Climate change attribution and legal contexts: evidence and the role of storylines

Abstract

In a recent very influential court case, Juliana v. United States, climate scientist Kevin Trenberth used the “storyline” approach to extreme event attribution to argue that greenhouse warming had affected and will affect extreme events in their regions to such an extent that the plaintiffs already had been or will be harmed. The storyline approach to attribution is deterministic rather than probabilistic, taking certain factors as contingent and assessing the role of climate change conditional on those factors. The US Government’s opposing expert witness argued that Trenberth had failed to make his case because “all his conclusions of the injuries to Plaintiffs suffer from the same failure to connect his conditional approach to Plaintiffs’ local circumstances.” The issue is whether it is possible to make statements about individual events based on general knowledge. A similar question is sometimes debated within the climate science community. We argue here that proceeding from the general to the specific is a process of deduction and is an entirely legitimate form of scientific reasoning. We further argue that it is well aligned with the concept of legal evidence, much more so than the more usual inductive form of scientific reasoning, which proceeds from the specific to the general. This has implications for how attribution science can be used to support climate change litigation. “The question is”, said Alice, “whether you can make words mean different things.” “The question is”, said Humpty Dumpty, “which is to be master — that’s all.” (Lewis Carroll, Alice’s Adventures in Wonderland).

Characterizing the development and drivers of 2021 Western US drought

“…Additionally, we compared the CMIP6 historical 'all forcing' runs with 'natural forcing' runs to examine potential role of external forcing in driving the extreme drought conditions using the RR framework (Allen 2003, Zhang et al 2016, Reisinger et al 2020, Lloyd and Shepherd 2021. The scientific community defines risk as the potential for adverse consequences through hazard, exposure and vulnerability and can be subjected to natural, unintended or deliberate (i.e.…”

“…without anthropogenic effects). For instance, probabilities for an event occurring in both worlds are estimated to construct the so-called fractional attributable risk, or risk ratio (RR), which determines the level of human influence on the event (Allen 2003, Zhang et al 2016, Lloyd and Shepherd 2021. The observation data, global climate models (GCMs), and the method used to estimate the probabilities affect attribution results (Sun et al 2022).…”

The KNMI Large Ensemble Time Slice (KNMI-LENTIS)

“…Robust statistical analyses are often not possible due to their sparsity, not even with large ensemble data. A recently developed technique into swan-type events is using story lines (Shepherd et al, 2018;Lloyd and Shepherd, 2021;Sillmann et al, 2021). By composing a story line the extreme event can be understood in terms of its spatial and temporal meteorological context.…”

Ethics of Probabilistic Extreme Event Attribution in Climate Change Science: A Critique

“…In climate science, an extreme event can be understood as a part of a causal chain from emissions of greenhouse gases that induce global and regional warming, to a particular climate change impact (such as permafrost thaw), and ultimately some form of damage (such as collapse of infrastructure), and where each causal link needs to be supported by science (Lloyd & Shepherd, 2020, 2021).…”