The study presents best estimates for climate change and its impacts at varying warming levels and associated stabilization targets, and it also briefly describes the level of understanding of the processes involved (e.g., physical, biological, etc.) in such changes. Presentation of best estimates provides a clear view of the best current understanding that may otherwise be poorly communicated, and this is one aim of this study. However, we also balance descriptions of best estimates with their corresponding uncertainties, along with appropriate coverage of issues of uncertain risks. Climate changes and their impacts are discussed on a global basis, and specific regional examples within America and American territories are also presented for the purpose of illustration. We summarize as appropriate the following factors: (1) the extent to which multiple studies are available, and the robustness of findings across work by a range of authors, (2) the scientific confidence in understanding of the underlying processes, and (3) studies that already attribute a contribution of observed changes to anthropogenic effects where available (see Section 1.2). Where attribution is already possible for current levels of climate change, confidence in further future changes is generally strengthened.

Many climate changes or impacts currently are understood only in a qualitative manner, and thus are not quantifiable as a function of stabilization target. The report assesses and identifies these unquantified factors to the extent practical based upon the available literature. It should be emphasized that these should not be considered negligible; indeed some of these could be very important, or even dominant, in evaluating future risks due to anthropogenic climate change.

Many studies involve the use of “pattern scaling” whereby it is assumed that the spatial pattern of future climate change computed for one level of perturbation (i.e., radiative forcing) may be scaled to derive values for another test case such as one with stronger forcing and larger perturbations. A variety of studies have shown that such methods generally simulate the results of atmosphere-ocean general circulation models rather well (see Section 4.2), although results are generally less robust for precipitation than for warming (see Section 4.3), and near regions where strong feedbacks such as sea ice retreat take place. We employ pattern scaling for many of the estimates in this report.

Earth’s history shows that climate changes can occur on time scales ranging from decades to centuries to millennia. All of these time scales are considered here. As there is abundant evidence that climate is already changing in part due to human activities, a focus of the report is on the next few decades to century, where the climate changes under increased



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement