Adaptation pathways is an analytical approach for undertaking planning and implementation under uncertainty and change. It provides a process for deciding when and under which conditions acceptable actions need to be implemented to cope with multiple plausible futures. In contrast, an "adaptive" pathway refers to the ability of a system to cope with different future uncertainties.

Adaptive decisions plan for change over time, such that management options can be tailored to different multiple plausible futures as they unfold, or as new information becomes available.

Adaptive pathways consist of sequences of options that can be taken and the circumstances in which alternative sequences might be selected. Adaptiveness is an attribute of a system allowing it to cope with different future uncertainties. Planning of "adaptation pathways" contributes to adaptiveness by enabling investment in an initial management option knowing that changes can be made during implementation of the plan.

Bottom-up vulnerability assessment refers to a form of vulnerability analysis that tests management options across a broad range of conditions, as opposed to evaluating performance of management options against a small set of "top-down" scenarios.

Breakpoint analysis identifies values of a variable at which a change occurs, e.g. a management option fails, or the preferred solution changes.

CART (Classification and Regression Trees) predicts outcomes based on binary splits of input variables. It can be used for scenario discovery to identify vulnerable scenarios for policy measures.

Crossover point scenarios describe states of the world or coherent storylines where the preferred alternative will change. They are intended to prompt discussion on whether the scenario is plausible, and iteratively build understanding of the problem, including possible contingency actions.

DAP is an approach to designing adaptive and robust plans based on concepts associated with Assumption-Based Planning (ABP) in which a plan is valid under an assertion (assumption) made about future conditions. The approach is also known as (Dynamic) Adaptive Policymaking.

DAPP is a framework to evaluate policy robustness and adaptiveness under deep uncertainty. The approach combines adaptive planning and adaptive pathways concepts.

Decision scaling is a method that links bottom‐up vulnerability assessment with multiple sources of climate information. It uses stochastic analysis to identify how climate affects objectives and uses runs from global climate models to then estimate probabilities of decision-relevant climate states.

Decision making under deep uncertainty (DMDU) is an umbrella concept with a focus on helping decision makers identify and evaluate robust and adaptive strategies.

Systems often exhibit a trade-off between efficiency and robustness. Management options that are efficient in one scenario may perform poorly in others. Options that perform well across scenarios may not perform as well as a less robust alternative. In tackling multiple plausible futures, this trade-off needs to be acknowledged, and can be explicitly analysed.

Expected shortfall is a probabilistic measure of risk (also called conditional value at risk, CVaR) which measures the average return of a management option for the worst cases, in the tail of a probability distribution. When faced with multiple plausible futures without probabilistic information, robustness metrics focused on the worst cases can be used as an alternative.

The best management option is the one that performs best on average if the future played out many times. With multiple plausible futures, without probabilities, the average is not well defined. An alternative is a robustness metric that selects options that perform well across many scenarios.

Exploratory modelling refers to the use of computer models to explore the implications of varying assumptions and hypotheses rather than performing experiments with a model as if it is the actual system.

A management strategy that defines an action based on the state of the system, relying on a constant flow of information in a closed loop, e.g. reservoir operation rules that specify releases as a function of water levels.

Flexibility of an initial management option relates to the ability to change options later to achieve desired outcomes, i.e. keeping options open. It can be evaluated with a range of dedicated robustness metrics.

Info-gap decision theory identifies the deviation from a best estimate (“uncertainty horizon”) within which an action will robustly provide a minimum acceptable “reward” or within which a windfall may be achieved, quantified respectively as “robustness” and “opportuneness”.

Decision making in the face of multiple plausible futures emphasises the need to design or select between competing management options to achieve certain objectives regardless of which of the plausible futures occurs. Depending on the context, management options may also be referred to as decision alternatives or policy alternatives.

The theoretical upper limit for adaptation of a water resource system, determined by identifying optimal management policies for all climate exposures of interest. Beyond this limit, infrastructure upgrades are required to manage future climate changes and reduce water resource system vulnerabilities.

