Once interested and affected parties have been identified and the smaller set of participants willing and able to work through a decision process has been established (as described in Chapter 6), the next task in the decision process is to clarify and structure a set of decision-specific objectives. Table 6.1 highlights the overlapping responsibilities and management objectives of agencies in the Spirit Lake and Toutle River system. Stated goals of the participants, however, do not align. In fact, when compiled, those goals may create a messy list of process concerns, positions, vaguely stated broad aspirations, and underlying interests. This chapter describes how the priorities and objectives of the different interested and engaged parties may be identified and organized and how metrics to forecast expected performance of those potential alternatives against those objectives may be derived. The discussion then leads to how the decision participants might start formulating a wide range of alternatives across the system to meet specific objectives.
Decision objectives are the identified goals that are to be attained or accomplished through decision making (Keeney, 1988). They are always phrased as verbs with a direction—for example, to maximize economic well-being, or to minimize adverse environmental impacts—but they are
only qualitatively defined until some metric (called an attribute in the language of decision theory) is assigned for their measurement. Objectives are linked to fundamental interests (Fisher et al., 2011) and provide the underlying motivations for participant positions on different management alternatives. Part of the process of identifying objectives is developing a common understanding of the motivations and priorities of decision participants. For instance, an objective that the committee heard repeatedly during public session meetings in Kelso was restoring the “naturalness” of the system. Like improving “sustainability” or “ecological health,” restoring naturalness seems a worthy goal, but naturalness means different things to different people. The term sometimes seemed to refer to the ability of fish to migrate past the sediment retention structure (SRS) and into Spirit Lake. At other times the term seemed to be an appeal for “pristineness” in the Mount St. Helens National Volcanic Monument (the Monument) or to management solutions that require little human intervention. Since the term “naturalness” seemed to be an important issue cared about by many, further discussions to define the elements of a “natural” system are warranted to help ensure that decision participants share a common understanding of the term. Furthermore, this common understanding should be adequately reflected in the objectives hierarchy (discussed later) in a way that can be useful for comparing alternatives.
People often refrain from generating or considering important decision objectives if not prompted through a careful and structured elicitation (Bond et al., 2008). At the same time, broad conversations with the public might lead to an unmanageably large list of overlapping concerns that are a mixture of potential solutions, process concerns, or issues not related to the decision at hand (Gregory et al., 2012). The decision process should result in development of a complete, compact, and structured set of decision objectives that includes all the objectives identified by the agencies and other interested and affected parties. This list will serve as the basis for exploring trade-offs among decision alternatives later in the decision process. Since participants in these processes often do not come into the discussion with clearly defined values, the decision process should help
participants iteratively construct and refine their values over time. The list of decision objectives will then need to be organized and structured.
Explicit elicitation of the participants’ decision objectives helps to build trust in the process on the part of all participants. The subsequent sections provide guidance as to how a small group of interested and affected parties might develop a set of decision objectives related to managing catastrophic and chronic flood risk as well as sedimentation in the region.
Drafting Decision Objectives
If the purpose of developing a list of decision objectives is to help compare alternatives for managing flood risk and sedimentation, the planning objectives used in the U.S. Army Corps of Engineers’ (USACE’s) original plan (USACE, 1983: IV-11) might work as a starting point for discussion. Management alternatives could be compared based on how they impact
- Flood control, including both the risk of a catastrophic breakout of Spirit Lake and the chronic flood risk arising from sedimentation and high seasonal flows;
- Water quality;
- Fish and wildlife; and
- Maintenance of cultural resources.
Based on comments received from interested and affected parties during the committee’s public meetings, this preliminary list might be elaborated to include
- Ecosystem services (e.g., reestablishing the pre-eruption landscape of the Toutle River);
- Cost, including both the expected cost to implement an alternative and the cost risk arising from an alternative not performing as expected; and
- Safety, including the operational safety of workers inherent in the different alternatives.
A list such as this could be a starting point for discussion, although its format is not useful for comparing management alternatives.
