The Scenario That Ate the Future

Publication note: This essay concerns emissions scenarios, public meaning, and risk communication. Greenhouse gases warm the planet, and severe climate risk remains real. The argument is narrower: one severe pathway was too often allowed to sound like the future itself.

The sentence in the paper

A reader could pass over the line in a hurry. It appears in a 2011 paper in Climatic Change, in a section headed “Storyline and main scenario drivers of RCP8.5.” The paper reads like technical work: figures, model names, footnotes, forcing levels, emissions pathways, assumptions about primary energy and land use.

Then comes the sentence.

“Compared to the scenario literature,” the authors write, “RCP8.5 depicts thus a relatively conservative business as usual case with low income, high population and high energy demand due to only modest improvements in energy intensity.”¹

The phrase was easy to read narrowly at the time. It had a local meaning inside a literature review. It compared a modeled pathway with other scenarios available to researchers in the late 2000s. It belonged to a paragraph, not to a campaign slogan, a newspaper headline, or a public forecast.

The phrase acquired force. RCP8.5 was no ordinary climate scenario. It was the highest of the four Representative Concentration Pathways selected for the climate-modeling community in the years before the Intergovernmental Panel on Climate Change’s Fifth Assessment Report. It described a world of rising greenhouse-gas emissions, no explicit climate policy, high population, slow income convergence, weak technological change, and a coal-heavy energy system. It was a stress path, a high-end path, a “baseline” in the language of the modelers, and eventually a public symbol.

By the time many readers encountered its consequences, they did not encounter “RCP8.5” at all. They saw maps of flooded cities, crop failures, deadly heat, mass displacement, vanishing coasts, and a late-century planet four or five degrees warmer. They read phrases like “business as usual,” “where we are headed,” “if nothing changes,” or “on our current path.” They saw a severe future described as an ordinary future, not always in so many words, but often enough in the logic of the story.

That is the saga. Not the invention of a fake scenario. Not the collapse of climate science. Not proof that warming is harmless. The real story is more exacting and, in some ways, more damning. A severe pathway was created for a defensible scientific purpose. The original papers themselves contained both caution and ambiguity. The scenario then gained a public life larger than its original warrant. It became a convenient climate future for scientists needing a high signal, journalists needing a sharp headline, public officials needing a warning, advocates needing urgency, insurers and banks needing stress tests, and critics needing an example of expert overreach. For years, every side got something from it.

Then, in 2026, the scenario machinery began catching up with the world that had changed underneath it. A new ScenarioMIP-CMIP7 paper, led by Detlef van Vuuren and more than forty co-authors, said that the CMIP6 high-emission levels represented by SSP5-8.5 had become implausible for the twenty-first century. The new high scenario for CMIP7 would remain high. It would assume policy failure. It would be built to explore severe climate risk. But it would lie below SSP5-8.5. RCP8.5 and SSP5-8.5 are not the same scenario, but they occupy the same high-end public role in successive scenario generations; this essay treats them together only when discussing that scenario-family role.²³

The public argument that followed split almost at once into two evasions. Some treated the change as a confession that climate science had been wrong all along. That claim is false. Some treated the change as a small technical update with no civic meaning. That claim is evasive. The correction did not prove fraud. It did prove that a public story had been allowed to run far ahead of its best-supported meaning.

The central claim can be stated plainly:

RCP8.5 began as a deliberately high-end baseline pathway meant to span the scenario literature, not as the most-likely climate future. But its early “business as usual” language, the inertia of climate-model cycles, and the incentives of science, media, advocacy, policy planning, and risk analysis helped turn it into a public image of the climate future. That public image then helped supply the background against which emergency-style carbon politics could present disputed choices about cost, timing, tradeoffs, subsidies, bans, and targets as the practical meaning of ’the science.’ The 2026 reset rebuts the claim that climate science was fake. It records a late correction to a framing problem that should have been corrected much earlier.

A number instead of a prophecy

The “8.5” in RCP8.5 does not mean 8.5 degrees Celsius. It refers to radiative forcing: roughly 8.5 watts per square meter by 2100. The RCPs are pathways of concentrations and forcing, not probabilities. They were not born as a ranked forecast sheet with one future labeled “likely” and the others labeled “less likely.” They were a technical device for letting climate modelers, impact researchers, and integrated assessment modelers work from a common set of inputs.