In the context of adaptive decisions, a monitoring plan describes the information that will be collected and tracked in order to change path as the future unfolds or prompt the need to review plans. It may also be directly connected to triggers for actions.

MORDM is a method for robust decision making which combines interactive visual analytics with multi-objective optimisation to generate alternatives, evaluation of robustness metrics, scenario discovery to illuminate vulnerabilities, and trade-off analysis.

Management Option Rank Equivalence is a form of breakpoint analysis that reports the changes in each variable required to change the preferred management option.

"Multiple plausible futures" describes a distinct way of describing uncertainty about the future, with its own distinct approaches for handling that uncertainty.

Optimisation-based hypothesis testing seeks to identify plausible model scenarios that falsify a hypothesis. It can be used for vulnerability analysis.

A management option is "Pareto optimal" if there is no alternative that would perform better on all objectives. When the relative importance of objectives is contested, Pareto optimal solutions are commonly illustrated with a trade-off curve, such that decision makers can use information about the trade-off to select a preferred alternative.

Current decisions are conditioned by past decisions or actions, and affect the ability to implement required future actions in response to uncertain future changes.

Pareto Optimal Management Option Rank Equivalence (POMORE) uses multi-objective optimisation to locate minimum combined parameter changes that result in a change in the preferred management option.

PRIM (Patient Rule Induction Method) is a "bump-hunting" algorithm that identifies "boxes" in the input parameter space containing scenarios matching specified properties.

Reliability relates to the probability of success or frequency of failure. In the context of multiple plausible futures, without probabiliities, an alternative is the use of robustness metrics. In problems such as water supply, reliability can also be quantified within each plausible future, such that reliability and robustness are often seen together.

Risk is the effect of uncertainty on objectives; risk management is one approach to managing uncertainty. In the context of multiple plausible futures, risk can only be quantified in terms of how performance might play out across scenarios.

Robust decision making describes a variety of approaches that characterize uncertainty with multiple plausible futures and use robustness, rather than optimality, as a decision criterion.

Robust decisions perform well across a range of scenarios. Their robustness can be measured, or we can evaluate how their performance varies across scenarios, including circumstances where the plan of action fails.

Robust optimisation primarily refers to a form of optimisation where constraints are uncertain and defined by an exhaustive set of options. The need to be exhaustive is not usually met in the context of multiple plausible futures, but can be used as an approximation. Robust optimisation focuses on satisfying constraints in the worst case, which is one particular robustness metric. Problems where uncertainty is probabilistic can be solved using stochastic optimisation.

Robustness metrics measure how performance of an alternative changes in different scenarios

Scenario development describes multiple plausible futures either in terms of states of the world (e.g. values of different of variables), or coherent storylines. Each scenario should be based on different assumptions about the future.

Scenario discovery involves identifying policy-relevant plausible future scenarios as a form of vulnerability analysis or to identify a limited number of future scenarios to focus on

Scenario neutral climate impact assessment uses a stress testing approach to evaluate performance of a system across a range of plausible climate changes rather than individual scenarios.

Signposts specify the information and variables that should be monitored to track performance/progress, and identify whether anticipated vulnerabilities or opportunities are occurring. They are typically included in monitoring plans or adaptive pathways, and may be associated with triggers.

Stress testing involves evaluating how decision alternatives perform in different conditions.

Tipping points are thresholds or triggers beyond which adaptation strategies would be required for a system to cope with future changes. The term often also refers to a 'no-return point', in which a system shifts to a new state or regime.

In the context of adaptive decisions, triggers define conditions in which to implement new management options or seek additional information. Triggers therefore help track different multiple plausible futures as they unfold.

Delegation of uncertainty management passes responsibility onto others who are better informed or have greater capacity to address it, e.g. later in time at a level closer to the consequences involved.

Value of information relates to how much information will help improve a decision. In the context of adaptive decisions, it can help prioritise data collection, inform monitoring plans, or evaluate the benefit of keeping options open until information becomes available.

Vulnerability analysis identifies conditions in which management options can fail, either in terms of model parameter values or as scenarios.