Ultimately, identifying a widely acceptable solution requires that each participant observe that his or her concerns are reflected in the group’s decision objectives. Each agency should be able to trace some subset of the group’s decision objectives developed within this process back to its own overall management objectives (i.e., such as those listed in Table 6.1). Similarly, nonagency participants should be able to do the same when reflecting back on their respective goals stated early in the process. From a process perspective, this provides a formal check that concerns are being heard and considered. The process requires consideration of each participant’s concerns and objectives; it does not assume adoption of each participant’s proposed solution.
The Objectives Hierarchy
The bulleted list of decision objectives above could be a starting point for discussion. As dialogue continues among the decision participants, however, a more refined and structured set of decision objectives should be developed. Decision objectives can be nested within an “objectives hierarchy” (Keeney, 1996), a method of organizing the objectives into a manageable format. An objectives hierarchy allows decision participants to better understand the relationships between specific goals. Participants create an objectives hierarchy by deciding which objectives represent the highest-level goals (i.e., minimize adverse impacts of erosion of the debris blockage) and which represent more detailed goals that are necessary to reach the highest-level objectives (e.g., minimize cost, minimize risk of a Spirit Lake breakout, minimize adverse impacts to downstream residents). Each of those subgoals can be further broken down into more subgoals. The relationship between objectives can be illustrated in a hierarchy such as the example in Figure 7.1. This committee developed this figure based
on individual opinions expressed during its public meetings in Kelso and on information published by the USACE: for example, see USACE, 1983 (section IV, 11-12). It represents only a small subset of high-level objectives and therefore should be considered rough and incomplete. The objectives in it, however, might be similar to those of a hierarchy developed for decisions related to control of water levels in Spirit Lake and the implications for erosion management.
The decision objectives are phrased and organized to clarify both the preferred direction of change (e.g., decreasing) and the subject of concern (e.g., cost risk). Tension may arise among participants because not all these objectives can be reached at their minimum or maximum levels, but the diagram could be used in later discussions about management alternatives
and their consequences to determine how well those consequences align with decision objectives. This should aid the identification of a preferred alternative. More specificity can be included in the objectives hierarchy to differentiate how alternatives align with the various objectives. It might be desirable among the decision participants to include goals related to different fish species or specific impacts to fish above and below the SRS. Doing so may help participants understand how different decision alternatives will require certain trade-offs among the objectives.
Developing an objectives hierarchy is an iterative process. The exact wording of the objectives in the hierarchy will evolve with greater understanding of the relationships between management alternatives, their consequences throughout the system, and the objectives that matter to people. It is important to avoid premature comparisons of decision objectives because priorities among objectives are likely to shift as management alternatives are better understood. It is more useful to focus on creating a complete and compact set of clearly worded objectives and sub-objectives. While the objectives need to be designed with an eye toward highlighting future trade-offs, the comparison of objectives will be more productive when trade-offs associated with choosing among decision alternatives and consequences are explored (discussed later).
A structured objectives hierarchy needs to be organized into categories relevant to the decision at hand and meaningful to participants. Each participant’s (and agency’s) decision objectives should appear somewhere in the hierarchy and should be presented so as to be meaningful to all participants. Doing so is an important trust-building measure. The objectives hierarchy developed by the small group may be different in some respects from what would emerge from a benefit-cost analysis or other agency planning processes. Reconciling the results of this proposed objectives hierarchy and the objectives hierarchies used by other agencies is discussed briefly at the end of this chapter. Box 7.1 describes how decision objectives might be established for managing water levels at Spirit Lake.
When Is a Set of Objectives Complete?
Identifying the right number of discrete objectives is challenging, and there are no exact rules for doing so. It is possible to consider too few or too many objectives. Each decision process will result in a different number of objectives, and the objectives and their number will evolve as information is gathered and viewpoints are explored. The completeness of an objectives hierarchy might be gauged by considering a preliminary list of decision alternatives to the decision problem. When comparing alternatives, if no new decision objectives are raised by the interested and affected parties, then the list may be adequate. If new decision objectives are revealed, then additional work is needed to include the newly raised objectives into the hierarchy.
