APPENDIX D
Uncertainty Terminology
In assessing the state of knowledge about climate change, scientists have developed a careful terminology for expressing uncertainties around both statements of fact about a current situation (for example, “most observed warming can be attributed to human action”) and statements about the likelihoods of various future outcomes (for example, “sea level could rise by several feet by 2100”). The IPCC, in particular, has devoted serious debate and discussion to appropriate ways of expressing and dealing with uncertainty around such statements (Moss and Schneider, 2000), and all recent IPCC assessments have adopted a set of terminology to describe the degree of confidence in conclusions (see, e.g., Manning et al., 2004). In estimating confidence, scientific assessment teams draw on information about “the strength and consistency of the observed evidence, the range and consistency of model projections, the reliability of particular models as tested by various methods, and, most importantly, the body of work addressed in earlier synthesis and assessment reports” (USGCRP, 2009). It is easier to employ precise uncertainty language in situations where conclusions are based on extensive quantitative data or models than in areas where data are less extensive, important research is qualitative, or models are in an earlier stage of development. Statements about the future are also generally more uncertain than statements of fact about observed changes or current trends.
Table D.1 shows the language adopted by the IPCC to describe confidence about facts or the likelihood that a statement is accurate. The U.S. Global Change Research Program’s recent assessment report on Global Climate Change Impacts on the United States(USGCRP, 2009) uses similar language. In this report, Advancing the Science of Climate Change, when we draw directly on the statements of the formal national and international assessments, we adopt their terminology to describe uncertainty. However, because of the more concise nature and intent of this report, we do not attempt to quantify confidence and certainty about every statement of the science.
TABLE D.1 Language Adopted by the IPCC to Describe Confidence About Facts or the Likelihood of an Outcome
Terminology for Describing Confidence About Facts |
|
Very high confidence |
At least 9 out of 10 chance of being correct |
High confidence |
About 8 out of 10 chance |
Medium confidence |
About 5 out of 10 chance |
Low confidence |
About 2 out of 10 chance |
Very low confidence |
Less than 1 out of 10 chance |
Terminology for Describing Likelihood of an Outcome |
|
Virtually certain |
More than 99 chances out of 100 |
Extremely likely |
More than 95 chances out of 100 |
Very likely |
More than 90 chances out of 100 |
Likely |
More than 65 chances out of 100 |
More likely than not |
More than 50 chances out of 100 |
About as likely as not |
Between 33 and 66 chances out of 100 |
Unlikely |
Less than 33 chances out of 100 |
Very unlikely |
Less than 10 chances out of 100 |
Extremely unlikely |
Less than 5 chances out of 100 |
Exceptionally unlikely |
Less than 1 chance out of 100 |
SOURCE: IPCC (2007a). |