The need for that device was real. In 2010, Richard Moss and colleagues published “The next generation of scenarios for climate change research and assessment” in Nature. They described a research world in which future climate consequences would depend on the Earth system’s response to radiative forcing and on human choices about technology, economies, lifestyles, and policy. They argued that prior scenario processes had not adequately examined mitigation and adaptation, and had slowed the exchange of information among physical, biological, and social scientists.⁴

The next year, van Vuuren and colleagues published the RCP overview paper. The paper says the IPCC had asked scientific communities to develop new scenarios, but that the IPCC also decided those scenarios would not be developed as part of the IPCC process. Scenario development was left to the research community. That distinction carries weight because public claims often collapse scenario designers, model intercomparison projects, assessment authors, and the IPCC into one all-purpose authority.⁵

The RCP overview paper defines the role of the four pathways. The RCPs were meant to span the forcing levels in the open literature, from 2.6 to 8.5 W/m² by 2100. The set included one low mitigation pathway, two medium stabilization pathways, and one “very high baseline emission” pathway: RCP8.5.⁶⁷

That description is enough to correct the first false public lesson. RCP8.5 was not the middle of the distribution. It was not the default future in the ordinary sense. It was the high end of the menu.

The same overview paper makes the point more sharply. RCP8.5, it says, was representative of the high range of non-climate-policy scenarios. “Most non-climate policy scenarios,” the authors write, predicted end-century emissions closer to RCP6.⁸ That sentence is a key exhibit in the case. Even inside the original 2011 scenario architecture, RCP8.5 was high. Most no-policy scenarios were lower.

But the record is not clean enough to close the question. The companion RCP8.5 paper says the scenario is a “baseline scenario with no explicit climate policy,” the highest RCP in greenhouse-gas emissions. It also calls RCP8.5 a “relatively conservative business as usual case” compared with the scenario literature.⁹¹⁰

So the original record contains two truths in tension. RCP8.5 was technically high-end. It was also labeled in ways that made later misuse easy.

That tension did not destroy the science. It did shape the story.

The coal world

RCP8.5 was a model world with named assumptions: concrete, countable, and built from explicit premises.

It built on the A2r scenario, itself an updated and revised quantification of the older IPCC SRES A2 storyline. Its story was a heterogeneous world with a continuously rising global population, reaching about 12 billion people by 2100. Per-capita income growth was slow. International and regional convergence between richer and poorer countries remained limited. Global GDP reached about $250 trillion in 2005 dollars by 2100. Slow development implied weak progress in energy use per unit of output.¹¹

Those assumptions produced high energy demand. Technological progress was modest. Trade in energy and technology was limited. The world leaned toward domestic resources. Easily available conventional oil and gas became relatively scarce relative to harder-to-harvest unconventional fuels. Low-carbon technology improved slowly. The energy system moved toward coal-intensive choices.¹²

One number from the paper captures the character of the world. RCP8.5 assumed primary energy intensity improvement of about 0.5 percent per year across the century, well below the roughly 1 percent annual historical average between 1940 and 2000.¹³ In other words, it did not simply assume that the world burned a lot of fossil fuel. It assumed that the world became only slowly better at producing output per unit of primary energy.

The scenario’s critics later focused hard on coal. They were right to do so. A high-emission pathway with weak climate policy can take different forms depending on coal, gas, oil, land use, population, technology, and economic growth. RCP8.5’s severe forcing required an energy story in which coal and other fossil fuels remained central much longer and much more heavily than the world now appears likely to tolerate, afford, or demand under current technology and policy conditions.

The 2011 paper did not hide the fossil-heavy character. It says the future energy system moves toward coal-intensive technology choices. Later, in its summary, it describes a high-emission, business-as-usual scenario with slow economic development, high population, slow technological change, and a fossil-dominated energy system supported by extraction of large unconventional resources.¹⁴¹⁵

The public problem is not that researchers wrote down a severe fossil future. They had reason to explore one. The public problem is that, once written down and given the label “business as usual,” the scenario could travel into public speech without its coal, population, and technology assumptions attached.