Choosing Metrics to Predict Performance Against Objectives
To define objectives properly, there is a need to determine by what means those objectives would be measured. Performance metrics can be chosen to measure the degree to which objectives are expected to be attained. Establishing such metrics gives decision participants a means to quantify a desired outcome so that expected progress toward or away from those outcomes can be modeled. Measuring expected progress is challenging, however. Metrics used by physical scientists and engineers to assess, for example, projected infrastructure performance will not necessarily measure progress with respect to decision objectives. Keeney and Raiffa (1993) note that there are many types of metrics and care is needed to select the appropriate metrics for each objective. Consideration of what is measured, where, and over what time frame is partially subjective and so requires both technical and nontechnical considerations. This allows easily overlooked links to be revealed and tested against the management alternatives and their impacts. Moreover, a deeper and more concrete understanding of what the pursuit of an objective may lead to can stimulate the search for new and creative solutions. This underscores the importance of using an iterative and broadly collaborative process. Metrics developed in isolation by technical experts may miss key elements relevant to decision participants.
Metrics used in performance forecasts need to capture what is expected to happen as well as the uncertainty associated with those expectations. The latter is usually expressed as a probability distribution over the scale of the metric. For instance, given one option, models might forecast that 500 fish of a specific species might migrate past a physical barrier in a specific amount of time—this is the expectation. The range of fish passing, however, may be 300 to 700 and expressed as a probability distribution over the scale of the metric.
When beginning to develop metrics, a deliberative discussion with participants could consider different kinds of metrics and how they perform under different scenarios that may be of interest. These discussions are both technical in nature and driven by the concerns of decision participants. Metrics need to be developed with explicit consideration of how they do or do not incorporate risk and attitudes toward risk.
Some metrics directly measure the consequences of interest in their own terms. These are considered measurements of “natural attributes” in decision science literature (e.g., Keeney, 1996). Reliance on natural attributes versus other types of metrics is, to the extent possible, desirable. As used here, the term “natural attributes” should not be confused with what natural scientists might use for characterization; rather, it refers to attributes that can be measured using the same units as those with which the objective might be expressed. An obvious natural metric for the maintenance cost of an alternative, for example, is dollars because there is a direct link between the metric (dollars) and the underlying objective (minimizing maintenance cost). Even in this example, however, care is needed since apparently “obvious” natural attributes might actually include hidden assumptions about participant values and priorities (NRC, 1996).
Using natural attributes as metrics tends to separate causal links from participants’ values (Slovic, 2010). It is seldom the case, however, that all the important objectives have natural metrics. More typically, metrics are best stated as proxies (also known as indicators or correlates) for the consequences of an alternative, and still others might be represented on a constructed scale (see Box 7.2). For example, the area (e.g., acres or hectares) of accessible fish spawning habitat might be a useful proxy for
fish abundance since it is easier to measure area and that measurement is tied more directly to management alternatives. In another example, the probability of a catastrophic breakout of Spirit Lake might be a useful proxy for reducing various downstream impacts. This value might be more meaningful to participants and might be easier to measure while exploring management alternatives than to determine the myriad of downstream consequences that would arise from a breakout—even given that the downstream impacts may be what really matter to people in the area. Delibera-
tive discussion among participants regarding proxies is vital; it is necessary to determine whether proxy metrics meet the needs of those comparing the alternatives. Deliberative discussion also stimulates the search for new and creative solutions.
Some metrics, such as flood damage, might be monetized. Flood damage may be translated from inundation depth and the number of structures impacted to an aggregated dollar amount. It may be desirable during a later step in the decision process to convert other impacts into dollars.
This would be done with input from the interested and affected parties and to the extent that this translation is found to be useful for comparing alternatives (see the later discussion of trade-offs).
