While temperatures rose and fell 3 million years ago as part of the ice-age cycle, carbon dioxide levels oscillated by 20 ppm. Rather than give up the Net Zero fantasy, climate alarmists argue this finding shows climate is ‘even more sensitive’ to CO2 levels since, they say, ‘even small changes’ can trigger ice-age shifts
By Chris Morrison, Daily Skeptic, March 25, 2026
The climate science world (“settled” division) is in shock after the discovery in ancient ice cores that levels of carbon dioxide remained stable as the world plunged into an ice age around 2.7 million years ago.
Levels of CO2 at around 250 parts per million (ppm) were said to be lower than often assumed, with just a 20 ppm movement recorded for the following near three million-year period. In addition, no changes in methane levels were seen in the entire period.
In other words, massive decreases in temperature with occasional interglacial rises appear to have occurred without affecting “greenhouse” gas levels, and this revelation has caused near panic in activist circles.
New ice core evidence challenges prevailing assumptions
The assumed level three million years ago of CO2 was around 400 ppm, a convenient mark that has been used to explain the subsequent ice age and a drop to 250 ppm. However, the recently published paper brings this explanation into question but, alas, similar explanations are mostly ignored in discussing today’s climate changesto keep promoting the Net Zero fantasy.
Some alarmists still cling desperately to a dominant CO2 role, including one of the authors of the findings published in Nature. The co-author states that the results suggest even greater climate sensitivity to the warming effect of CO2, even though the data show CO2 had almost nothing to do with the ice-age cycle at that time. In short, there is a great deal of applying the laws of physics and chemistry to one era, while applying different laws of physics and chemistry to another era.
Natural variability remains underplayed in the debate
The title of the paper, produced by 17 America-based scientists, was enough to set alarm bells ringing in the “settled” science, Net Zero-obsessed community: “Broadly stable atmospheric CO2 and CH4 levels over the past three million years.” A related paper examining ocean heat content derived from the ice core record was also published.
Carrie Lear, Professor of Past Climates and Earth System Changes at Cardiff University, claimed that the papers “don’t rewrite the role of CO2, they underline how sensitive the climate system is… that is why today’s rapid CO2 rise is so alarming.”
So even if CO2 movements are minimal, probably within a margin of potential error, they are (somehow) still responsible for large variations in temperature. That’s because the laws of climate science are “settled”—whether the trace atmospheric gas CO2 is rising, falling or generally stable, it is nonetheless almost wholly responsible for large movements in global temperature. Under this rather shaky assumption, humans must stop burning hydrocarbons and return to a neo-Malthusian pre-industrial age.
The dominant role of CO₂ questioned
Study lead author Julia Marks-Peterson noted: “We definitely were a bit surprised. If correct, the findings may suggest that even small changes in greenhouse gas levels could trigger major shifts in climate.” That’s a little bit of a scary thought, she added, possibly with an eye on future grant funding. “May suggest” is doing a lot of the work here, and it may also be suggested that more plausible opinions are available.
Quoted in New Scientist magazine, Tim Naish, Professor of Earth Science at Victoria University in New Zealand, said it was “way too early to thrown the baby out with the bathwater.” Perish the thought that baby should be given its marching orders, ending a science-lite 40-year demonization of CO2 and related promotion of a hard-Left Net Zero dream.
New data extends the climate record
The latest Nature-published research gives a snapshot from ancient Antarctica “blue” ice drilled in the Allan Hills area. It looks back further in time, well past the usual 800,000-year ice core records. For the first time, the work has pushed the direct gas measurements back into the late Pliocene era.
Over the last three million years moving into the Pleistocene, global temperatures showed a long-term cooling trend of several degrees Celsius, interrupted by increasingly large interglacial oscillations. Interglacial temperature swings, as in the current Holocene, often see temperatures rise by 5°C and more.
Critics seeking to downplay ice-core evidence often suggest it is too imprecise to provide a wholly accurate record of gas levels and temperature. But it is accurate enough to give a broad cyclical insight. It remains the source of some of the best data we have on the past climate. It is undoubtedly more accurate than most proxy evidence from millions of years ago.
But whatever the evidence used, it is hard to detect any obvious and continuous link between CO2 and temperature across the entire geological record going back 600 million years to the start of abundant life on Earth. Certainly none to justify the political notion that humans control the climate thermostat by burning hydrocarbons.
In fact the evidence is so slim that Les Hatton, Emeritus Professor in Computer Science at Kingston University, was recently able to determine from ice core records that 100-year rises of 1.1°C in the current interglacial, which started 20,000 years ago, have occurred in one in six centuries. Going back 150,000 years, the frequency was around one in six to one in 20 centuries.
None of these findings suggest that current warming is either unusual or primarily caused by human activity. Needless to say, none of these findings trouble the headline writers in narrative-addicted mainstream media.
This article has been slightly edited. To read the original version, click here.
REFERENCES
Marks-Peterson, J., Shackleton, S., Higgins, J. et al. Broadly stable atmospheric CO2 and CH4 levels over the past 3 million years. Nature 651, 647–652 (2026). https://doi.org/10.1038/s41586-025-10032-y
Shackleton, S., Hishamunda, V., Yan, Y. et al. Global ocean heat content over the past 3 million years. Nature 651, 653–657 (2026). https://doi.org/10.1038/s41586-026-10116-3