Editing IPCC:AR6/SR15/Chapter-1 (section)

== 1.6 Confidence, Uncertainty and Risk ==

<div id="article-1-6-block-1"></div>

This report relies on the IPCC’s uncertainty guidance provided in Mastrandrea et al. (2011) <sup>[[#fn:r297|297]]</sup> and sources given therein. Two metrics for qualifying key findings are used:

'''Confidence:''' Five qualifiers are used to express levels of confidence in key findings, ranging from ''very low'' , through ''low'' , ''medium'' , ''high'' , to ''very high'' . The assessment of confidence involves at least two dimensions, one being the type, quality, amount or internal consistency of individual lines of evidence, and the second being the level of agreement between different lines of evidence. Very high confidence findings must either be supported by a high level of agreement across multiple lines of mutually independent and individually robust lines of evidence or, if only a single line of evidence is available, by a very high level of understanding underlying that evidence. Findings of low or very low confidence are presented only if they address a topic of major concern.

'''Likelihood:''' A calibrated language scale is used to communicate assessed probabilities of outcomes, ranging from ''exceptionally unlikely'' (&lt;1%), ''extremely unlikely'' (&lt;5%), ''very unlikely'' (&lt;10%), ''unlikely'' (&lt;33%), ''about as likely as not'' (33–66%), ''likely'' (&gt;66%), ''very likely'' (&gt;90%), ''extremely likely'' (&gt;95%) to ''virtually certain'' (&gt;99%). These terms are normally only applied to findings associated with high or very high confidence. Frequency of occurrence within a model ensemble does not correspond to actual assessed probability of outcome unless the ensemble is judged to capture and represent the full range of relevant uncertainties.

Three specific challenges arise in the treatment of uncertainty and risk in this report. First, the current state of the scientific literature on 1.5°C means that findings based on multiple lines of robust evidence for which quantitative probabilistic results can be expressed may be few in number, and those that do exist may not be the most policy-relevant. Hence many key findings are expressed using confidence qualifiers alone.

Second, many of the most important findings of this report are conditional because they refer to ambitious mitigation scenarios, potentially involving large-scale technological or societal transformation. Conditional probabilities often depend strongly on how conditions are specified, such as whether temperature goals are met through early emission reductions, reliance on negative emissions, or through a low climate response. Whether a certain risk is considered high at 1.5°C may therefore depend strongly on how 1.5°C is specified, whereas a statement that a certain risk may be substantially higher at 2°C relative to 1.5°C may be much more robust.

Third, achieving ambitious mitigation goals will require active, goal-directed efforts aiming explicitly for specific outcomes and incorporating new information as it becomes available (Otto et al., 2015) <sup>[[#fn:r298|298]]</sup> . This shifts the focus of uncertainty from the climate outcome itself to the level of mitigation effort that may be required to achieve it. Probabilistic statements about human decisions are always problematic, but in the context of robust decision-making, many near-term policies that are needed to keep open the option of limiting warming to 1.5°C may be the same, regardless of the actual probability that the goal will be met (Knutti et al., 2015) <sup>[[#fn:r299|299]]</sup> .

<span id="storyline-of-the-report"></span>