e., attainment or non-attainment of legal criteria. Simulation is used to evaluate possible decisions under unique combinations of uncertainties, levers, and relationships (Figure below). This procedure is known as a vulnerability analysis or a stress test. If a potential future scenario might lead to a shortage in supply or nonattainment of water quality criteria, one has identified a potential failure in performance. By evaluating a range of strategies, one may identify strategies that perform well over the broadest range of futures. Such a set of strategies is then considered “robust.” By revealing vulnerabilities, RDM and vulnerability analysis make it possible to anticipate and evaluate potential responses (Tariq et al. 2017).
Decision scaling focuses vulnerability analysis or “stress testing” on the climate components of a potential decision. One may use historic, observed conditions for one set of simulations. If a reliable statistically downscaled dataset is not available for evaluating climate change, one may statistically modify the observed dataset to generate alternative climate conditions, such as warm, hot, dry, wet, and any combination of those or other factors. Such perturbed climate time series data may be used to stress test a plan. Such a strategy allows one to quickly evaluate the potential influence of climate change on system performance.
The most important step is to get started. You could recruit a scientific advisory panel to advise you on how to integrate some of these methods and through a range of plausible futures for your region. You can also adopt these methods incrementally.
One possible sequence for mainstreaming these into your organization is to just to first off embrace the idea of having multiple futures. The next step could be to employ qualitative methods to evaluate ranges of futures for which a system would perform better or worse. A third step could be to quantify and rigorously think through how ranges of futures could affect your system using modeling and evaluation tools.