ConservationBytes.com

Uncertainty is to science what the score is to music. Everything scientists measure is subject to various types of ‘error’ — from measurements to models. And climate change is no exception. The Intergovernmental Panel on Climate Change (IPCC) is the United Nation’s scientific body informing the political response to the human disruption of our planet’s climate. In a recent paper, we reveal that the IPCC’s language to address climatic uncertainty is overly conservative in its assessment reports.

—

Uncertainty dominates our lives. We might grumble, but we generally accept the bus arriving late, the risotto being a little colder than we might want, or a blood test unexpectedly announcing higher cholesterol than usual. We even treat uncertainty as an inevitable outcome of many professions: meteorologists might fail to predict rain for our camping weekend, an investor might make back the money today that was lost yesterday, or Messi might fail to score in the last minute of a Champions League final.

So, we shouldn’t be surprised that scientific research is suffused with uncertainty. In fact, scientists indulge themselves with uncertainty. For them, acknowledging an uncertain outcome is not enough; instead, a critical component of their work is to quantify the amount of that uncertainty by, say, measuring the degree of variance (uncertainty) in the average height of a tree in a forest, or in the exact number of bees in a hive.

The treatment of uncertainty even gains added value when we deal with themes that directly affect our society — like climate change. But whoever communicates the uncertainty of climate science faces two formidable challenges. First, people often have a flawed understanding of science as a source of statements of fact. Instead, science is, in essence, about asking questions (1) and gradually refining the accuracy of our answers to those questions. Second, people perceive climate change through the lens of their ideology, education, and personal interests (2, 3, 4).

Therefore, what scientists say about the climate is not necessarily what some people will hear, even to the point that what scientists collectively recommend as ‘useful’ for addressing climate change might instead be deemed ‘useless’ by policymakers (5). 

Mathematical jargon

The IPCC evaluates the technical literature about climate change, and publishes such comprehensive evaluations in the form of periodical ‘assessment reports’ (AR). Those reports — which receive vast, global attention — provide policy-relevant, but not policy-prescriptive, advice to national governments and international organizations. The reports also catch the eye of a variety of audiences such as business people, educators, scientists, and the public as a whole.

To date, the IPCC has published five assessment reports in 1990 (AR1), 1995 (AR2), 2001 (AR3), 2007 (A4) and 2014 (AR5), and the next one (see progress updates here) is scheduled for 2022 (AR6) (6). With a focus on climate change, each report comprises three components: (i) the physical basis of climate, (ii) its socio-economic consequences, and (iii) adaptation and mitigation options. Every component is elaborated by a different ‘working group’ in charge of editing, integrating, and synthesizing the inputs of multiple experts.

Aware of the magnitude and various interests of its audiences, the IPCC has adopted a vocabulary known as “calibrated language” (last updated in 2010) to gauge the certainty of its evaluations using two terms (7):

• Confidence: a scientific finding is assigned to five categories from very low to very high confidence depending on the degree of available evidence and the consensus across the scientific community about that finding.
• Likelihood: a scientific funding is assigned to seven categories from virtually certain (99-100%) to extremely unlikely (0-1 %) whenever the panel can actually quantify certainty for the finding in question.

For instance, the IPCC’s AR5 states that, in the Northern Hemisphere, it is very likely that 1983 to 2012 was the hottest 3-decade period in the last 800 years (high confidence), and likely so in the last 1400 years (medium confidence) (8). As can be expected, the two statements attest that the certainty of climatic predictions into the past (or the future) should decrease with the length of time considered.

The uncertainty triplet

In our latest study just out in the journal BioScience (9), we analyse the massive effort the Intergovernmental Panel on Climate Change (IPCC) has made to communicate to its target audiences what we know and don’t know about climate science. To do so, we counted the number of times the IPCC’s AR5 allocated the different categories of confidence and likelihood to qualify scientific findings regarding the Physical Science Basis of Climate Change by Working Group I (10). Because the IPCC endorses the overwhelming scientific consensus about the human impact on climate (11), one would expect the assessment to be dominated by extreme confidences and likelihoods.