Finally, progress toward certain decision objectives may be best tracked via constructed scales. Topics that arose during the committee’s public meetings suggest that some people would compare management alternatives in ways that would be difficult to measure. For example, it is likely to be difficult to determine the extent to which management alternatives protect the “naturalness” of the Monument or the degree to which they may improve or promote collaboration among agencies.
Additional Considerations When Defining Metrics
Various issues that will likely require metrics were raised by interested and affected parties during the committee’s open session meetings. Some are described briefly below. Many more topics will probably be raised during future decision making.
Restoring “naturalness.” As already described, “naturalness” is defined differently by different interested and affected parties in the region. Participants will need to agree on a meaningful definition that is grounded in cause-effect logic. The definition will need to be expressed so as to be linked usefully to decision objectives and metrics. This might link naturalness to modeled or quantifiable estimates of fish and wildlife abundance, or it might develop a constructed scale.
Cost uncertainty. While expected costs have been used to characterize built management alternatives, cost uncertainty (also called cost risk) may also need to be addressed. A wide range of beliefs has been aired related to the future performance of the Spirit Lake tunnel (and therefore the need for and size of repairs) versus an open channel alternative (see Chapter 5). Decision participants may find that cost risk is an issue because of institutional barriers—sudden and unexpected repairs stress the USFS budgeting process. Deeper exploration of metrics might
yield better understanding of the consequences of management alternatives, which, in turn, could inspire more imaginative management approaches to more directly address decision objectives.
Types of cost. Metrics for expected costs may require discussion among decision participants. Some interested and affected parties may place more importance on aggregate costs, while others may think in terms of present value costs. Because different agencies have different budgets allocated for different parts of the system (or different budgets for different mitigation drivers in the same part of the system), division of costs of various management alternatives among parties (and how they are tied to the intended mitigation) might also be of interest.
Adaptive choice. Metrics based on expected costs may be misleading if conditional choices are to be made over time. For example, it might be desirable to compare today’s cost of raising the SRS spillway to the cost of a more adaptive approach such as dredging as needed until sediment loads reach a certain threshold before the spillway is raised. Comparing the expected cost of dredging versus a spillway raise without the application of appropriate metrics that capture the expected cost of the dynamic dredging options—using, perhaps, a decision tree, for example—will not capture alternative values of the dredging strategy and would also ignore the regret value of raising the SRS in advance of need.
Fish abundance. Objectives related to fish survival and recovery were repeatedly raised by interested and affected parties. These include the impacts of dredging on fish migration; potential for upstream migration past the SRS, through the sediment plain, and even into Spirit Lake; access to spawning areas (including tributaries that may or may not be cut off from the main stem through sedimentation); and possibilities for downstream migration through the sediment field. It is not apparent, however, that a system-wide assessment of the fish populations and the factors that affect them has been performed, let alone
been translated into metrics useful for comparing alternatives. In a similar approach to that presented in Table 7.1, a first step would be to identify factors that impact fish survival and recovery on a system-wide basis and assemble them to compare management alternatives. This would require deeper deliberation to develop common understanding regarding the different dimensions of “fish abundance” (i.e., different individuals might equate population size, population productivity, or aggregate health of individual fish in a population as defining characteristics of fish abundance). If trade-offs are complex because there are different impacts in different locations for multiple management alternatives (e.g., raising the SRS spillway avoids dredging and therefore the associated negative impacts on fish downstream, but it blocks passage and inundates side channels above the SRS), then further work is needed to aggregate the impacts into one summary statistic that
TABLE 7.1 Screening Table for Alternatives and Their Associated Uncertainties.
|Closed Conduit (Alternatives 1-4)||Open Channel (Alternative 5)|
|Known engineering design and performance||Yes||No|
|Potential for mechanical failure||Yes||No|
|Outflow scales with inflow||No||Yes|
|Vulnerability to principal regional hazards|
|Geomorphic||None||Moderate to High|
|Time scales of recession post-hydrologic event||Weeks to months||Days|
|Time scale for intervention in the event of failure||Weeks to months||Hours to Days|
aThere are subtleties associated with a buried conduit that increase vulnerability to these hazards.
