Climate Scenarios and Reality

Progress on the important issue of climate change requires a framework for evaluating the likely consequences of different courses of action. Science can powerfully inform public decisions on energy systems, infrastructure, and economic policy when researchers explore, using the best available evidence, a range of possible futures using emissions scenarios. The process of constructing, describing, and using these scenarios is challenging for many reasons. The continued evolution and improvement of emissions scenarios is an important element of the future of climate-change research. But in “How Climate Scenarios Lost Touch With Reality” (Issues, Summer 2021), Roger Pielke Jr. and Justin Ritchie are wildly off base in declaring that the “misuse of scenarios in climate research has become pervasive and consequential—so much so that we view it as one of the most significant failures of scientific integrity in the twenty-first century thus far.”

Their characterization is wrong for three main reasons. First, the scenario developers and the Intergovernmental Panel on Climate Change have been explicit about the features of the scenarios and the limits on their relevance to specific applications. In particular, the high-emissions RCP8.5 scenario has long been described as a “business-as-usual” pathway with a continued emphasis on energy from fossil fuels with no climate policies in place. This remains 100% accurate, even if RCP8.5 does not appear to be the most likely high-emissions pathway.

Second, one of the main motivations for emissions scenarios is to provide a basis for comparing futures with and without policies related to climate change. Until recently, it has been reasonable to expect that a no-policy future would be a world of continuing high emissions and ongoing emphasis on fossil fuels, namely RCP8.5. As greater understanding of climate change spurs new policies and advances in technology, the notion of a no-policy world becomes increasingly abstract. But a no-policy endpoint remains an important point for comparison, even after the world has begun to diverge from the no-policy path. Referring to this no-policy endpoint as business-as-usual is imprecise, but it is not a significant failure of scientific integrity.

Third, at least part of the reason that the world is moving away from RCP8.5 and toward lower emissions is that effective communication of risks from a changing climate (and the unacceptable consequences to society of the business-as-usual scenario) has stimulated technology advances, incentives, and policies that now make RCP8.5 unlikely. Progress in tackling the risks of a changing climate, even if progress is still too slow, should be celebrated. It should not be converted into an implied failure of scientific integrity. Around the world, tens of thousands of scientists are working hard to understand the details of climate change and the risks it brings. The research tools are imperfect, and the future has many features that are unknowable. In this setting, the key to maintaining the highest standards of scientific integrity is maintaining commitments to professionalism and transparency, including continuing to fine-tune the development, use, and interpretation of emissions scenarios.

Chris Field

Marcia McNutt

“All models are wrong,” said the renowned statistician George Box, “but some are useful.” The same could be said of future predictions. Climate models have proved enormously useful and minimally wrong: they have captured the observed pattern and magnitude of human-caused global warming stunningly well. But they don’t even try to predict the future. Instead, they make projections: incomplete but informative pictures of possible worlds conditional on different carbon dioxide emissions scenarios.

I agree with Roger Pielke Jr. and Justin Ritchie’s statement that we shouldn’t call the high-emissions RCP8.5 scenario “business as usual,” and they are right to call for the climate community to end this sloppy wording. The world appears to be off that particular nightmare trajectory, but horrors still await us if we fail to rein in greenhouse gas emissions. We don’t know what the future holds, but we are clear that the biggest wild card is completely within our control. This is the message that emerges from the best available climate science, a complex and remarkable picture assembled from climate models; basic theory; observations of temperature, ice, precipitation, sea level, cloud cover, and many other variables; as well as reconstructions of past climate.

I was, however, saddened and confused by the authors’ contention that the use of RCP8.5 threatens the integrity of that science. Neither the most recent Intergovernmental Panel of Climate Change report nor the National Climate Assessment claims RCP8.5 is “business as usual,” but even an unrealistic scenario can yield interesting science if used appropriately. After all, we can do experiments in a climate model that we’d never be able or allowed to do in the real world. We abruptly quadruple carbon dioxide in the atmosphere, return it to preindustrial levels, or increase it steadily by 1% every year. I am using RCP8.5 in my research right now—not because I believe it to be business as usual or our inevitable future, but because I am interested in what happens to the climate as Earth passes temperature thresholds as it warms. There is not much difference between a world that passes 1.5 degrees Centigrade and eventually warms by three degrees and a world that exceeds that threshold on its way to something hotter.

