World may warm by about 25% more than current official projections suggest, according to a New Scientist article (19 August 2026) reporting on new research.

Official climate projections are already severe, but the study indicates they understate future warming.

Lead researcher Gergana Gyuleva (ETH Zurich) states: “For a given emission trajectory, we expect warming to be 25 per cent higher.”

For example, the Climate Action Tracker’s projection of roughly 2.6°C of warming by 2100 under existing policies would rise to about 3.25°C under the new assessment.

How much the planet warms depends on cumulative CO₂ emissions and the climate’s response (climate sensitivity). Models disagree on the transient climate response (TCR)—the near-term warming from a doubling of atmospheric CO₂. Some project ~1.6°C; others up to ~3°C.

For the latest IPCC report, many higher-sensitivity models were down-weighted because they simulated more warming than observed in the recent historical record when driven by past emissions. This led to a TCR range of 1.2–2.4°C with a best estimate of 1.8°C.

Gyuleva’s team took two main steps:

Accounting for natural variability

Using Earth’s energy imbalance

Combining both approaches, the better-performing models give a TCR of 1.9–2.6°C, with a best estimate of 2.25°C (roughly 25% higher than the prior central estimate). The authors note that the results still need confirmation from additional independent lines of evidence.

In short, the research argues that previously discounted higher-sensitivity models better match both adjusted historical temperatures and satellite energy-budget trends, implying substantially more warming for any given emissions path than many current policy-relevant projections assume.

_____________________________________________________________________________________

Recent Temperature and Energy Imbalance Trends Point to Higher Estimates of Future Warming

“Recent Temperature and Energy Imbalance Trends Point to Higher Estimates of Future Warming” is a 2026 paper by Gergana Gyuleva, Erich Fischer, Reto Knutti (ETH Zurich), and Sebastian Sippel (Leipzig University), published in Earth’s Future (DOI: 10.1029/2026EF008356).

It is the study covered in the New Scientist article discussed previously. The paper addresses an apparent discrepancy in how well climate models match observations and revises upward estimates of the Transient Climate Response (TCR)—the near-term global warming expected from a doubling of atmospheric CO₂.

Climate models produce a wide range of TCR values (roughly 1–3 K, or more specifically ~1.3–3 K in the relevant ensembles). Earlier “emergent constraints” based on historical surface temperature trends consistently favored the lower end of this range and tended to down-weight or exclude high-sensitivity (“hot”) models. In contrast, recent observations of Earth’s energy imbalance (EEI—the difference between incoming solar radiation and outgoing long-wave radiation at the top of the atmosphere) show that models with higher TCR better reproduce the observed EEI trends.

The authors reconcile the discrepancy in two main ways:

• They improve temperature-based constraints by statistically removing the influence of internal (natural) variability from global temperature and EEI records, using spatial surface temperature anomalies. Earlier constraints were biased low because of cooling contributions from variability (e.g., patterns related to La Niña) in recent decades, particularly around 1981–2014.
• They develop an EEI-based constraint, focusing on trends in the short-wave and long-wave components of the energy imbalance. These show larger inter-model spread than surface temperature alone and therefore have strong potential to constrain future warming.

When both approaches are applied—especially to the more recent period (~2001–2024/2025)—variability-adjusted temperature trends and EEI trends both support higher TCR values.

The paper provides a revised TCR range of 1.9–2.6 K.

This aligns with the New Scientist reporting of a best estimate around 2.25°C (roughly 25% higher than the prior central estimate of ~1.8°C) and the implication that, for a given emissions pathway, expected warming is substantially higher (e.g., shifting a 2.6°C-by-2100 projection under current policies toward ~3.25°C).

Implications and caveats

• Constraints based on EEI are presented as a valuable additional line of observational evidence alongside temperature trends.
• Previous TCR constraints need to be revised upward due to the combined effects of accounting for variability and incorporating EEI data.
• Shorter, more recent trends carry greater uncertainty.
• The results make it increasingly difficult to exclude high-sensitivity models from the plausible range of future warming.
• The authors (and external commentators in related coverage) note that further independent confirmation is still needed.

In short, the paper argues that once natural variability is properly filtered and the fundamental planetary energy imbalance is taken into account, both recent temperature and radiation observations point to stronger future warming than many earlier observational constraints had suggested.

Climate sensitivity measures how much the Earth’s global average surface temperature rises in response to a sustained doubling of atmospheric CO₂ (from pre-industrial levels of ~280 ppm to ~560 ppm). It is one of the most important numbers in climate science because it helps translate emissions pathways into expected warming.

Yes, the authors know what they are doing — this is competent, mainstream climate science from a strong group. But a skeptical reading is still warranted, and the paper does not magically close the case on higher future warming.

Reto Knutti (ETH Zurich) is one of the more careful and prominent voices on climate sensitivity, model evaluation, and uncertainty quantification. He has a long track record in IPCC work and has repeatedly warned against over-interpreting single lines of evidence or “hot models.”

The co-authors (Gyuleva, Fischer, Sippel) are active researchers in the same community, working on internal variability, emergent constraints, and observational constraints.

