The Third National Climate Assessment was released on Tuesday. The headline messages were not actually news: the climate is changing, average temperatures are increasing, sea levels are rising, humans are primarily responsible for changes over the last 50 years, and all of these trends are projected to continue. What happens in the next 20-30 years is already largely determined by warming that is “in the pipeline,” and what happens by 2100 is determined more by what we collectively choose to do to restrain emissions than by some uncertainty about the exact impacts of rising concentrations of heat-trapping gases. All of these conclusions were expressed with “very high confidence,” in part because they are based on continuing confirmation of conclusions drawn in previous assessments, both the U.S. National Assessments and the assessments of the Intergovernmental Panel on Climate Change (IPCC).
Confidence and uncertainty in the Third National Climate Assessment
Nonetheless, one can read the same document and see areas where scientists have concluded that uncertainty remains very great. For instance, our confidence remains low that the frequency of many types of extreme events – like hurricanes and tornadoes – are increasing due to human caused climate change. Our confidence in projections of sea level rise is only high when the range is given as “between 1-4 feet.” Similarly scientists understanding of the ecological consequences of ocean acidification are only “low –to-medium.”
The question that must be asked, then, is what actions should follow from our confidence in some judgments and our remaining uncertainty about others? And on this, the answer seems clear – we know more than enough already to warrant taking action to mitigate climate risks. We are already seeing harmful impacts to people, property and ecosystems. There are many reasons why it is hard to be certain about particular trends or specific projections, but we must always consider that reality may turn out to be on the bad end of things. And as has been shown in two recent papers (located here and here), under standard assumptions about risk and damage, greater uncertainty justifies greater, rather than lesser, precaution.
“Traceable Accounts” provide transparency for judgments of confidence
To its credit, the NCA has — for the first time I am aware of — explicitly included “traceable accounts” of the judgments that were made about the level of confidence in the various findings. This is in contrast to the IPCC, in spite of the fact that the idea of “traceable accounts” was introduced in the IPCC context by the late Stephen Schneider and his colleague Richard Moss (who is a lead author for chapter 26 in the NCA) in 2000. The traceable accounts in the NCA provide three categories of information for each “key message”:
- Description of the evidence base: typically 1-3 paragraphs with descriptions of broad types of evidence, as well as citations of specific papers, especially reviews that themselves incorporate multiple citations.
- New information and remaining uncertainties: a description of evidence that has been added to knowledge available at the time of the previous assessment, and of remaining gaps and the prospects for reducing them.
- Assessment of confidence based on evidence: A sentence or two specifying whether confidence in the judgments is very high, high, medium or low.
These four categories of confidence are in turn given the following definitions:
- Very High: Strong evidence (established theory, multiple sources, consistent results, well document and accepted methods, etc.)
- High: Moderate evidence (several sources, some consistency, methods vary and/or documentation limited, etc.), medium consensus
- Medium: Suggestive evidence (a few sources, limited consistency, models incomplete, methods emerging, etc.), competing schools of thought
- Low: Inconclusive evidence (limited sources, extrapolations, inconsistent findings, poor documentation and/or methods not tested, etc.), disagreement or lack of opinion among experts.
In addition, each chapter includes a description of the number of conference calls and meetings that the chapter authors had in determining what confidence levels would be assigned to what findings.
We know more than enough to know what to do
The absence of contrarian talking points in assessments like this shows that there is actually a system of quality control that operates within science. While discredited characterizations of climate science proliferate among policymakers and the media, the NCA bases its assessment on the full scope of the best-available science. And, as I argued above, that science is clear that our knowledge of the risks is more than enough to justify rapid emissions reductions.
It really is the same problem as with smoking. There are still political actors that cherry-pick the science to suggest that smoking is less risky than mainstream scientists know it is. But anyone who held that, literally, smoking is not a health risk, and that it’s OK to keep smoking while we do further research, would not be taken seriously in a scientific assessment, and deservedly so. There are of course uncertainties remaining about smoking, but we know enough for policymakers to feel justified in adopting policies that could reduce its prevalence. The same is true about carbon emissions. But we remain stuck in the phase where policy-making is deadlocked by policymakers’ mistaken emphasis on the views of people who stand outside mainstream science.
This is not to say that projections of the patterns of future climate change are not important, and that we shouldn’t be striving to make more accurate models. We will need the best information we can get to prepare for the changes that will occur – to adapt our lives, our economy and infrastructure to new patterns of heat and water. But that’s like saying we need to study cancer to help treat it; yes, but that’s for the scientists; as citizens, we also just need to quit smoking. Now. Yesterday. We know that already.