Amount of certainty in the physical basis of climate evaluated in the 5th Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) (9). Histograms encompass the Synthesis Report (Summary for Policymakers and Technical Summary), and the main text of the in-depth chapters with their respective summaries. Qualifiers of intermediate certainty predominate. The Technical Summary and the in-depth chapters result from inputs of hundreds of international scientists (contributing authors), those inputs are in turn integrated by an expert team (lead authors), and ultimately revised for completedness and style by the coordinating lead authors. The overall process is observed by two review editors, while government representatives revise line by line and contribute to the Summary for Policymakers (19). The bottom-right photo (IISD Reporting Services) shows the 47th IPCC Plenary in Paris (March, 2018) when this panel reached 30 years of existence. IPCC reports are publically available free of charge.

Surprisingly, that is not the case. On average, confidence is medium, and statements of intermediate likelihood predominate (likely, unlikely), while the proportion of categories of top/bottom certainty never reached 8%. This result holds for the 13 theme chapters (and their executive summaries), as well as for the “Synthesis Report”, including a “Technical Summary” and a “Summary for Policymakers”.

Our finding aligns with other work showing that the IPCC’s language is more cautious than that of organisations denying climate change (12). It also resonates with scientific evidence in conflict with IPCC assessments. Two instances are that the rate of global warming is expected to be 15% stronger and 30% more probable (13), and the lower bounds of seal-level rise have been argued to be > 10 cm higher (14), than the IPCC estimates. We argue that the tepid tone of the IPCC’s calibrated language arises for three main reasons:

(1) Calibration code: The IPCC’s recommendations to assign confidence and likelihood categories to climate findings (7) are intrinsically conservative. Paradoxically, this means that there is no category for findings that are 100% certain. Additionally, the Panel encourages the reporting of statements of low certainty because they can sensibly bear important implications to contrast rival scenarios for adaptation and mitigation.

Additionally, likelihood intervals are open to be interpreted conservatively by readers — think of it like being given the choice to travel in two different aeroplanes, one of which has a 66-100% chance of crashing versus the other with a 99-100% chance; clearly, our attention anchors at the lower part of the interval (66 versus 99%), despite the chance of crashing being up to 100% for both aeroplanes.

(2) Lobbies: The IPCC must carefully consider the costs in credibility that can occur when one of their evaluations proves incorrect. This forces them to “to err on the side of least drama” (15) for inherently uncertain statements. Of course, exaggeration of climate risk is bound to incite criticisms from economic and political lobbyists; in contrast, understating risk will, at most, concern a small group of scientists whose job is to improve certainty.

A painstaking case was the so-called ‘Glaciergate’, where the IPCC fell into the spotlight by mistakenly writing a 90-100% chance (very likely) for the complete melting of Himalayan glaciers by 2035, when instead they meant 2305 (16).

(3) Climate complexity: It is important to remember that the IPCC evaluates the state of the Earth’s climate as a whole, and humans are one of multiple factors influencing the climate system — the latter is of course a complex expression of different atmospheric, biotic, oceanic, and anthropogenic phenomena.

Therefore, the scientific certainty about anthropogenic climate change in the IPCC’s assessments can be masked and confused by the scientific certainty associated with a myriad of climatic observations, experiments, and trends. Thus, the confidence that different factors are causing the planet to absorb more heat than it loses (positive radiative forcing = warming) is very high for CO₂ and NO₂, but it varies from low to high for aerosols, albedo caused by land changes, carbon monoxide, dust, halocarbons, methane, non-methane volatile organic compounds, and sun irradiance (8). Each intervening factor receives a confidence rank of various strength.

What next

Uncertainty is a two-edged sword: what the IPCC and the majority of the scientific community regard as a paradigm of rigour and transparency (i.e., being explicit about the uncertainties of climate data), is exactly what the “merchants of doubt” (17) — those who benefit politically and financially from climate inaction — regard as a weakness. And they exploit that weakness by proffering their own pseudoscientific counter-positions in terms that tolerate no uncertainty.

