Uncertainty is a concept about value. When we want to know something we desire to know how accurate the estimate is or how close it is to truth. Climate change scientists and adaptation experts recognize many sources of uncertainty associated with climate change and adaptation. Much has been written about sources of uncertainty, how to quantify and express them, and how to accommodate them in responding to climate change. Two major sources of uncertainty are the likelihood of an outcome (i.e., properties of the system being studied) and the confidence in the information used. Below we provide a few key references for understanding uncertainties and for responding to them in adaptation strategies. Broadly speaking, climate uncertainties are associated with: 1) projections of future climate change (especially for longer time horizons), 2) evaluations of impacts of this change on species and ecosystems, and 3) the effectiveness of strategies designed to abate and mitigate these impacts.

The Intergovernmental Panel on Climate Change (IPCC 2007) defines uncertainty as "An expression of the degree to which a value (e.g., the future state of the climate system) is unknown. Uncertainty can result from lack of information or from disagreement about what is known or even knowable (i.e., among experts). It may have many types of sources, from quantifiable errors in the data to ambiguously defined concepts or terminology, or uncertain projections of human behavior. Uncertainty can therefore be represented by quantitative measures (e.g., range of values calculated by various models) and/or by qualitative statements (e.g., reflecting the judgment of a team of experts)."

In 2013, the IPCC will release its fifth major assessment of climate change. This assessment will speak to uncertainties in projections of future climate change. And, despite substantial improvements in scientists’ understanding of climate change, the models used in this assessment are likely to produce even greater ranges of uncertainty in their predictions than past efforts (Maslin and Austin 2012). There are several reasons for this. First, models are not reality but are mathematical expressions that attempt to approximate reality by incorporating as many factors as possible that are involved in complicated climate systems to simulate the Earth’s climate. However, they cannot possibly accurately incorporate all mechanisms because of an incomplete understanding of factors that affect climate and create climatic patterns. Second, the many different global climate models each have somewhat different assumptions (e.g., how clouds will behave as temperatures rise) and perform differently when tested against historical climate data. Third, the spatial resolution of climate models vary but are improving as more detailed models are being used to predict regional climate patterns. Nonetheless, these regional climate models have additional large uncertainties similar to those associated with the models used to predict local weather. For example, precipitation is highly variable over relatively small spatial scales. Finally, various forms of uncertainty all add up. For example, climate models take as inputs estimates of the amount of greenhouse gasses in the atmosphere – but this, in turn, relies on assumptions and models of population growth, emissions regulation policies, and economics—all with their own obvious and less obvious uncertainties. Although models used in the IPCC’s fifth assessment will make fewer assumptions and account for more complex climate factors, there is still a great deal of uncertainty surrounding these complex factors.

The outputs of global climate models are used, in turn, as inputs to models that project the impact of climate change on species, communities, and ecosystems. These ecological models have their own sources of uncertainty (Glick et al. 2011). Among them are limited or unreliable data on the natural history of species and ecosystems and their sensitivities to climate change (including basic information on species distributions), unknown interactions of these species and systems with non-climate stressors (e.g., invasive species, human –related habitat fragmentation), uncertain effects of human responses to climate change (Turner et al. 2010), and scientific disagreements on what we know about species and ecosystems and their responses to climate. In reality, these sources of uncertainty parallel those that physical and meteorological scientists face in predicting future climate changes.

We outlined six broad adaptation objectives in the Yale Framework. Implementing these objectives involves a set of uncertainties. In some cases, these uncertainties will already be accounted for by those related to predicting future climate and estimating climate impacts on species and ecosystems. For others, there will be new uncertainties and assumptions introduced. For example, "protecting the ecological stage" assumes that most species distributions are closely correlated and tied to underlying abiotic or geophysical features, when in fact there are many species for which we do not know if this is the case. Furthermore, different analytical approaches can produce radically different sets of facets or units. cancan produce In the case of the "ecological process" adaptation objective, scientists know that many ecological processes are important to the persistence of species and that these processes are being altered by climate change. However, scientists also recognize that much is unknown about how species respond to changes in ecological process. These examples illustrate the many uncertainties associated with implementing broad adaptation objectives.

We now understand there are multiple sources of uncertainty in assessing climate change, its impacts, and moving forward with adaptation responses. In some cases, uncertainties interact, usually resulting in an increase in overall uncertainty. Understanding interactions can be accomplished qualitatively through the use of conceptual models or expert judgment or quantitatively through models and decision support systems (Glick et al. 2011).

How to move forward in the face of uncertainty? Although it would be easy for a natural resource manager to be paralyzed by uncertainty, there is also a considerable amount of good advice and recommendations for moving forward with adaptation strategies amidst this uncertainty. First many proposed broad adaptation objectives are relatively robust to uncertainties associated with climate change – that is good actions regardless of whether projected changes in climate play out as we think they will (Groves et al. 2012). Second, there are a number of techniques to help reduce uncertainties such as using simulation analyses that account for uncertainties, sensitivity analyses that explore how robust certain models or adaptation strategies are to various assumptions, and scenario analyses that examine a range of possible outcomes of either impact projections or results of implementing adaptation objectives (Glick et al. 2011). Finally, employing an adaptive management approach in which adaptation management objectives are evaluated through monitoring and evaluation (Lawler et al. 2010, Cross et al. 2012) is critical to providing managers the feedback necessary to make continual changes to address new information and uncertainty.