A scenario without its assumptions becomes an omen.

The 8.5 world was built from assumptions: high population, slow efficiency, weak policy, and a long coal afterlife.

Why scientists used it

A fair history has to explain why RCP8.5 became popular inside science before it became notorious outside it.

Climate impact studies often need signal. If a researcher wants to know how heat extremes, water stress, sea level, fire weather, crop yields, habitat shifts, or infrastructure risk might change under strong warming, a high-end scenario gives a clearer contrast. It produces larger deltas, stronger maps, and more visible model responses. It can help expose vulnerabilities that might be hidden under a milder scenario, especially when local effects are noisy.

This is a legitimate scientific use. Engineers test loads beyond ordinary use. Insurers care about tail risk. Financial regulators run stress tests. Public officials ask what happens if a flood, drought, storm, fire season, or heat wave arrives outside the range of normal experience. A low-probability high-impact future can be relevant outside the central estimate.

RCP8.5 also had practical advantages. It became part of CMIP5, the Coupled Model Intercomparison Project phase that fed much of the AR5-era climate-model literature. Once the model runs existed, thousands of downstream users could use them. That is how shared scientific infrastructure works: a common input saves time, lets studies compare results, and creates a literature that can be assessed.

The 2026 ScenarioMIP-CMIP7 paper acknowledges the same production logic. It says ScenarioMIP experiments were among the most used scenario-based simulations of CMIP6 and contributed to IPCC assessment reports. It also names one of the key difficulties: the long period from integrated assessment model runs and harmonization to final climate-model runs meant that by the time climate output was available, some underlying projections had already become historical. Impact studies had to begin after that because they often needed downscaling and bias adjustment.¹⁶

That lag is one of the most important facts in the whole saga. Scenario systems are slow. Public reality is not. Energy costs move. Technology diffuses. Coal demand changes. Population projections shift. Policies appear. Wars, recessions, pandemics, trade fights, and supply chains interrupt forecasts. But the model archive remains useful, and because it remains useful, it remains cited.

Scientific inertia is not always a vice. A stable model archive lets researchers build comparable studies. But when the world changes under the archive, every user inherits a duty to label the work carefully. That duty was not met consistently.

The literature became too large to ignore

The footprint of RCP8.5 in the scholarly world is not anecdotal.

The three foundation papers alone show the scale of the scenario system’s scientific adoption. As retrieved for this package in late May 2026, the three foundation papers together showed roughly 15,000 citations across their publisher pages: Moss et al. 2010, van Vuuren et al. 2011, and Riahi et al. 2011.¹⁷¹⁸¹⁹

Those counts measure the central scenario-design literature. A census of every RCP8.5 impact paper is a separate task. The floor alone establishes the point: this was a deeply embedded research architecture, no stray modeling exercise.

The direct downstream count is harder. Bibliometric totals vary depending on database, search string, time window, article type, and whether “RCP8.5,” “RCP 8.5,” “RCP85,” “SSP5-8.5,” or “business as usual” is searched.

OIP also ran a metadata-index bibliometric pass for this essay using OpenAlex search and a limited Crossref metadata check. The pass retained 42,295 raw records and deduped them to 42,036 records across RCP8.5 and SSP5-8.5 scenario-family terms; for 2010 through 2020, it found 14,742 RCP8.5-family records. The same pass found continued 2026 publication activity at a combined RCP8.5 / SSP5-8.5 rate of 19.41 records per day through May 27, 2026. This is a metadata-index count, not full-text hand-coded classification.²⁰

OIP then ran a narrower framing-classification pass over a 52,710-article climate-policy and climate-risk corpus drawn from 16 journals. Within that corpus, 367 title/abstract records mentioned RCP8.5. Because this was title-and-abstract classification, not full-text hand coding, the result should be read as a measure of public-facing scholarly framing. On that surface, the dominant category was ambiguous projection: 305 of 367 RCP8.5 title/abstract records used ordinary scenario, projection, pathway, model, or simulation language without clearly marking RCP8.5 as a stress case, high-end case, high-emissions case, baseline, or business-as-usual case. Only 11 records, 3.0 percent, clearly labeled it as a stress or high-end case; 37 records, 10.1 percent, clearly labeled it as high-emissions; and 13 records, 3.5 percent, used baseline or business-as-usual framing. This does not measure every caveat inside every paper. It measures the public face of scholarship. On that surface, the tail case usually entered as normal projection machinery.²¹