However, climatic uncertainties are nothing but an expression of the climate risks we face, and should inspire action rather than indifference (18). This is analogous to being advised by a doctor that a bad diet or smoking is making us sick, but choosing to do nothing despite knowing that being overweight and smoking greatly increases the chances of dying from heart attack and lung cancer.

We argue that the IPCC can enhance its international relevance and credibility if its assessments incorporate a clear connection between the certainty of thousands of scientific findings under evaluation and the certainty of its overarching conclusions about the overriding human impacts on climate (9). The climate is changing, it is us, and it has adverse consequences — those three facts are beyond reasonable scientific doubt, and they deserve to emerge as such from IPCC reports.

We also advocate the creation of a fourth working group of communicators, educators, linguists, pedagogues, and psychologists pivotally centred on the effective dissemination of IPCC reports, and networking with national organizations to convey the message accurately.

Despite the irrefutable climatic footprint of human-made greenhouse gases, simply providing more and larger reports with the sole motive of improving and strengthening the message will not suddenly kick society’s response into high gear. The decision to change our fossil fuel-dominated energy model is mostly political, and so relies on our politicians. Regular citizens have a voice though; you might want to keep that in mind the next time you have to vote in an election.

If evolutionary history is enough to go on, the Earth will ultimately survive more aridity, more hurricanes, more floods, more sea-level rise, more species extinctions, and more degraded ecosystems, but our society as we know it today might not unless we clearly articulate the magnitude of the threat it poses.

References

1. Corner, A et al. (2015). The Uncertainty Handbook (University of Bristol, UK), pp. 19
2. McCright, AM & Dunlap, RE (2011). The politicization of climate change and polarization in the American public’s views of global warming, 2001-2010. Sociological Quarterly 52: 155-194
3. Kahan, DM et al. (2012). The polarizing impact of science literacy and numeracy on perceived climate change risks. Nature Climate Change 2: 732-735
4. Lee, TM et al. (2015). Predictors of public climate change awareness and risk perception around the world. Nature Climate Change 5: 1014-1020
5. Lemos, MC et al. (2012). Narrowing the climate information usability gap. Nature Climate Change 2: 789-794
6. Lynn, J (2018). Communicating the IPCC: challenges and Opportunities, in Handbook of Climate Change Communication: Vol. 3: Case Studies in Climate Change Communication, W. Leal Filho et al. editors. (Springer International Publishing), pp. 131-143
7. Mastrandrea, MD et al. (2011). The IPCC AR5 guidance note on consistent treatment of uncertainties: a common approach across the working groups. Climatic Change 108: 675-691
8. IPCC (2014). Climate Change 2014: Synthesis Report (IPCC, Geneve, Switzerland)
9. Herrando-Pérez, S et al. (2019). Statistical language backs conservatism in climate-change assessments. BioScience doi:10.1093/biosci/biz1004
10. IPCC (2013). Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Intergovernmental Panel on Climate Change, Cambridge, UK)
11. Cook, J et al. (2016). Consensus on consensus: a synthesis of consensus estimates on human-caused global warming. Environmental Research Letters 11: 048002
12. Medimorec, S & Pennycook, G (2015). The language of denial: text analysis reveals differences in language use between climate change proponents and skeptics. Climatic Change 133: 597-605
13. Brown, PT & Caldeira, K (2017). Greater future global warming inferred from Earth’s recent energy budget. Nature 552: 45-50
14. Horton, BP et al. (2014). Expert assessment of sea-level rise by AD 2100 and AD 2300. Quaternary Science Reviews 84: 1-6
15. Brysse, K et al. (2013). Climate change prediction: erring on the side of least drama? Global Environmental Change 23: 327-337
16. Cogley, JG et al. (2010). Tracking the source of glacier misinformation. Science 327: 522-522
17. Oreskes, N & Conway, EM (2010). Defeating the merchants of doubt. Nature 465: 686.
18. Lewandowsky, S et al. (2015). Uncertainty as knowledge. Philosophical Transactions of the Royal Society A 373: 20140462
19. InterAcademy Council (2010). Climate Change Assessments. Review of the Processes and Procedures of the IPCC (Amsterdam, The Netherlands)