Alarmism and Accountability in Climate Communication During Extreme Events

Abstract

The attribution of extreme events to climate change is now a recurrent feature of public communication. These statements are part of a wider refashioning of science and societal institutions to address the catastrophic risks of climate change. This incorporation of catastrophe into climate communication has heightened the tension between scholars concerned with undue alarmism and scholars emphasizing accountability to those most at risk or harmed by climate change. I discuss this tension in the context of extreme event attributions, their role in public discussions of losses and damages, and their implication in problems of climate communication more generally. I suggest it is time for assessments of accountability to balance fears of alarmism during discussions of disaster and denial.

Almost a decade ago, a scandal developed around questions of climate change and catastrophe in the wake of Superstorm Sandy. It was precipitated by the website, FiveThirtyEight, styled by statistician Nate Silver as a data-driven alternative to the punditry of conventional journalism. Roger Pielke Jr., a science and environmental policy expert, had written a blog post for Silver’s site. He argued that while disasters like Sandy were more damaging, the costs of such events were not attributable to climate change. Pielke (2014) claimed that discussions linking catastrophic loss to climate change were an effect of alarmist content (“climate porn”) circulating in social media and without a basis in data or research.

The claims generated intense criticism of Pielke and of Silver for using his site to amplify climate denial. FiveThirtyEight had promised to puncture media myths by grounding public discourse in data and evidence. Yet, the claim that linkages between costly disasters and climate change were unsupported by data and solely a result of perceptions distorted by social media was inaccurate. Silver responded by acknowledging mistakes, offering apologies for failing to uphold evidentiary standards, and publishing a response from hurricane expert, Kerry Emanuel.

Emanuel’s (2014) piece was notable for its careful tone amid the loud charges of alarmism and denial circulating on blogs and in social media, but also for its assumptions about the relationship of science and decision-making on questions of catastrophic risk. While Pielke had failed to detect the influence of climate change in his data, and concluded there was no relationship to increasing losses and damages due to disaster, Emanuel suggested other lines of evidence and analysis to consider when dealing with such questions. He analogized the current dangers of hurricanes to an increasing number of bears in a forest. You might not detect this increase by consulting the available data on bear maulings, but there are other ways of knowing whether the bear population has increased and grown more dangerous; if so, you might still warn those in the woods even if the evidence from documented bear attacks is inconclusive and you are not completely certain about the situation. In the case of hurricanes, Emanuel argued, there are more bears in the woods. The potential for catastrophic risk calls for a different orientation to climate knowledge and its communication (Russill, 2016).

This is one instance, hardly the first, in which climate communication centered questions of catastrophe. It is notable because instead of reducing epistemic diversity and multiple standards of evidence to a competitive struggle or polarized debate, the disagreement was moderated to expand the scope of climate communication. Raman and Pearce (2020) refer to such episodes as “cosmopolitan moments,” and suggest a different relationship of climate science and communication might be generated from these opportunities. Whether or not “cosmopolitan” is the best description of these moments, the salience of disaster in climate communication does not simply reflect activist desires to use alarming scenarios to gather attention and raise concern. Nor should it be dismissed as symptomatic of an anxious and apocalyptic cultural moment. It reflects a growing belief that decisions about the degree of certainty or confidence needed to act on climate knowledge is a decision that can no longer be withheld from those facing catastrophic harm. The FiveThirtyEight moment is worth recalling because it articulated the tension between undue alarmism and overdue accountability in the context of what we now call extreme event attribution. This opens up a different way of considering science and communication.

In the years since Superstorm Sandy, the science of extreme event attribution has grown immensely. Wide scope is now accorded to public communication in generating associations between climate change and catastrophic harm. Whether it is the specific events chosen for study, the speed with which knowledge is produced, or the determination of which standards of evidence and certainty to apply, catastrophe is now attributed to climate change publicly, without the earlier expectation to wait on peer review or scientific consensus before acting (de la Garza, 2021; Lloyd et al., 2021; Otto, 2019/2020; van Oldenborgh et al., 2021; Winsberg et al., 2020). Often this occurs while the events in question are still ongoing and the subject of intense media coverage.

The proliferation of “real-time” extreme event attribution during disasters owes a great deal to World Weather Attribution (WWA), founded in late 2014 as a collaboration between two climate scientists, Geert Jan van Oldenborgh and Friederike Otto, and Climate Central, a news organization specializing in media coverage grounded in climate science (Otto, 2019/2020). Put simply, WWA speaks to whether an event and its impacts can be associated with anthropogenic climate change (Stuart-Smith et al., 2021; van Oldenborgh et al., 2021; Wehner & Reed, 2022). The collaboration is intended to make attribution research faster and more relevant to journalists while retaining scientific legitimacy. While the immediate goal is to link disaster to climate change during moments of heightened media attention, and to respond to the needs of news organizations while they are covering disasters, the broader goal is to have scientifically credible knowledge available for moments that are often flooded by misinformation (Otto, 2019/2020). Climate lawsuits are also beginning to draw on these attribution statements to meet legal requirements for establishing causation and responsibility for harm (Lloyd et al., 2021; Stuart-Smith et al., 2021) and there are calls to integrate WWA into operational weather forecasting services so they are more readily available for determinations of losses and damages due to climate change (Wehner & Reed, 2022).