Vox reported in May 2026, citing Nature, that between 2011 and 2020 more than 2,000 climate impact studies used RCP8.5 as their default future. OIP’s metadata-index count lands in the same order of magnitude for that narrower class. The larger point no longer rests only on anecdote: RCP8.5 and SSP5-8.5 became a standing research infrastructure large enough to feed public climate imagination through studies, press releases, assessments, and journalism.²²

Even with that caution, the literature record supports the core claim. RCP8.5 was widely used in major scientific publication. It was not confined to a few fringe studies. It became a standard severe pathway for late-century impact research, and in some of that literature it was treated or described as a default, baseline, business-as-usual, or no-policy future.

That last distinction is the one the essay must keep making. “Used” is not the same as “misused.” A coastal engineer using RCP8.5 as a high-end sea-level stress case is not doing the same thing as a headline presenting an RCP8.5-derived outcome as the future children should expect to inherit. Counts have force only after the use is classified.

The signal is not mass explicit business-as-usual labeling; it is the rarity of visible stress-case labels in a literature large enough to shape public imagination.

How the public heard it

Science rarely enters public life in its full grammar. It enters through compression.

The compression of RCP8.5 followed a predictable route. Researchers used the scenario to study high-end impacts. Press releases described those impacts. Journalists translated the results into ordinary language. Editors selected dramatic findings. Advocacy groups recirculated them. Some political and educational materials folded those warnings into broader lessons about responsibility, urgency, and inherited risk. The public heard less about forcing pathways and more about “what happens if we do nothing.”

That phrase is not always wrong. A no-climate-policy baseline can legitimately be described as a future without climate policy. But by the 2010s, “if we do nothing” could mean several different things: no explicit climate policy; no additional policy beyond measures already enacted; no policies beyond announced pledges; no change in emissions trends, technology costs, market response, adaptation, policy learning, demography, or political bargaining.

Those are not the same world.

A public audience is rarely told which one it is seeing.

A British Columbia public-sector building standards document makes the problem concrete. In an appendix asking what future climate will look like, the document describes RCP 8.5 as a business-as-usual pathway in which past emissions practices continue; nearby, it describes SSP5 as fossil-fueled business-as-usual development. This is not a newspaper scare line. It is public-sector guidance for climate-resilient buildings. The document also contains context and caveats, and it does not prove that B.C. policy depended on RCP8.5. It proves the narrower point: the language had traveled. A high-end/no-policy scenario could enter ordinary infrastructure guidance as the public image of future climate.²³

By the time a model result reached public language, the assumptions had often fallen out of the chain.

That is how “business as usual” became a trap. The phrase sounds ordinary. It carries the scent of common sense. It says: keep doing what we are doing, and this is where we end up. But the RCP8.5 world went beyond simple continuation of the present. It was a particular high-population, slow-technology, fossil-heavy, weak-policy future. By the time the clean-energy cost curve bent, global population projections came down, and coal did not follow the extreme path, the old phrase had become more misleading with every year.

In January 2020, Zeke Hausfather and Glen Peters made that point in Nature with unusual directness. Their comment was titled “Emissions ~ the ‘business as usual’ story is misleading.” The subheading was even clearer: “Stop using the worst-case scenario for climate warming as the most likely outcome ~ more-realistic baselines make for better policy.”²⁴

This critique came from analysts working inside the climate and energy debate. It cited the original scenario literature and the energy-economics critiques of Justin Ritchie and Hadi Dowlatabadi.²⁵ The warning was available years before the 2026 reset.

When public labels change only after a new model cycle forces the issue, the failure reaches beyond technical error. It is a failure of public correction.

