Earth’s surface temperature is climbing at roughly 0.35 degrees Celsius per decade, nearly double the rate recorded before 2015, according to a peer-reviewed study by Grant Foster and Stefan Rahmstorf published in Geophysical Research Letters. The finding, drawn from five independent global temperature datasets after stripping out natural fluctuations like El Niño and volcanic eruptions, points to a sharp and statistically significant jump in the speed of human-driven warming. With NASA having already confirmed 2024 as the warmest year on record, the acceleration reframes the climate debate from whether the planet is warming to how quickly the warming itself is speeding up.
What the Data Actually Shows
The core claim rests on a straightforward comparison. Before roughly 2015, global mean surface temperature was rising at just under 0.2 degrees Celsius per decade, a rate consistent with long-term estimates summarized in climate.gov assessments and in peer-reviewed literature such as Allen et al. 2018. The new analysis, published in Geophysical Research Letters, finds that rate has jumped to approximately 0.35 degrees Celsius per decade in recent years. That is not a modest uptick. It represents a near-doubling of the warming pace within a single decade.
What makes this result harder to dismiss as a statistical fluke is the method behind it. Foster and Rahmstorf did not simply draw a line through raw temperature records. They used an iterative adjustment approach to filter out three major sources of natural variability, the El Niño–Southern Oscillation cycle, volcanic aerosol cooling, and solar variability. Once those influences were removed, the residual warming signal, the portion attributable to human activity, showed a clear acceleration after approximately 2015. The preprint of their study details this backfitting technique and the five global mean surface temperature datasets employed: NASA, NOAA, HadCRUT5, Berkeley Earth, and ERA5.
Five Datasets, One Consistent Signal
A single temperature record showing acceleration could reflect quirks in how that particular dataset handles coverage gaps or ocean measurements. Five records converging on the same result is a different story. The study drew on NASA’s GISTEMP v4, which provides monthly and annual global surface temperature anomalies and is maintained by the Goddard Institute for Space Studies. That record is part of the broader observational work carried out by NASA GISS, which has long tracked the planet’s warming trajectory.
It also used NOAA’s NOAAGlobalTemp v6, a separate government-maintained land–ocean temperature product with its own methodology for gridding and quality control. The inclusion of HadCRUT5, produced by the UK Met Office Hadley Centre and the University of East Anglia’s Climatic Research Unit, is especially telling. The HadCRUT5 dataset uses different coverage and infilling choices and generates uncertainty ensembles that test how sensitive results are to methodological decisions. Berkeley Earth and the ERA5 reanalysis product round out the group, each bringing distinct data processing pipelines. The acceleration held across all five, which sharply limits the chance that a single dataset’s bias is driving the headline finding.
Consistency across independent records also aligns with broader institutional observations. Global monitoring efforts led by agencies such as NASA’s Earth science division and international partners have repeatedly confirmed that the last decade contains the warmest years in the instrumental record. Foster and Rahmstorf’s work adds a new layer by showing that, once natural oscillations are removed, the underlying human-driven trend itself has steepened.
Why Natural Variability Cannot Explain the Jump
Skeptics of acceleration claims have pointed to the powerful 2023-2024 El Niño as a likely explanation for recent record heat. That argument has some surface appeal. NASA confirmed that 2024 was the warmest year on record and noted that individual years can be influenced by natural climate fluctuations such as El Niño. Climate scientists have also been investigating factors beyond greenhouse gas emissions that may have contributed to 2023’s exceptional heat, including unusual ocean patterns and short-term atmospheric changes.
But the Foster and Rahmstorf analysis was designed precisely to address this objection. By mathematically removing the influence of ENSO, volcanic aerosols, and solar cycles before measuring the trend, the study isolates the human-driven component. The acceleration it identifies is not an artifact of a single strong El Niño year layered on top of a steady warming trend; it is a change in the trend itself. Reporting in Nature’s climate coverage has framed this as a central question in climate science: whether recent record temperatures represent true acceleration or merely variability. The new study lands firmly on the side of acceleration, with statistical significance across multiple independent records.
Volcanic eruptions and solar cycles, the other two natural factors accounted for in the analysis, also fail to provide an alternative explanation. Major eruptions tend to cool the planet temporarily by injecting reflective aerosols into the stratosphere, while the roughly 11-year solar cycle modulates incoming solar energy by a small amount. Foster and Rahmstorf’s method explicitly regresses out these influences. What remains is a warming curve that bends upward in the past decade, even after the known natural drivers have been removed.
A Darker Planet, Faster Warming
If the acceleration is real and not explained by El Niño alone, what is driving it? One emerging line of research points to changes in Earth’s reflectivity, or albedo. Reduced low-level cloud cover over the oceans may be allowing more solar energy to reach the surface, effectively making the planet absorb more heat. Satellite observations have supported the idea that Earth is getting darker, meaning it reflects less incoming sunlight back to space.
This cloud feedback hypothesis is still being tested. Primary institutional data on the role of reduced cloud cover in the post-2015 acceleration remains limited, and no official IPCC attribution statement has directly addressed the speedup identified by Foster and Rahmstorf. The gap matters because it means the acceleration is better documented than it is explained. Scientists can show the warming rate has jumped; they are still working to quantify exactly how much comes from cloud changes, how much from shifts in ocean heat uptake, and how much from other feedbacks in the climate system.
What is clear is that the basic physics underpinning global warming have not changed. Greenhouse gases such as carbon dioxide trap heat in the atmosphere, and their concentrations continue to rise. Decades of satellite observations and surface measurements, supported by institutions like NASA’s Goddard center, have documented the resulting energy imbalance at the top of the atmosphere. An acceleration in surface warming suggests that, on top of steadily increasing greenhouse forcing, feedbacks and internal variability have recently combined in a way that delivers more of that trapped heat into the air and upper ocean.
Implications for Climate Targets
The difference between 0.2 and 0.35 degrees Celsius per decade may sound abstract, but it has concrete implications for global climate goals. At the slower rate, the world had more time to stay within the 1.5-degree and 2-degree warming thresholds set under the Paris Agreement. At the faster rate, those thresholds come into view years earlier than many scenarios assumed. That compresses the timeline for emissions cuts and adaptation planning.
An accelerated trend also raises the stakes for near-term policy choices. Infrastructure built today, whether power plants, buildings, or transportation systems, will operate in a climate that is changing more quickly than past averages implied. Coastal defenses sized for a certain rate of sea-level rise, for example, may be underdesigned if warming and associated ice loss proceed on the faster track suggested by Foster and Rahmstorf’s findings.
For scientists, the acceleration is both a warning and a research agenda. It underscores the need to refine climate models so they capture the interplay of clouds, aerosols, ocean circulation, and greenhouse gases that appears to be amplifying warming in the present decade. It also highlights the importance of maintaining and expanding high-quality observational networks, from satellites to ocean buoys, to track how the energy imbalance evolves.
For the public and policymakers, the message is simpler. The planet is not only warming; the warming itself is speeding up. That does not mean every year will be hotter than the last, or that natural variability has ceased to matter. It does mean that the baseline around which weather fluctuates is climbing more steeply than before. The Foster and Rahmstorf study, rooted in multiple independent datasets and careful removal of natural influences, suggests that the window for avoiding the most disruptive levels of climate change is narrowing faster than many had hoped. The physics of greenhouse gases have given us decades of warning. The new acceleration signal is a reminder that the bill for delay is coming due more quickly than expected.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
