We Cleaned the Air. The Planet Got Hotter. Warming Has Nearly Doubled Its Pace Since 2015.

Zero point three five degrees Celsius per decade. That is the new pace of planetary warming, confirmed with statistical significance for the first time in a March 2026 study published in Geophysical Research Letters by Grant Foster of Tempo Analytics and Stefan Rahmstorf of the Potsdam Institute for Climate Impact Research, a number that represents nearly a doubling from the 0.2°C per decade rate that had prevailed for the previous three decades.

Put differently: the planet is warming 75% faster than it was a decade ago, and the most important reason is one of the great public health triumphs of the 21st century.

We cleaned the air.

The Fingerprint Across Five Thermometers

Foster and Rahmstorf tested their finding against all five major global temperature datasets: NASA GISS, NOAA's Global Temperature dataset, the UK Met Office's HadCRUT series, Berkeley Earth, and the ERA5 reanalysis from the European Centre for Medium-Range Weather Forecasts. The acceleration appeared in every single one, surviving a battery of statistical controls designed to filter out short-term noise from El Niño oscillations, volcanic eruptions, and solar irradiance cycles, which are the three forces that have historically muddied attempts to detect changes in warming trends over timescales shorter than several decades.

Even including 2023 and 2024, the two hottest years in the instrumental record, the result held. The acceleration is not an artifact of a couple extreme data points riding an El Niño wave into the record books and skewing the regression. Remove those years entirely and the statistical significance remains because the underlying upward shift in the rate began around 2015, years before those record-breaking temperatures materialized.

“We can now demonstrate a strong and statistically significant acceleration of global warming since around 2015,” Foster wrote in the study’s abstract, a sentence that carries decades of frustration from climate scientists who had been watching their data outrun their models without possessing the statistical power to prove it was more than fluctuation.

The Paradox: Cleaner Air, Hotter Planet

The leading explanation is aerosol unmasking, a mechanism that climate modelers have understood theoretically since the 1990s but that is now manifesting in the temperature record with alarming clarity. When fossil fuels burn, they produce not only CO₂ but also sulfate aerosols, tiny reflective particles that scatter incoming sunlight back into space and, critically, serve as nuclei around which water vapor condenses into brighter, more reflective cloud droplets.

Dirty air, counterintuitively, cools the planet. The IPCC’s Sixth Assessment Report estimates that aerosols mask between 0.0°C and 1.7°C of warming, with a best estimate in the range of 0.5°C to 1.0°C. That uncertainty range is itself one of the most consequential unknowns in all of climate science, which is to say that we have been running a global experiment for a century and we still do not know, within a factor of three, how much of the warming we have caused is being hidden from us by the pollution we are simultaneously causing.

Three policy shifts have reduced aerosol loading significantly in the past decade. China’s Blue Sky campaign, launched in 2013 and dramatically expanded thereafter, slashed sulfur dioxide emissions from Chinese coal plants by roughly 70% over a decade, an intervention that has prevented an estimated 400,000 premature deaths per year but simultaneously reduced the cooling effect of one of the planet’s largest aerosol sources. The European Union’s progressively tightened air quality directives continued to drive sulfate emissions downward across the continent. And the International Maritime Organization’s 2020 regulation, which cut allowable sulfur content in shipping fuel from 3.5% to 0.5%, removed in a single year a substantial fraction of the sulfate aerosols that had been brightening low-level marine clouds over the busiest shipping lanes on Earth.

A 2024 study in Communications Earth & Environment by researchers from CIRES, NOAA, and Imperial College London examined the IMO 2020 effect specifically and found that the regulation measurably decreased cloud reflectivity over major shipping corridors, allowing more solar radiation to reach the ocean surface below, a net warming effect imposed not by increasing greenhouse gas concentrations but by reducing the very pollution that had been partially compensating for them.

The Timeline Math Nobody Updated

The arithmetic here is not complicated, which makes it more damning that it has not penetrated mainstream climate policy discourse.