Journalism liked the tail

The media record should be judged directly. The problem was avoidable because the basic questions were obvious. What assumptions drive this scenario? How likely are they? Is this a stress test or a forecast? Are the coal, population, technology, and policy assumptions tracking the world that actually exists? What happens to public trust if we turn an upper-bound pathway into the road ahead?

Those questions were ordinary. They were the minimum standard for honest risk communication. Too often, the institutions carrying RCP8.5 to the public did not ask them with enough force, did not answer them in plain language, or treated the people asking them as obstructionists. That posture inverted the scientific habit. Science is not a loyalty oath to a preferred scenario. It is a method for asking whether a claim survives contact with evidence.

RCP8.5 helped move a contested catastrophe pathway into the public language of “settled science.” Once there, it became easier to treat disputes over cost, timing, tradeoffs, and policy design as disputes over physics itself.

RCP8.5 became useful because it made uncertainty easy to dramatize. Academic authors got clearer signals and more dramatic impacts. Journals got sharper abstracts. University press offices got warnings that fit inside a headline. Journalists got stakes. Advocacy groups got urgency. Public bodies got justification. Risk analysts got a tail case. Critics of climate policy, once the old high end became implausible, got an indictment.

That is the failure. It does not require intentional lies or coordinated deception. Ordinary incentives can produce serious public distortion when every institution benefits from escalation and almost nobody gains from asking the dull question that would have changed the story: does this frightening number rest on a coal-heavy, high-population, slow-technology world that is already slipping out of plausibility?

The irony is sharper than the insult: a public-facing scientific system treated a testable scenario question as a loyalty test. The questioners were not asking for permission to ignore physics. They were asking whether a particular scenario continued to fit the world. Treating that question as heresy did damage to science itself, because it turned a testable claim into dogma and turned scrutiny into blasphemy.

The media’s role remains central. Most people never read Riahi et al. or van Vuuren et al. They received the scenario through stories, graphics, videos, public-facing materials, policy summaries, and cultural warnings. Too often they heard a stress case as a forecast and a high-end pathway as the road ahead.

By 2026, the correction had entered mainstream journalism. Vox’s self-audit matters as one public concession of what defenders had often minimized: climate journalism leaned on RCP8.5 outcomes without consistently telling readers that the future being described depended on severe assumptions. Vox also names David Wallace-Wells’s 2017 “The Uninhabitable Earth” as a major climate journalism event built largely on RCP8.5 projections, and notes that Wallace-Wells later revised his view in 2022.²⁶²⁷

AP and Le Monde then described the reset in plainer terms: scientists were jettisoning old extremes as no longer plausible, the new proposed worst case sat below the old coal-heavy path, and RCP8.5 had often been used in media and some scientific publications as business as usual.²⁸²⁹ Those accounts put the correction in public view. They do not erase the years of bad framing. They show what should have been true all along: no scenario deserves protection from ordinary questions just because someone has wrapped it in the phrase “settled science.”

The Reset and the Defense

The critique deserves primacy because the public failure came through use, emphasis, and repetition. The original RCP8.5 paper supplied the dangerous vocabulary: baseline, no explicit climate policy, business as usual. The overview paper supplied the limiting facts: very high baseline, high range of non-climate-policy scenarios, and most no-policy scenarios closer to RCP6. Public language absorbed far more of the first vocabulary than the second.

Once RCP8.5 became familiar in the model archive, it developed a self-sustaining presence. Papers used it because model output existed; impact studies fed press releases; press releases fed headlines; headlines fed public understanding. By the time internal critics objected, the scenario had become a climate-cultural object.

The 2026 reset sharpened the critique because it came from inside the scenario process. The CMIP7 paper says the high-end CMIP6 levels quantified by SSP5-8.5 have become implausible for the twenty-first century, based on renewable-cost trends, climate policy, and recent emissions trends; it also notes prior critiques of SSP5-8.5 and RCP8.5.³⁰³¹ That is a formal correction, not a fraud confession. It means the old high end no longer fits the plausible scenario space for the coming model cycle.

