Geology topics

Eve McDonald-Madden

Publications and source records attributed to Eve McDonald-Madden.

5 recordsLinked to original sources

USGS2018

Helping decision makers frame, analyze, and implement decisions

All decisions have the same recognizable elements. Context, objectives, alternatives, consequences, and deliberation. Decision makers and analysts familiar with these elements can quickly see the underlying structure of a decision. There are only a small number of classes of decisions. These classes differ in the cognitive and scientific challenge they present to the decision maker; the ability to recognize the class of decision leads a decision maker to tools to aid in the analysis. Sometimes we need more information, sometimes we don’t. The role of science in a decision-making process is to provide the predictions that link the alternative actions to the desired outcomes. Investing in more science is only valuable if it helps to choose a better action. Implementation. The successful integration of decision analysis into environmental decisions requires careful attention to the decision, the people, and the institutions involved.

Decision Point Online↗

USGS2018

Navigating the field of decision analysis

Managers, policy makers, and decision makers with responsibility for environmental decisions have an extraordinarily difficult job. The systems they manage are complex (coupled human-natural systems), with many dimensions and complicated dynamics. Our knowledge of how those systems respond to management actions is often limited, so many of the decisions have to be made in the face of uncertainty. The field of decision analysis provides a comprehensive set of tools for structuring, analyzing, and making decisions. Arising initially as a way to understand and manage risk, the field has expanded in the last 80 years to cover such topics as multipleobjective trade-offs, time-dependent linked decisions, the value of information, and competition among multiple decision makers. At the same time, cognitive psychologists and behavioural economists have studied human decisions to understand when our innate processes work and when they fail. Formal decision analysis is increasingly being used in many sectors, including economic, industrial, manufacturing, agricultural, transportation, and medical sectors, by individuals, corporations, non-profits, and government agencies. Occasional application of decision analysis to environmental decisions began in the mid-1970s, but the environmental science and management world was largely unaware of decision analysis until the late 1990s and early 2000s. (A major aim of the ARC Centre of Excellence for Environmental Decisions – CEED – was to better connect decision analysis and environmental management). Today, the field of environmental decision analysis is coming of age with the maturation of a rich set of tools to help decision makers frame, analyze, and implement decisions. Sometimes the wealth of tools is daunting, but there are core principles that tie decision analysis together. To help the reader navigate this complexity, we summarize the field of decision analysis in four key messages.

Decision Point Online↗

USGS2017

Oak habitat recovery on California's largest islands: Scenarios for the role of corvid seed dispersal

Seed dispersal by birds is central to the passive restoration of many tree communities. Reintroduction of extinct seed dispersers can therefore restore degraded forests and woodlands. To test this, we constructed a spatially explicit simulation model, parameterized with field data, to consider the effect of different seed dispersal scenarios on the extent of oak populations. We applied the model to two islands in California's Channel Islands National Park (USA), one of which has lost a key seed disperser. We used an ensemble modelling approach to simulate island scrub oak ( Quercus pacifica ) demography. The model was developed and trained to recreate known population changes over a 20-year period on 250-km 2 Santa Cruz Island, and incorporated acorn dispersal by island scrub-jays ( Aphelocoma insularis ), deer mice ( Peromyscus maniculatus ) and gravity, as well as seed predation. We applied the trained model to 215-km 2 Santa Rosa Island to examine how reintroducing island scrub-jays would affect the rate and pattern of oak population expansion. Oak habitat on Santa Rosa Island has been greatly reduced from its historical extent due to past grazing by introduced ungulates, the last of which were removed by 2011. Our simulation model predicts that a seed dispersal scenario including island scrub-jays would increase the extent of the island scrub oak population on Santa Rosa Island by 281% over 100 years, and by 544% over 200 years. Scenarios without jays would result in little expansion. Simulated long-distance seed dispersal by jays also facilitates establishment of discontinuous patches of oaks, and increases their elevational distribution. Synthesis and applications . Scenario planning provides powerful decision support for conservation managers. We used ensemble modelling of plant demographic and seed dispersal processes to investigate whether the reintroduction of seed dispersers could provide cost-effective means of achieving broader ecosystem restoration goals on California's second-largest island. The simulation model, extensively parameterized with field data, suggests that re-establishing the mutualism with seed-hoarding jays would accelerate the expansion of island scrub oak, which could benefit myriad species of conservation concern.

California↗

USGS2017

Two-step adaptive management for choosing between two management actions

Adaptive management is widely advocated to improve environmental management. Derivations of optimal strategies for adaptive management, however, tend to be case specific and time consuming. In contrast, managers might seek relatively simple guidance, such as insight into when a new potential management action should be considered, and how much effort should be expended on trialing such an action. We constructed a two-time-step scenario where a manager is choosing between two possible management actions. The manager has a total budget that can be split between a learning phase and an implementation phase. We use this scenario to investigate when and how much a manager should invest in learning about the management actions available. The optimal investment in learning can be understood intuitively by accounting for the expected value of sample information, the benefits that accrue during learning, the direct costs of learning, and the opportunity costs of learning. We find that the optimal proportion of the budget to spend on learning is characterized by several critical thresholds that mark a jump from spending a large proportion of the budget on learning to spending nothing. For example, as sampling variance increases, it is optimal to spend a larger proportion of the budget on learning, up to a point: if the sampling variance passes a critical threshold, it is no longer beneficial to invest in learning. Similar thresholds are observed as a function of the total budget and the difference in the expected performance of the two actions. We illustrate how this model can be applied using a case study of choosing between alternative rearing diets for hihi, an endangered New Zealand passerine. Although the model presented is a simplified scenario, we believe it is relevant to many management situations. Managers often have relatively short time horizons for management, and might be reluctant to consider further investment in learning and monitoring beyond collecting data from a single time period.

Ecological Applications↗

USGS2016

Detecting failure of climate predictions

The practical consequences of climate change challenge society to formulate responses that are more suited to achieving long-term objectives, even if those responses have to be made in the face of uncertainty 1 , 2 . Such a decision-analytic focus uses the products of climate science as probabilistic predictions about the effects of management policies 3 . Here we present methods to detect when climate predictions are failing to capture the system dynamics. For a single model, we measure goodness of fit based on the empirical distribution function, and define failure when the distribution of observed values significantly diverges from the modelled distribution. For a set of models, the same statistic can be used to provide relative weights for the individual models, and we define failure when there is no linear weighting of the ensemble models that produces a satisfactory match to the observations. Early detection of failure of a set of predictions is important for improving model predictions and the decisions based on them. We show that these methods would have detected a range shift in northern pintail 20 years before it was actually discovered, and are increasingly giving more weight to those climate models that forecast a September ice-free Arctic by 2055.

Nature Climate Change↗