Current global average temperature sits at approximately 1.35°C above the preindustrial baseline, after removing the temporary El Niño signal from 2023–2024. The Paris Agreement’s 1.5°C threshold is therefore 0.15°C away.

At the old rate of 0.2°C per decade, that gap closes in roughly 7.5 years, placing the threshold crossing around 2033. Every nationally determined contribution, every corporate net-zero pledge, every infrastructure investment timeline set before 2024 was calibrated, explicitly or implicitly, to this pace.

At 0.35°C per decade, the gap closes in 4.3 years. The 1.5°C threshold falls in 2029 or 2030.

Three to four years vanished from the clock, and most of the institutions responsible for responding to that clock have not updated their watches.

The carbon budget tells the same story from a different angle. The Global Carbon Budget 2025, published in Nature, reported that global fossil CO₂ emissions hit a record 38.1 billion tonnes in 2025, up 1.1% year over year, with land-use change adding another 4.1 billion tonnes. Total annual emissions: roughly 42.2 billion tonnes. The remaining carbon budget for a 50% chance of staying below 1.5°C was estimated at 170 gigatonnes as of January 2025.

Divide 170 by 42.2. The budget runs out in 4.03 years.

Early 2029. The two independent approaches, one thermometric and one budgetary, converge on the same date like witnesses corroborating a timeline in court, neither having consulted the other’s testimony.

Who Is Still Emitting

Thirty-five countries have now demonstrated the capacity to grow their economies while reducing emissions, double the number from a decade ago, which is genuine progress and worth naming as such. But the aggregate numbers moved in the wrong direction in 2025. China’s emissions rose 0.4%, India’s 1.4%, and the United States, which had been on a decade-long declining trend, increased 1.9%, a reversal attributed partly to a rebound in industrial activity and partly to the energy demands of a domestic data center construction boom that is adding electricity load faster than renewable capacity can match it.

The Global Carbon Budget noted a second finding that compounds the first: climate change itself is weakening the planet’s carbon sinks. Roughly 8% of the rise in atmospheric CO₂ concentrations since 1960 is attributable not to increased human emissions but to the reduced capacity of oceans and forests to absorb them, a feedback loop in which warming diminishes the very processes that slow warming, leaving more of each tonne emitted in the atmosphere where it accumulates and amplifies the problem it was expected to partially solve.

The Full Forcing We Have Actually Imposed

Here is the number that reframes the conversation. If aerosols are masking approximately 0.5°C of warming (the IPCC’s midpoint estimate), then the total anthropogenic forcing on the climate system is not the 1.35°C we measure in the temperature record but something closer to 1.85°C, a figure that already exceeds the 1.5°C threshold and sits uncomfortably close to 2.0°C.

The measured temperature is the warming that has made it past the aerosol shield. The committed warming is everything behind it, waiting to arrive as the shield continues to thin.

This does not mean we should stop cleaning up air pollution. Seven million people die prematurely from air pollution every year, according to the World Health Organization, and aerosol masking is not a technology anyone chose to deploy. It is a side effect of killing people slowly with particulate matter, and the fact that it happened to slow the rate at which we are also changing the climate does not convert it into a policy option.

It does mean that the 1.5°C target, as a policy instrument, was calibrated to a planet where dirty air was doing part of the work, and as we remove that dirty air, the target becomes more difficult precisely because the underlying forcing was always higher than the thermometer showed.

What This Does Not Prove

One decade of accelerated warming, even statistically significant acceleration, does not guarantee the new rate persists. The Atlantic Multidecadal Oscillation may be entering a phase that temporarily amplified recent warming. The Hunga Tonga-Hunga Ha’apai eruption in January 2022 injected an estimated 150 teragrams of water vapor into the stratosphere, a novel forcing mechanism not fully captured by the standard natural variability filters applied in Foster and Rahmstorf’s analysis, and whose warming contribution remains actively debated in the literature.

