There's a particular kind of intellectual violence that happens when an entire policy edifice, carbon taxes, net-zero mandates, the gutting of industrial economies, rests on a single causal claim that almost nobody is allowed to interrogate. Jonathan Cohler's work is an attempt to do exactly that interrogation, and he does it with the tools the establishment claims to revere: direct observation, isotope chemistry, and mass balance arithmetic.
The central provocation is as blunt as it gets: the rise in atmospheric CO₂ over the instrumental record is not primarily driven by human emissions. The fossil-fuel contribution, in his accounting, is negligible, a rounding error against the natural carbon cycle.
Cohler's most elegant argument is also his most difficult to dismiss, because it doesn't rely on any model. It relies on a controlled perturbation that humanity itself accidentally created.
Between 1955 and 1963, above-ground nuclear testing roughly doubled the atmospheric concentration of carbon-14, a chemically identical isotope of carbon that rides along inside CO₂ molecules. When the test ban went into effect, that radiocarbon pulse began to drain out of the atmosphere, and scientists have been measuring its disappearance for over five decades.
What the measurements show, according to Cohler and his co-author Willie Soon, is a near-perfect single exponential decay with an e-folding time of roughly 17 to 18 years. This is devastating to the IPCC's carbon cycle framework, because the Bern model, the standard tool for attributing CO₂ rise to fossil fuels, is built on multi-exponential decay: a fast component, several slow components, and a permanent airborne fraction that supposedly lingers for centuries or millennia.
The bomb pulse shows no such permanent component. It drains on a single, modest timescale. If the Bern model's long-lived reservoir actually existed, the radiocarbon pulse would have to betray its existence. Cohler's claim is that it doesn't, and that the IPCC's parameter choices have shifted substantially between assessment reports, which is the signature of curve-fitting to a predetermined conclusion rather than measurement of an immutable physical property.
Fossil fuels are ancient, and ancient carbon is depleted in carbon-13. If burning coal and oil were truly the dominant source of the CO₂ accumulating in the atmosphere, the carbon-13 signature of the net atmospheric input should have become progressively more negative over time, the atmosphere should be getting isotopically "lighter," like a cocktail being steadily diluted with old, ¹³C-poor carbon.
Cohler leans on an analysis by Demetris Koutsoyiannis covering four monitoring stations in both hemispheres across four decades, from 1978 to 2022. The finding: the isotopic signature of the net input stayed essentially constant. No significant trend. That flat line is incompatible with the claim that fossil emissions are the engine of atmospheric growth. If human carbon were doing the heavy lifting, the isotope record would have to move. It doesn't.
Human emissions constitute roughly 4 percent of the annual global carbon flux. The natural exchange: oceanic outgassing, soil respiration, terrestrial photosynthesis, dwarfs it by more than an order of magnitude. Cohler's mass-balance analysis shows natural flux rising by more than 30 percent over the study period while tracking atmospheric CO₂ mass remarkably closely.
The implication is blunt: a small perturbation in a massive, non-linear natural system, driven by temperature, by ocean chemistry, by biology, can easily produce the observed rise without any meaningful human contribution. The correlation between emissions and atmospheric CO₂, which is treated as causal proof, may simply be two things rising together in a warming, outgassing world.
The CO₂ argument is only the last link in a chain Cohler claims to have broken. He argues that the global mean surface temperature is a physically questionable construct, an average over a chaotic, non-equilibrium system with no single thermodynamic meaning. He argues that the ocean heat content records and the alleged planetary energy imbalance are insufficiently constrained by observation. He points out that the CMIP3, CMIP5, and CMIP6 model runs show near-zero correlation with unadjusted temperature and sea-ice records.
And underneath it all sits a single, low-variability solar irradiance reconstruction that the IPCC adopted while 27 alternatives existed: alternatives whose higher variability aligns far more closely with observed warming. Choose a different solar forcing, and the entire "anthropogenic fingerprint" narrative loses its statistical legs.
Cohler's thesis has not been refuted so much as it has been declared inadmissible. That's the tell. A claim this specific, single-exponential radiocarbon decay, flat carbon-13 trends, a 4 percent flux share, is either falsifiable or it isn't. If the Bern model's permanent airborne fraction were real, the bomb pulse would show it. If fossil carbon were dominating the atmosphere, the isotopes would show it. These are empirical questions with empirical answers sitting in public datasets.
The fact that the institutional response is coordinated dismissal rather than engaged falsification tells you what's really at stake. The CO₂ attribution claim isn't a scientific conclusion being defended with data. It's the load-bearing wall of a multi-trillion-dollar regulatory and financial apparatus. And load-bearing walls are not subjected to stress tests, they are protected.
Cohler's presentation at the EIKE conference in Halle was delivered by videoconference, which is its own quiet irony: a man calmly dismantling the carbon narrative from a screen, while the institutions that could falsify him with a single clean dataset choose instead to pretend the argument doesn't exist.
https://clintel.org/jonathan-cohler-the-rise-in-atmospheric-co%e2%82%82-is-not-caused-by-humans/