Skip to main content



Climate change is here, it’s real, and it won’t be easy for humans to deal with. But few things are all good or all bad, and so it may be for climate change, at least with respect to environmental science and management.

A vast literature has accumulated in the past two or three decades in geosciences, environmental sciences, and ecology acknowledging the pervasive—and to some extent irreducible—roles of uncertainty and contingency. This does not make prediction impossible or unfeasible, but does change the context of prediction. We are obliged to not only acknowledge uncertainty, but also to frame prediction in terms of ranges or envelopes of probabilities and possibilities rather than single predicted outcomes. Think of hurricane track forecasts, which acknowledge a range of possible pathways, and that the uncertainty increases into the future.

Forecast track for Hurricane Lili, September 30, 2002. The range of possible tracks and the increasing uncertainty over time are clear. Source: National Hurricane Center.

Applied environmental science and environmental management has traditionally been based on what I think of as a medical-style model—if a patient is presenting certain symptoms, then this is the treatment. The environmental analogy, be it forest or range management, agricultural policy, stream restoration, etc., has been similar—if this is the problem or the situation, then that is what you do. Sometimes this works; sometimes it doesn’t.

An approach more consistent with the way environmental systems actually work is based on flexibility, adaptability, and identifying ranges of options. This is the case with or without climate change, but perhaps dealing with the uncertainties, path dependence, complex interconnections, and inevitable surprises we know are associated with ongoing and near-future climate change, we can change the culture of environmental and natural resource policy and management. Rather than (at least implicitly) pretending certainty in forecasts, we acknowledge uncertainty, contingency, and possible surprises. Rather than single prescribed management options, we develop and present ranges of options, with multiple objectives and multiple potential pathways. And rather than implementing a policy, restoration scheme, etc., and “riding it until it crashes,” we build in flexibility and adaptive capacity.