The aerosol masking estimate carries enormous uncertainty. A range of 0.0°C to 1.7°C is not a precision instrument; it is a confession that one of the most consequential variables in the climate system remains poorly constrained after four decades of satellite observation and modeling. The “total forcing” calculation above uses the midpoint of that range, but the true value could be substantially higher or lower, and the policy implications shift dramatically with it.

The 1.5°C “crossing” that the timeline math predicts refers to an annual average temperature anomaly, not the decadal average that the IPCC formally uses to define threshold exceedance. A single year above 1.5°C is not the same as the sustained warming that triggers the coral die-offs, ice sheet instabilities, and monsoon disruptions the Paris target was designed to prevent, though it is a waypoint on the road to exactly those outcomes.

The Strongest Case Against This Analysis

Rahmstorf himself offered the best counterargument: “This may not occur if the past decade turns out to be an outlier.” Natural climate variability operates on scales of decades and longer, and a ten-year acceleration embedded within multidecadal oscillations is exactly the kind of signal that has fooled researchers before. The Pacific Decadal Oscillation created an apparent “warming pause” from roughly 1998 to 2013 that climate contrarians seized upon and that climate scientists had to spend years explaining was consistent with long-term trends despite appearing to contradict them on a shorter timescale. The current acceleration could be the mirror image: a temporary overshoot of the long-term rate driven by favorable phasing of ocean cycles, which will revert toward the mean in the 2030s as those cycles shift.

If that happens, the 1.5°C threshold might hold until 2033 or later, and the acceleration will have been a false alarm in the specific quantitative sense even as the qualitative direction, toward faster warming in a cleaner-air world, remains correct.

The problem with this counterargument is practical, not scientific: waiting another decade to see whether the acceleration is real costs the same decade you would have spent preparing for it.

What You Can Do

If you set national emissions targets: Every Nationally Determined Contribution filed with the UNFCCC before 2025 was implicitly calibrated to 0.2°C per decade of warming. At 0.35°C per decade, the emissions reduction required to stay below any given temperature target increases by roughly 40 to 60%, depending on how much of the acceleration is aerosol-driven versus natural variability. NDCs should be stress-tested against the accelerated rate and revised accordingly, because the alternative is discovering the inadequacy in the temperature record five years from now.

If you price climate risk for a living: Insurance actuarial tables, real estate valuations in heat-vulnerable or flood-prone regions, and infrastructure depreciation schedules all embed assumptions about the pace of warming. A 75% acceleration in that pace reprices coastal flood risk, urban heat mortality, agricultural yield projections, and wildfire exposure on timelines that current financial models are not capturing because the models were built on the old rate.

If you fund climate research: Constraining the aerosol masking estimate from its current 0.0°C to 1.7°C range to something approaching the precision of CO₂ forcing estimates would be the single highest-leverage investment in climate science today. The difference between 0.3°C and 1.0°C of masked warming is the difference between manageable and terrifying, and right now we cannot tell which world we live in.

If you are a person alive in 2026: The 1.5°C target is, for practical purposes, no longer achievable through emissions reduction alone. This is not defeatism but arithmetic: 170 gigatonnes of budget at 42 gigatonnes per year leaves no scenario in which incremental policy adjustments close the gap. Adaptation planning, for your property, your career, your health, and your family’s location, should assume a 2°C world arriving between 2050 and 2065 as the central case, not the worst case.

The Bottom Line

The planet is warming 75% faster than it was a decade ago, confirmed across every major temperature dataset and robust to the removal of short-term climate noise. The leading cause is one of the cruel ironies of industrial civilization: the same aerosol pollution that was killing seven million people a year was also reflecting enough sunlight to hide roughly half a degree of warming from the instruments we use to measure how much trouble we are in. As China, Europe, and the global shipping fleet clean their emissions, the warming those aerosols were masking has arrived in the temperature record, and the 1.5°C Paris threshold, once projected for the early 2030s, now converges on 2029 from both thermometric and carbon budget analyses. We did not fail to act on climate change. We acted on air pollution first, and the climate change that was hiding behind it stepped into the light.