The fair defense remains narrow. High-end scenarios belong in serious analysis. Public risk work needs more than central estimates, and RCP8.5 had legitimate early uses, including nearer-term comparisons when early emissions tracked portions of high pathways more closely than later debate remembered. Climate risk also remains serious under lower scenarios. A world headed toward roughly 2.5°C to 3°C of warming is unsafe, costly, and unevenly damaging.

The 2026 paper keeps high scenarios in the toolkit, but with clearer discipline. It proposes a new high-emissions scenario, including rollback of current mitigation policies, expected to fall below SSP5-8.5; it says the scenario set should cover plausible outcomes from policy failure to stringent mitigation.³²³³ It also explains the production lag that made some CMIP6 assumptions dated by the time downstream impact research began.³⁴

The authors propose seven scenarios across a plausible range, define plausibility as a non-negligible likelihood, and refuse to attach formal probabilities to individual scenarios.³⁵ The main correction cuts both ways: the CMIP6 high-emission levels represented by SSP5-8.5 have become implausible, while some old low-end trajectories also became inconsistent with observed 2020-2030 trends.³⁶ The new high scenario assumes severe policy failure but lower forcing than SSP5-8.5; the medium scenario extends current policies and trends and is explicitly not the most-likely future.³⁷³⁸

That is the lesson the old RCP8.5 fight should have taught sooner: if experts do not say what a scenario is not, the public may decide for them. The CMIP7 paper keeps uncertainty alive, including subjective plausibility, conditional assumptions, and possible futures outside the ScenarioMIP range.³⁹ The old high pathway is gone as a plausible century-long emissions path. Severe climate risk is not gone. Public communication needs both sentences.

What the Reset Did Not Prove

The 2026 reset does not mean the IPCC admitted RCP8.5 was fraudulent. The paper is a ScenarioMIP-CMIP7 paper in Geoscientific Model Development, designed to support future scenario work and assessments. Le Monde reported that the project is separate from the IPCC and that the IPCC said it does not own the new scenarios or run climate models; it assesses literature.⁴⁰

The distinction is important because “the UN admitted fraud” collapses a research process, a model-intercomparison project, and an assessment body into one slogan. The reset also does not make RCP8.5 useless. It had legitimate high-end uses, including near- and mid-century comparisons, even after its century-long emissions path lost plausibility.

Nor does the reset make current-policy warming mild. AP reported a middle scenario around 3°C by 2100; Vox put current policies around 2.6°C; Le Monde described plausible pathways around 2.5°C to 3°C.⁴¹⁴²⁴³ The plausible center moved down from old worst-case rhetoric, not down to safety.

A cleaner public story would have said this years ago: the world is not heading for the old coal-heavy RCP8.5 path as the central case; the world remains headed toward damaging warming under current policy; severe tail risk remains relevant; and every study using a high-end pathway should say so in the headline, abstract, press release, and first public paragraph.

Instead, the correction arrived late enough that both sides could misuse it.

From Scenario to Carbon Politics

RCP8.5 did not need to appear in every statute to matter. Policy rarely moves that cleanly. It moves through assessments, summaries, risk models, speeches, agency documents, central-bank scenarios, public-facing materials, and the moral atmosphere those documents create. Once a high-end pathway was heard as business as usual, it helped supply a public background image in which severe carbon policy could be presented less as one contested political choice among others and more as the necessary meaning of climate science.

The IPCC assessment process gave the scenario architecture enormous institutional reach. The panel did not create the RCPs, and AR6 later used SSP pathways instead of the original RCP labels. But IPCC summaries placed very low, low, intermediate, high, and very high emissions futures beside one another as the official grammar of climate risk.⁴⁴ In public translation, that grammar often lost its caveats. A scenario meant to explore consequences under severe assumptions could become what “the science” said would happen if governments did not act.

Net-zero politics operates through a separate chain. Physical risk becomes social priority; social priority becomes policy target; policy target becomes a claim about permitted cost. The European Climate Law writes climate neutrality by 2050 into binding law and requires EU institutions and member states to make progress toward that target. The UK Net Zero Strategy sets out the government’s plan to decarbonize the economy and deliver net zero by 2050, with a climate-science annex supplying the physical-risk frame for that program.⁴⁵⁴⁶⁴⁷

Those documents do not need to say “RCP8.5 made us do this.” The stronger point is about permission. A public already taught to imagine the old high-end pathway as the road ahead is easier to govern through emergency. Costs that would normally require argument can be presented as the price of listening to science. Doubt about policy design can be recoded as doubt about warming itself. That is the settled-science category error: a real physical finding becomes a shield around a disputed political program.