This is not fringe work. It was published in Earth’s Future and presented at EGU. The authors themselves are cautious in places, noting that results need confirmation from independent evidence and that shorter recent trends carry more uncertainty.

What the paper does well

• It directly confronts a real tension: temperature-based emergent constraints had been pushing TCR downward, while Earth’s energy imbalance (EEI) trends (especially the short-wave/long-wave split) favor higher-sensitivity models.
• Attempting to remove internal variability (the “hiatus”-era cooling influence) from the temperature record is a legitimate and necessary step. Natural variability can bias short-to-medium-term trend constraints.
• Using satellite EEI data (CERES era, roughly post-2001) as an additional constraint is scientifically interesting because the energy imbalance is a more fundamental quantity than surface temperature alone.

Legitimate grounds for skepticism

Short observational record for EEI
Reliable continuous satellite EEI data essentially begin around 2000–2001. A ~20–25 year trend is still short in climate terms. Internal variability, potential small instrumental drifts, and the influence of recent large El Niño events can still matter. The paper acknowledges higher uncertainty on the more recent periods.

Method dependence of the variability filter
Statistically “removing” internal variability from global temperature using spatial patterns is clever, but it is not unique or assumption-free. Different statistical choices or different periods can change the residual forced trend. Previous temperature-based constraints that pointed lower were not obviously incompetent; they used different (also defensible) approaches.

Emergent constraints have a mixed track record
These methods can be sensitive to the model ensemble, the exact period, and shared model biases. They have sometimes been revised when new data or better understanding arrived. A single new constraint that moves the range upward should be treated as one update, not a final answer.

Still within (or only modestly above) previous ranges
Their revised TCR of 1.9–2.6°C sits mostly inside or only modestly above the IPCC AR6 very likely range of 1.2–2.4°C. It is not a radical leap into completely new territory. The “25% higher” framing in media coverage is eye-catching but depends on which baseline you choose.

Broader context still matters more than one paper
Climate sensitivity assessments rest on multiple lines of evidence (process understanding, paleoclimate, historical energy budget, models, etc.). AR6’s ECS best estimate of ~3°C and TCR of ~1.8°C reflected that synthesis. One solid paper using recent EEI + adjusted temperatures is useful, but it does not automatically override the rest of the evidence base. Other recent emergent-constraint studies have not all shifted upward by the same amount.

Recent rapid warming and rising EEI are real — but the attribution is not trivial
Earth’s energy imbalance has increased noticeably. Part of this is expected from rising greenhouse gases; part may involve aerosol changes, cloud responses, or pattern effects. Whether it cleanly implies a permanently higher TCR/ECS, or partly reflects transient factors, is still being debated.

The authors are skilled researchers doing serious work on a hard problem. The paper is a legitimate contribution that highlights why earlier temperature-only constraints may have been biased low and why EEI data deserve more weight. It is reasonable to take the upward pressure on TCR estimates seriously.

At the same time, climate sensitivity has resisted precise narrowing for decades.

Short satellite records, statistical filtering choices, model dependence, and the usual difficulties of separating forced response from variability all introduce residual uncertainty.

Treat this as an important data point that shifts the distribution somewhat higher, not as definitive proof that “official projections are 25% too low.”

Healthy scientific skepticism here means:

• Respect the expertise and the methods.
• Demand replication and integration with other lines of evidence.
• Avoid both dismissive rejection and media-style overclaiming.

The field will sort this out the usual way — through further papers, longer records, and the next major assessment.

Published: Earth’s Future

DOI: 10.1029/2026EF008356

Provided: New Scientist (19 August 2026)

Authors: Gergana Gyuleva, Erich Fischer, Reto Knutti (ETH Zurich), and Sebastian Sippel (Leipzig University)

Abstract

Climate models simulate a wide range of 21st century warming for a given forcing scenario. Constraining this uncertainty is a central challenge in climate science because of its implications for climate policy and adaptation. The transient climate response (TCR) is a key idealized metric used to quantify future warming in response to an exponentially increasing CO₂ concentration. Climate models span a range of 1.3–3 K for TCR. In attempts to constrain this range, emergent constraints on TCR based on historical temperature trends consistently pointed toward TCR values at the lower end of the range of models. However, recent evidence from trends in the short-wave and long-wave components of Earth’s energy imbalance (EEI) at the top-of-atmosphere suggests that models with higher TCR lie closer to the observed EEI trends. Here, we reconcile this discrepancy and provide a revised range for TCR of 1.9–2.6 K. Using a statistical variability-filtering approach, we show that previous temperature-based constraints were biased low due to internal variability, according to our method. We then provide an EEI-based constraint and demonstrate that short- and long-wave EEI trends have a strong potential to constrain future warming, due to the much larger inter-model spread in EEI compared to surface temperature. When considering the recent 2001–2025 period, our results show that both surface temperature and EEI trends support higher TCR values than previously estimated. This result implies that it is increasingly difficult to exclude high climate sensitivity models from the plausible range of future warming.