Thousands of scientists use this scenario for other perfectly legitimate reasons: to understand signals of forced change against a background of natural variability, for instance, or to compare state-of-the-art climate models to earlier generations. They do so while facing constant criticism, much of which I worry is in bad faith. As Pielke Jr. and Richie note, “Groups such as the Global Warming Policy Foundation in London and the Competitiveness Enterprise Institute in Washington, DC, are highlighting the misuse of RCP8.5 to call into question the quality and legitimacy of climate science and assessments as a whole.” I think it’s wrong to claim that the existence of a high-forcing scenario compromises scientific integrity. But for some, it’s certainly useful.

Kate Marvel

Roger Pielke Jr. and Justin Ritchie make a number of provocative claims that deserve additional scrutiny.

Since the beginning of global climate modeling, scientists have been acutely aware of the need to maximize the ratio of climate change signals to the noise of chaotic internal variability. Two approaches are widely used. One is employing large-magnitude “forcings” (such as projecting abrupt increases of carbon dioxide concentrations by as much as four times current levels, or increasing carbon dioxide levels by 1% annually) to establish patterns of future climate change. The second is using wide spreads of storyline-based scenarios, where emissions and land use/land cover change as functions of varying underlying assumptions about energy use, economic growth, and other factors. These will hopefully bracket potential future changes and explore thresholds and non-linearities in the transient climate system response.

However, as climate models have become more comprehensive, creating coherent storyline-based scenarios for all relevant inputs has become more challenging. The increased coherence requires a substantial length of time (years) for the mostly unfunded volunteer economic and energy modelers around the world to create the input files for the climate modelers who, in turn, take another couple of years to complete the multi-model simulations and make the results available. It is thus neither remarkable nor surprising that the literature available for assessments such as those by the Intergovernmental Panel on Climate Change (IPCC) relies heavily on scenarios established a decade ago, including a high-emissions scenario (RCP8.5) that was originally described as “business-as-usual,” in the event society made no efforts to cut greenhouse gas emissions.

Over time, assumptions underlying the storylines can become more or less plausible, and specific scenarios, more or less useful. This was true for scenarios devised in the early 1980s that didn’t envisage the success of 1987’s Montreal Protocol on Substances that Deplete the Ozone Layer in curbing emissions of chlorofluorocarbons or foresee China’s rapid industrialization. Notably, we agree that the concept of a business-as-usual scenario in today’s fast-moving policy environment is poorly defined—particularly for a general audience—though neither recent IPCC reports nor the National Climate Assessment use such terminology.

Despite claims by Pielke Jr. and Ritchie, the use of a wide range of plausible scenarios is neither a blunder on par with misidentified cancer cell lines (an absurd claim) nor an issue of “scientific integrity.” Rather, the scientific community is already responding to the need for increased diversity and real-world grounding of projections, as well as new conceptual approaches. New scenarios are continually developed for many different purposes, for example, to assess the climate impact of the COVID-19 pandemic. Additionally, there is already movement to assess impacts based on the commonly projected “scenario-free” global warming levels of 1.5 degrees Centigrade, 2ºC, 3ºC, and so on, which can be used broadly to quantify impacts for any new proposed scenarios.

Faster updates could be accelerated by institutionalizing scenario development and associated climate model input files. More focus on scenario-free analyses would also be useful. Certainly, increased communication between economic and energy modelers, climate modelers, and impact modelers is welcome. We stress that the use of a scenario such as RCP8.5 tells nothing about whether the results depend on the realism of the scenario itself. Thus, assessing the worth of scientific contributions by counting which scenarios are mentioned is like assessing honesty by counting the number of times the word integrity is used in an article; it is both pointless and misleading.

Gavin A. Schmidt

Peter H. Jacobs