SOURCE: G. Grant, personal communication (November 1, 2016).
allows a comparison of management alternatives based on their overall system-wide impact on fish. This is likely to be a highly technical exercise (e.g., comparing the importance on overall fish abundance of spawning, rearing, and migration life stages). But this is also likely to be a value-laden exercise (e.g., determining which fish matter more and where). Care will be needed to structure the interaction between the technical analysis and the decision participants.
Wildlife habitat. Habitat above the SRS was frequently mentioned but rarely quantified. There are several countervailing trends, including loss of habitat due to the steady accumulation of sediment, the natural re-vegetation of areas impacts by the events of 1980, and the possible revegetation of portions of the sediment plain. Furthermore, habitats suitable for different categories of species (e.g., large mammals, birds, rodents) can be differently affected by decisions. Useful metrics would be species category–specific and would isolate the effects of management alternatives as opposed to natural trends.
The third step in a PrOACT-like process addresses alternatives. The goal is to craft multiple and diverse alternatives to address the management decisions at hand. This section will look at how this can be accomplished and describe how future management efforts might build on previous efforts. Management alternatives should be complete packages of actions or policies (i.e., a full system-wide strategy) put together in a way that addresses all the possible individual actions and policies available to decision participants within the scope of decisions being addressed.
A common mistake in decision making is to construct alternatives from a too narrow subset of possible actions, thus failing to represent a full range of possible solutions (Bond et al., 2008). One effect of constraining alternatives in this way is the tendency to miss interdependencies. This mistake is reflected in past decision documents in the Spirit Lake and Toutle River region. For example, as mentioned earlier, the original USACE plan (1983)
included alternatives to manage Spirit Lake levels as well as alternatives to manage sedimentation in the system. While it was recognized that these two management levers could be interdependent (USACE, 1984a), analyses presented in the documents treated them as independent; each set of alternatives was treated separately. The opportunity to formulate and evaluate management alternatives recognizing interdependence was missed. It is often the tendency of individual agencies to concentrate on actions that are possible within their respective authorities whereas representatives of the Cowlitz Indian Tribe, the community residents, and other interested and affected parties may be interested in a more holistic management of the challenges that affect the entire system, as expressed during the committee’s open session meetings in Kelso.
Possible actions and policies should be bundled into system-wide alternatives so that their full impact and interrelationships can be assessed. Such an approach will capture the important impacts of, for example,
- Spirit Lake water level management alternatives on the risk of catastrophic breakout through the debris field;
- Spirit Lake water level management alternatives on sedimentation rates at and below the SRS;
- Spirit Lake, SRS, and dredging alternatives on fish survival and recovery; and
- Spirit Lake and SRS alternatives on wildlife from Spirit Lake to the Cowlitz River.
This list captures some issues that are apparently secondary to the primary issue of preventing a catastrophic breakout of Spirit Lake. It might be argued that the prevention of a Spirit Lake breakout is so overwhelmingly important as to make connections to secondary issues irrelevant. Even so, it is helpful to include these secondary issues to avoid derailing decision processes when participants feel that their concerns are being ignored and potential solutions to address them are being dismissed. Inclusion of apparently peripheral issues creates a means for agencies to engage with interested and affected parties on the “primary” issues being addressed.
Broad inclusion of interested and affected parties when creating decision alternatives potentially improves decision quality (NRC, 2008), avoids an overly narrow scoping of alternatives (which can preclude buy-in from a broad audience), and possibly contains and resolves debate over the feasibility of alternatives so that outcomes are mutually satisfactory.
