barrier to using models relates to relevancy issues, which are resolved when
the conceptual models are adequately developed. The second barrier to
model usage is confidence that the model will provide reliable results that
are sufficient for decision making; this criterion relates to a host of specific
details in model development, validation, and application. Each of these
details has to be addressed satisfactorily in order to overcome those
barriers. But that in itself is inadequate, because experiences by decision
makers, nonmodelers, and stakeholders involving the use of other models
for other environmental assessments or resource management decisions
may greatly influence and bias their perspective of the barriers for the
assessment at hand. Barriers are diminished the more that models are
successfully used and are perceived as having significantly improved the
decision-making process by making more correct, timely, cost-effective, or
defensible decisions. Conversely, the more models are used unsuccessfully,
or at least perceived as so used, the more barriers are raised for future
model use. For example, models that lead to the wrong decision, to a decision that does not hold up to adversarial scrutiny, to a decision that is too
costly, or to a decision that addresses the wrong problem decrease confidence in the use of models in general. This vulnerability is not limited to
models; the same result occurs for other technological or scientific bases for
uncertain or bad decisions, ranging from decisions about the swine flu to
decisions about having pure oxygen atmospheres for Apollo capsules. But
models are especially at risk. By definition, they are approximations of
reality and, therefore, have inherent uncertainty, which in some circumstances can be substantial. Also, models are often presented as magical tools
yet are too obscure for the decision maker to understand, thus requiring
almost blind faith in their usage.
We are at a critical juncture in the use of models to solve environmental
problems. Unlike the historic regulatory issues of the 1970s and 1980s that
focused on local scales and point-source controls, environmental managers
and decision makers must be prepared to make decisions at the watershed
and regional scales and for periods that encompass multigenerational time
scales. The economic and societal consequences of decisions at these scales
are tremendous. To address this emerging class of environmental problems
effectively, scientists and decision makers must now begin to rely on not a
single model but a suite of models (e.g., climate, hydrologic, hydrodynamic,
ecotoxicological, population, community, and ecosystem) to provide the
necessary complexity to evaluate management options at these scales. By
directing careful attention to the issues discussed above, especially the
issues relating to confidence-building and to the communication of models,
and taking particular care to select the most appropriate and defensible
models in the first place, modelers and scientists can make great strides in
overcoming the barriers to understanding and using models in environmental decision making.
5. Overcoming Barriers to the Use of Models
105
the conceptual models are adequately developed. The second barrier to
model usage is confidence that the model will provide reliable results that
are sufficient for decision making; this criterion relates to a host of specific
details in model development, validation, and application. Each of these
details has to be addressed satisfactorily in order to overcome those
barriers. But that in itself is inadequate, because experiences by decision
makers, nonmodelers, and stakeholders involving the use of other models
for other environmental assessments or resource management decisions
may greatly influence and bias their perspective of the barriers for the
assessment at hand. Barriers are diminished the more that models are
successfully used and are perceived as having significantly improved the
decision-making process by making more correct, timely, cost-effective, or
defensible decisions. Conversely, the more models are used unsuccessfully,
or at least perceived as so used, the more barriers are raised for future
model use. For example, models that lead to the wrong decision, to a decision that does not hold up to adversarial scrutiny, to a decision that is too
costly, or to a decision that addresses the wrong problem decrease confidence in the use of models in general. This vulnerability is not limited to
models; the same result occurs for other technological or scientific bases for
uncertain or bad decisions, ranging from decisions about the swine flu to
decisions about having pure oxygen atmospheres for Apollo capsules. But
models are especially at risk. By definition, they are approximations of
reality and, therefore, have inherent uncertainty, which in some circumstances can be substantial. Also, models are often presented as magical tools
yet are too obscure for the decision maker to understand, thus requiring
almost blind faith in their usage.
We are at a critical juncture in the use of models to solve environmental
problems. Unlike the historic regulatory issues of the 1970s and 1980s that
focused on local scales and point-source controls, environmental managers
and decision makers must be prepared to make decisions at the watershed
and regional scales and for periods that encompass multigenerational time
scales. The economic and societal consequences of decisions at these scales
are tremendous. To address this emerging class of environmental problems
effectively, scientists and decision makers must now begin to rely on not a
single model but a suite of models (e.g., climate, hydrologic, hydrodynamic,
ecotoxicological, population, community, and ecosystem) to provide the
necessary complexity to evaluate management options at these scales. By
directing careful attention to the issues discussed above, especially the
issues relating to confidence-building and to the communication of models,
and taking particular care to select the most appropriate and defensible
models in the first place, modelers and scientists can make great strides in
overcoming the barriers to understanding and using models in environmental decision making.
5. Overcoming Barriers to the Use of Models
105