Methodology
Information on this page describes the approach used on this website to provide information on the following topics:
- Climate Observations and Projections
- Projecting Population
- Water Quality Tool
- Low Flow Calculator
- Projecting Sea-Level Rise
- Mapping Sea-Level Rise
- Projecting Land Use Patterns
OVERVIEW
Data
The CARA website allows users to view many types of data at scales ranging from the entire mid- and upper-Atlantic region down to the county level (and smaller areas for case studies) and the ability to combine data layers. This can make it easier to understand interactions among different influences.
An emphasis in compiling the data has been to use publicly available data for which documentation exists to ascertain its quality. A similar emphasis in presenting the data has been to rely on software that is free to users. The CARA research team has a strong commitment to peer review so that the analysis of data and tools will be scientifically credible—and thus more useful to decision makers.
Accounting for Uncertainty
The more complete our information about a future outcome, the more certain, or confident, we are about that outcome. When we have complete information about all variables - like the roll of a die - we can calculate the likelihood of an outcome, and be confident about it. When not enough information is available to calculate a probability - a measure of how likely it is that some event will occur - sometimes we can determine a range for the likelihood. Despite gaps in information we may still feel relatively confident that the likelihood will lie within this range.
Nearly all decisions involve uncertainties, but are easier to make if there is information about the degree of uncertainty. Decisions affected by potential changes in climate or land use have uncertainties because it is impossible to be absolutely sure about events and conditions in the future. We rely on historical data, theoretical understanding, model projections, and published scientific literature to judge the likelihood of future outcomes. We have more confidence that a particular outcome is “very likely” (or “very unlikely”) when numerous experts agree on the supporting evidence.
Good decisions rely on sound information for identifying potential future events and determining their likelihood. If there is a lot of uncertainty about how a particular factor might affect a decision that could produce undesirable outcomes, then it is valuable to conduct activities to close the information gaps and reduce uncertainty. This process still can leave some uncertainty and we need to be clear about the degree of confidence for the remaining uncertainty.
Misunderstandings about probability and uncertainty often can be avoided by using a shared language about the “considered judgment” of likelihood that is based on whatever supporting evidence is available. The CARA team and website rely on the language depicted in the graphic shown below:
( Used for the first US National Assessment
)
An outcome judged to be nearly certain (having a probability near one) is described as being “very likely” or “very probable.” An outcome whose probability is large (considerably more than 50%) but not near one would be described as “likely” or “probable.” When the probability is judged to be in the mid range, the outcome would be described as “possible,” and so on. The “pillows” convey that boundaries between the descriptors are fuzzy, rather than being tied to specific probabilities.