Financial supervision shows another part of the translation. NGFS scenarios were created to give central banks and supervisors common climate-risk pathways. The Bank of England describes climate scenario analysis and stress testing as tools for assessing climate-related financial risks, and the ECB has used climate stress testing to examine bank and company exposure under alternative transition and warming paths. This is the proper language for a stress test. It becomes dangerous when stress-test logic migrates into public politics without the stress-test label.⁴⁸⁴⁹⁵⁰

The Bank of England’s 2021 Climate Biennial Exploratory Scenario gives the mechanism a public form. The exercise asked major banks and insurers to test current balance sheets against Early Action, Late Action, and No Additional Action climate-policy scenarios. The Bank said the scenarios were not forecasts, but plausible representations under different government policy paths; it also said the No Additional Action scenario deliberately brought forward later-century physical risks so supervisors could examine more severe exposure.⁵¹ This does not prove that RCP8.5 dictated the exercise. It proves the narrower and more defensible point: climate pathway logic had entered the public machinery of financial supervision, where a policy assumption, a temperature path, and a risk model could be translated into exposures, business-model questions, and supervisory policy.

This is how scenario logic can become policy atmosphere. RCP8.5 helped supply a nightmare baseline in public culture. Science and media communication could convert high-end impact studies into public expectation. Policy institutions could convert public expectation into urgency. Financial regulators then translated climate pathways into balance-sheet exposure. “Settled science” language then fused those layers together, as if the evidence that greenhouse gases warm the planet also settled which carbon targets, costs, timelines, subsidies, bans, and tradeoffs democratic societies had to accept.

The youth fear question

The most morally charged version of the RCP8.5 story is the claim that a generation of children was frightened by exaggerated climate futures.

There is evidence that youth climate anxiety is real. Surveys reported in 2021 and 2024 found high levels of climate worry among young people, including fear about the future and statements that climate worries affected daily life. Those findings should not be waved away as mere politics or adolescent fashion. Young people grew up amid fire seasons, floods, extreme heat, climate protests, public warnings, social media, and repeated claims that their future was being stolen.

But a careful essay cannot assign that anxiety to RCP8.5 alone. The public climate imagination had many sources: observed disasters, media catastrophe, activist rhetoric, scientific reports, public-facing materials, political speeches, NGO campaigns, documentaries, social media, family conversations, and genuine uncertainty about future risk. RCP8.5-based studies were part of that world. They were one source among many.

The responsible claim is narrower and stronger than the slogan. Catastrophic climate messaging likely drew some of its force from high-end scenario studies, including RCP8.5 and SSP5-8.5. Those studies were often simplified in public culture into images of the future awaiting today’s children. That simplification plausibly contributed to fear. The evidence does not prove that RCP8.5, by itself, caused youth climate anxiety.

That distinction does not soften the indictment. Adults too often let a technical upper-bound world enter public culture as the future waiting for children. When young people are told the future has been stolen, the adults speaking owe them basic precision: which scenario, which assumptions, which probability, which time horizon, which uncertainty, and which agency.

RCP8.5 failed that standard too often in the world around the science, and sometimes in the language of the science itself. The damage reached beyond emotion. It trained parts of public culture to experience uncertainty as doom and to treat questions about assumptions as moral betrayal.

Who owns the error?

The easiest answer is: journalists.

They did own part of it. Headlines flattened scenario language. Stories foregrounded impacts and buried assumptions. Public-facing summaries used “business as usual” without enough explanation. Some climate journalism preferred the most vivid future to the most plausible one. The high-end path became narratively useful because it made the stakes legible.

But journalists did not create the scenario. They did not write the 2011 phrase “relatively conservative business as usual.” They did not design the model cycles. They did not decide which scenarios would be run by most modeling groups. They did not write every abstract, title, press release, grant description, assessment paragraph, or adaptation plan.