At first glance, the proliferation of these statements is interesting but poses no special problems to those studying climate communication. While there remain differences of opinion on the wisdom of embedding climate science within the dynamics of public discourse in this way, WWA’s approach to extreme event attribution has been largely validated by peer-reviewed publications, the Intergovernmental Panel on Climate Change, and scientific organizations like the American Meteorological Society and the US National Academy of Sciences. The importance of culture and inequality in shaping the availability, uptake, and significance of these statements is also part of an ongoing discussion by scholars in communication and science, technology, and society (Lahsen & Ribot, 2022).

However, the wider significance of attribution lies less in its scientific methodology than in the discussions of accountability it necessitates. In some cases, scientists have hesitated to release an attribution during disaster, generating criticism that the standards of certainty and confidence required for a statement to circulate are too severe. The desire to protect the reputation of extreme event attribution as a legitimate science has made scientists less willing to be wrong. In her WWA work on the impacts of Hurricane Harvey in Texas, Otto (2019/2020) worried about making a statement attributing elements of the storm to climate change:

So as not to make ourselves vulnerable, we must not under any circumstances make statements that are not 100 percent reliable. The U.S. attracts far more attention than the rest of the world, so we must take all the time we need. (p. 13).

The intensity of misinformation in U.S. media during extreme events can also increase the desire to have near certainty and to avoid disagreement on attribution statements in public discourse. “Ultimately, it is more important to ensure nobody can play groups of scientists against each other than to go public as quickly as possible” (Otto, 2019/2020, p. 111).

This problem is familiar to climate communicators and not entirely reducible to concerns with misinformation and denial (which are significant). It involves whether to withdraw from taking a position on alarming statements until one is completely confident in their validity, or to participate in discourse that emphasizes the dangers and harms for those most affected by climate change despite uncertainty and potential error. For the most part, climate scientists have been loath to introduce “false positives” into public discourse, or to circulate statements that wrongly connect climate change to trends and impacts, but the potential harms of failing to inform people of catastrophic risk has heightened concern with “false-negative results,” and with failures to associate threatening events with climate change in a timely fashion (Winsberg et al., 2020, p. 148).

These concerns have motivated calls to reorganize public communication of extreme event attribution around the threshold of “more likely than not,” a standard common to civil litigation in the United States, rather than “very likely” or “virtually certain” (Lloyd et al., 2021, p. 2). The argument is that the standards for evidence and proof needed to communicate publicly about extreme event attribution should reflect the conditions in which such knowledge becomes useful or actionable, not the conventions required for peer review or subjective determinations of what is necessary to protect one’s reputation. By making attribution statements available for determinations of liability, or to inform infrastructure and adaptation planning, public engagement with the varying degrees of certainty that scientists have in statements would be advanced and their claims made more useful. Perhaps, more importantly, decisions about the standards that are acceptable for acting on attributions of catastrophe to climate change would involve those most affected by its harms.

The impacts of Hurricane Harvey in Texas helped crystalize these debates given the surge of rapid attribution studies that appeared during and after the storm (Wehner & Sampson, 2021), but it is the impacts in Puerto Rico that may shape future discussions. A 2022 lawsuit filed by municipalities in Puerto Rico not only attributes losses, damages, and deaths from several 2017 hurricanes to climate change, but it also emphasizes the failure to disclose catastrophic risk as a key element of liability and responsibility. The complaint references the attribution studies connecting Harvey to damages incurred in Texas because no such studies were conducted for Puerto Rico, a fact reflecting the broader importance of colonial history to these questions (Rivera, 2020). It also references internal Exxon documents discussing the catastrophic risks of burning fossil fuels, which were not disclosed in Exxon’s public-facing communication on climate change. This raises the issue of accountability when one dampens concern or otherwise fails to inform those affected by catastrophic change—whether in advance of risks or after those harmed are trying to survive and make sense of the situation.

This complaint raises wider questions about the role of climate communication in amplifying or dampening catastrophic concerns more broadly. Attribution is an interesting and complex scientific question, but its wider significance involves assessments of accountability that have remained largely implicit or overwhelmed by debates over alarmism. It is time for more explicit questions of accountability, in addition to evidence of denial, to balance charges of alarmism in the communication of extreme event attribution, and to draw on a variety of intellectual traditions in doing so.