It is useful at this stage of the decision process to be deliberate about why various alternatives are being considered. Broad and creative thinking imposes a process burden (e.g., time and other resources) on the decision lead and participants, and pushback can be expected if alternatives seem unrealistic. Nevertheless, systematically and inclusively identifying alternatives can, at different phases in the decision process,
- Build trust among participants;
- Inform all engaged about modeling processes and system behavior given different management alternatives;
- Help participants better understand the relationship between participants’ decision objectives and what is possible to change through management; and
- Drive participants toward a solution.
Just as it is important to collaboratively identify decision objectives and their metrics, it is important to develop the list of alternatives through a collaborative process for many of the same reasons. A level of engagement consistent with the “involve” or “collaborate” levels of participation explicitly includes a commitment to incorporate other participants’ ideas in the generation of alternatives.
Generating Common Understanding About Alternatives
The need for common understanding of the decision problem and objectives extends into the next phase of the decision-making process. For example, the expressed desire for a “more natural” way to manage Spirit Lake water levels may be at odds with potential engineering solutions for the control of water in Spirit Lake as described in Chapter 5. Some par-
ticipants at the committee’s open session meetings discussed their understanding of the benefits of the open channel alternative to control Spirit Lake water levels. There seems to be a divergence between what advocates believe a functional, reliable permanent open channel solution would look like (i.e., a “natural” part of the landscape) and the possible reality of a less maintenance-intensive and more robust solution with predictable performance (i.e., an engineered spillway). There is a substantive knowledge gap regarding practical design issues that needs to be resolved before trade-offs can be considered between the open channel alternative and other alternatives. A concerted effort is required to create common understanding and agreement about performance characteristics and design goals (e.g., a more natural landscape). There may even be trade-offs to consider among high-level goals. This may require a smaller deliberative process to design one (or several) open drainage alternative(s) that then informs system-wide considerations.
A Broad Range of System-Wide Alternatives
A system-wide approach for making decisions is described and recommended in this report. This involves thinking about how all the natural and built elements of the Spirit Lake and Toutle River region might contribute to or be affected by any given alternative. Different kinds and possibly combinations of coordinated management actions and policies may result in the most realistic, cost-effective, and sustainable solutions. When considering any decision in the system, it is important to consider all components of a system-wide solution and to generate alternatives that include
- capital works (multiple and perhaps redundant ways to manage water levels at Spirit Lake);
- operational changes (including lowering Spirit Lake, different approaches to dredging);
- emergency response plans (including a road to ensure full-time access to drainage works, automating gate operations, enhancing downstream emergency management systems); and
- mitigation measures (e.g., levees, managing dredge spoils, property buyouts).
No single agency can shoulder the responsibility for all the management actions required, so explicit consideration of what agency or organization will be responsible for what actions needs to be made.
Alternatives Over Time
A systems approach also requires understanding system response over time. When generating alternatives, consider whether actions are necessary only once or are more dynamic and adaptive and based on a sequence of conditional decisions. When the SRS was first planned, the USACE chose a staged approach to future management (USACE, 1983). A sequence of spillway raises was not originally part of the plan, but that possibility became feasible when the probable maximum flood was downscaled in the early 2000s. Such an adaptive approach (given that the existing plan is truly adaptive and responsive to conditions) should also be considered among the alternatives for current and future spillway raise decisions at the SRS, where low-cost dredging alternatives might be explored before undertaking the next spillway raise—given the cost and likely irreversibility of the decision. Note, however, that these alternatives have implications across the system and across steps in the decision process. Where there are dynamic, adaptive management steps that can be taken, taking the expected value of the criteria must account for dynamic choices. Ignoring this modeling requirement means that the expected values will miss alternative value (the benefit of delaying a decision) and regret value (the lost benefit from making a decision too soon) and potentially lead to the wrong policy conclusions.
The Number of Alternatives
Taking a systems approach to management requires developing strategies that consist of one or more actions, addressing multiple goals, and consid-
ering all beneficial and adverse effects, both individually and in aggregate. Whereas thinking creatively about all aspects of the system at once can lead to creative solutions, it may also lead to an unmanageable proliferation of alternatives. There are constructive ways to limit this process, and collaborative deliberation guided by a decision analyst can help to negotiate these. Examples are provided below.