Scientists own part of it too. The charge is not fraud. The charge is failure to ask and answer basic questions in public while their tools were shaping public fear. A technical community that knows a term is being misunderstood has a duty to correct the misunderstanding while the misunderstanding carries public force. If the public hears a stress test as prophecy, the expert cannot retreat into jargon and call the confusion someone else’s problem.

Scientific institutions own part of it. Journals, universities, assessment bodies, research programs, and press offices gave the scenario institutional authority. They also helped create a social environment in which basic scrutiny could sound like disloyalty. That is where the behavior became most damaging. A community committed to truth should welcome the question, “Does this scenario fit the evidence now?” When that question is treated as denial, the institution has confused science with dogma.

Public bodies own part of it. A city, agency, regulator, central bank, or education agency may legitimately use a severe scenario for stress testing. But if it presents that stress case as the expected future, or lets readers infer that expectation, it has moved from prudence to distortion.

Advocacy groups that used high-end findings without clear probability language own part of it. Urgency gives no license to blur probability. Fear may mobilize attention, but fear purchased with sloppy scenario language has a long afterlife. It erodes trust when the public later learns that a future presented as the road ahead was the high end of an aging model menu.

Critics own part of it too. Some were right early about RCP8.5 misuse. Some then used that victory to imply that climate risk itself had been exposed as a hoax. That move is opportunism with better footnotes.

The error was distributed. So was the benefit.

That is why accountability is hard. There is no single villain. There is a system of small permissions: a phrase left vague, a scenario label left unexplained, a press release tuned for drama, a headline written for shock, a policymaker happy to quote the upper bound, an editor bored by caveats, a research institution content to let prestige do the work, and a culture ready to treat dissenting questions as bad faith.

The public did not need fraud to be misled. It needed a severe scenario that everyone found useful for slightly different reasons, plus an institutional reflex that treated ordinary scrutiny as disloyalty.

The publication test

Any public account of RCP8.5 should answer four questions: who made the claim, which scenario was used, how was it framed, and what caveats reached the reader?

The judgment after the caveat

The caveat is limited. The future is no point forecast; climate systems have feedbacks, economies surprise modelers, politics can fail, and a severe pathway can reveal vulnerabilities outside the likeliest path. But uncertainty makes language more important, not less. A high-end pathway deserved a clearer public label once business as usual began to sound like ordinary forecast. By 2020, the problem had been named clearly. By 2026, the scenario community had moved the high end down. The correction showed that public climate risk communication leaned too long on a pathway no longer defensible as a plausible century-long emissions future.

The deeper critique is cultural. Too many people who claimed to speak for science treated scenario scrutiny as an attack on science. That was backward. Scrutiny is the scientific act. Refusing it, or punishing it with social and institutional suspicion, weakened the truth claim.

RCP8.5 deserved a place in the scientific archive. It did not deserve the public role it came to play. The failure is not that researchers studied a bad future. The failure is that too many public messengers let that bad future sound like the one already promised.

After the old future

There is a strange quietness to a scenario table.

A table does not look like a cultural weapon. It has rows, labels, notes, references. It tells modeling teams what to run. It tells impact researchers where to begin. It tells assessment authors which futures will soon appear in the literature. It does not tell a child whether she should be afraid. It does not tell a city how to talk to residents about flood risk. It does not tell a journalist where to put the caveat. People do that.

The CMIP7 table will now carry a different high end. The old 8.5 future has not vanished from the archive. It will keep appearing in old papers, old reports, old maps, old presentation decks, old lawsuits, old adaptation plans, and public memory. Scenarios have afterlives.

The work now is not to pretend the mistake was everything. It was not. Climate risk remains. Nor is the work to pretend the mistake was nothing. It was not. A public culture that cannot distinguish a stress test from a forecast will not make better decisions by shouting “science” louder. It will make better decisions only by asking the basic questions that should have been asked before the fear was exported.

The 2011 sentence remains there in the paper. It did what sentences sometimes do after publication. It escaped its paragraph.

The next sentence is ours to write more carefully.