CO2-Temperature – 66 Million Years

March 16, 2025 rjdavison

CSS-67 – CO2-Temperature – 66 Million Years – U1 Download PDF

CSS-67 What is the CO₂/temperature relationship over the Cenozoic (the last 66 million years)? The peak temperatures and CO₂ concentrations were at their high during the Eocene Climate Optimum and were at their lows during the last glacial maximum. And you can (and people have) argued that there is a correlation between the two parameters. The “logical” conclusion, CO₂ has been driving temperatures over the Cenozoic. Of course, it is only “logical” if you conflate causation with correlation.

#climatechange #delaythegreen #globalwarming #showusthedata

This post looks at the relationship between CO₂ and temperature over the entire Cenozoic, in detail. What the data show are temperature and CO₂ moving independently of each other on the longer time scales (millions of years). This contrasts with CO₂ and Temperature over the Pleistocene Ice Age (the last few million years) where ice core and benthic foraminifera isotope data shows that the two parameters are correlating. They also show that temperature is driving the CO₂, not the other way around. That correlation/causation very likely continues throughout the Cenozoic. The Milankovitch Cycles did not suddenly pop into existence over the last few million years. Orbital dynamics and influences on the planet’s climate have always been there. They would have played out differently in the past, and data resolution becomes a bigger issue the further back in time you go.

How and why does the longer-term data look different? A more detailed discussion of the Cenozoic can be found in my CSS-10 – A Ride Though the Cenozoic post. As mentioned earlier, CO₂ and Temperature moved independently of each other (shown on the attached graph). The overall profile consists of a collection of stable temperature with variable CO₂ concentrations and stable CO₂ concentrations with variable temperature periods. They are not moving in unison until you get to more recent time intervals. There are a variety of reasons for the declining temperatures. Over the Cenozoic, the transition from the hothouse Eocene climate to the deep Pleistocene Ice Age was driven by the long-term rise in Cosmic Ray Flux (CRF) and on the slightly shorter time scales associated with plate tectonics (the closing of the Tethys Sea and the Panama Isthmus and the opening of the Drake Passage, etc.) and a couple of large extraterrestrial impacts. Plate tectonics affect the ocean cycles that redistribute energy around the planet.

There were four major periods of temperature reduction. Period 1: From 47 to 34 million years ago as the Tethys Sea gradually closed, choking off the east-west ocean circulation between Africa/India and Europe/Asia. Period 2: A much more drastic temperature drop occurred at the Eocene/Oligocene boundary (34 million years ago) that coincided with two large celestial impacts (Chesapeake Bay and Popigai), and the Drake Deep Passage opening, leading to Antarctic glaciation. Period 3: The Panama Isthmus began closing 13.9 million years ago, cutting off the east-west communication between the Pacific and Atlantic Oceans, cooling the planet further. Period 4: The Panama Isthmus closed completely 3.3 million years ago, accelerating temperature reduction into the Pleistocene Ice Age, adding in Arctic and Greenland glaciation. There were other influences (including the Cosmic Ray Flux background).

There are four periods where CO₂ is changing significantly (but temperatures are relatively stable). Period 1a: From 67 to 56 million years ago, CO₂ ranged between 250 and 500 ppm. Period 1b: From 55 to 47 million years ago, CO₂ ranged between 650 and 2,000 ppm. Period 2a: From 37 to 34 million years ago, CO₂ ranged between 700 and 1,400 ppm. Period 3a: From 34 to 14 million years ago, CO₂ ranged between 270 and 820 ppm.

Period 4: From 3.3 million years ago to the pre-industrial era, both CO₂ and temperatures have been trending down together. Reflecting the traditional relationship between ocean temperatures and CO₂ concentrations and the Milankovitch cycles. Atmospheric CO₂ concentrations will be affected by the temperatures, but the more continuous and dominant driver of CO₂ concentrations is the natural sequestration process where plants and ocean species remove CO₂ from the atmosphere/ocean, much of which eventually gets buried in new sedimentary deposits. That steady deposition is offset partially by the steady erosion of other exposed rock sequences (releasing some of that captured carbon). Major volcanic intrusions do provide some significant boosts to atmospheric CO₂ concentrations. The three major intrusions over the Cenozoic were the Deccan Traps, two phases in the North American Igneous Province, and the Columbia River Basalt Flood events.

A discussion on the CO₂/Temperature relationship must include CO₂’s climate sensitivity. But I will not relitigate all that here. A more detailed discussion can be found in OPS-80 – CO₂ Affects Temperature, but Does CO₂ Drive Climate post. For this discussion, we will assume that all the warming from the pre-industrial era to the present was due to rising CO₂ concentrations, the climate sensitivity would be 1.94 °C. Not a great assumption (since there are other forcings involved), but that is effectively the maximum the sensitivity can be, and it is a starting point. The implied sensitivity over the entire Cenozoic is 5.30 °C based on a 19 °C temperature change and a drop in CO₂ from 1200 to 180 ppm. A climate sensitivity of 1.94 °C would only drop the temperature by 5.31 °C (27.9% of the full 19 °C drop). Using a more realistic climate sensitivity (±0.75 °C), the temperature drop is just 2.05 °C (10.8% of the full 19 °C drop).

So, no CO₂ is not driving the climate over the Cenozoic. The CO₂ concentrations do affect the temperature but the other natural forcings dominate. How much CO₂ affects temperature depends on the climate sensitivity. I personally like the 0.8 °C range for a variety of reasons discussed in OPS-80, but there are peer reviewed, published papers that show much lower climate sensitivities. What do we know? The models are self-acknowledged to “run way too hot” (with a sensitivity range of 1.8 to 5.7 °C) and use unrealistically high emission scenarios. That sets the maximum sensitivity at 1.8 °C.

Using an average Total Solar Irradiance (TSI) reconstruction (instead of arbitrarily picking just one out of the 40+ available that fits the narrative), the Modern Temperature Record (MTR, 1850 to the present) can be history matched more accurately than using the current CO₂ focused models. That sets the lower limit closer to zero. CO₂’s influence is not zero, but it is not a major driver and will not lead to dangerously high temperatures that we cannot survive. The 0.8 °C range can be easily justified and would be a far better estimate for making policy decisions than our current dependence on computer models than are self-acknowledged to be wrong! That is not surprising because the climate sensitivities used are wrong!