- A collaborative participatory process may, in a purposeful manner, screen those alternatives that are obvious nonstarters for participants. The fatal flaws associated with those alternatives need to be carefully communicated so the reasoning behind their removal is understood.1 A mistake here is to prematurely eliminate alternatives best dealt with during later deliberations of trade-offs.
- Deliberation might reveal those specific actions or policies that could be decided on independently of other choices. These can then be added back into discussions as constants across system-wide alternatives.
- Strategy tables (Gregory et al., 2012), such as the hypothetical example shown as Table 7.2, are visualization tools that can be used to effectively manage the number of possible solution sets when there are a large number of decision elements.
- A careful review of decision objectives can be useful in guiding the generation of alternatives.
There is a tendency for alternatives to be adjusted from some base case or starting point. This may preclude consideration of some viable alternatives that are substantially different from that starting point. In the behavioral decision literature, this is called the anchor-and-adjustment bias (Kahneman and Tversky, 1982). On the other hand, alternative creation can stall when it becomes a mechanical working through of all possible permutations and combinations of decision elements. Using decision ob-
TABLE 7.2 Hypothetical Example of a Strategy Tablea
|Spirit Lake Levels||Other||Spirit Lake Drainage Alternatives||SRS Alternatives||Dredging Alternatives||Other|
|Maintain current operating regime
Reduce max height to XX
Reduce max height to YY
Permanently drain Spirit Lake
|Build access road to intake
Clear logs from lake
Reinforce log boom by intake
Alternate drainage tunnel
Pumping station on standby
Lower spillway by 7’
Maintain current spillway elevation
Raise spillway by 13”
Remove upstream migration impediment for salmon
|Dredge in Toutle as needed to maintain current flood protection
Dredge in Columbia as needed to maintain shipping
|Raise levees at XX, YY
Do ZZ with emergency management system
NOTES: Blue squares represent a primary set of alternatives associated with one possible strategy; red squares represent the alternatives associated with a second strategy. Contents of table represent examples and not recommendations.
jectives to drive the creation of alternatives can help navigate between these two extremes. Referring back to the examples of draft decision objectives described in the previous section, this process could consider
- A solution that minimizes the risk of a catastrophic breakout from Spirit Lake;
- A solution that minimizes overall cost;
- A solution that maximizes the “naturalness” of the system (if this decision objective is carried forward by decision participants);
- A solution that minimizes the negative impact of sediments; and, perhaps,
- A solution that maximizes the abundance and diversity of fish and wildlife populations.
While this set of alternatives may not contain the solution settled on ultimately by decision participants, these might serve as preliminary bounds for an otherwise almost limitless decision space, constraining the search for the best solution and perhaps yielding surprising insights that can be built upon to create a final, broadly acceptable solution. Box 7.3 provides a list of questions that decision participants should consider as a group when developing their list of alternatives.
In a strategy table (e.g., Table 7.2), one alternative might address the desire for a more “natural” solution by selecting an open channel and a lower SRS spillway level. This alternative is a package of all these alternatives—that is, the various actions required throughout the system to implement a specific strategy—selected at once, as shown by the blue solid circles above. A second strategy (highlighted with red boxes) could be minimizing sediment migration. In this case, Spirit Lake is drained to a lower level, an alternate drainage tunnel is chosen that does not discharge into the Toutle River system, and current dredging practices are continued in the lower river. Whether or not this idea is acceptable depends on how this package of alternatives, as modeled for the combined impacts across the whole system, compares to the performance of other packages of alternatives given the chosen decision objectives and metrics.
A strategy table is a useful tool for creating a mental model to compare the individual actions required to implement a given strategy with the actions embodied in other strategies. The table provides a visual image that may be helpful for narrowing strategies. A strategy table applied in a real-world setting would probably include more elements than appear in Table 7.2 and, as such, is likely to be more practical once the number of alternate strategies to be considered has been reduced.