Guest essay by Jennifer Marohasy (reposted from her website by request) with addendum by Anthony.
Picture this: it’s a hot day, and you grab a soda can that’s been in the sun. You crack it open—psssht—and CO₂ fizzes out, tickling your nose, maybe spraying your shirt if you’re slow. It’s a tiny chaos, a burst you can’t control. Now imagine that fizz across the ocean’s sun-warmed surface, covering 71% of Earth, bubbling CO₂ into the air we breathe. Wild, right? A bit mad. I reckon it’s a missing piece of the climate puzzle.
The IPCC pins it all on smokestacks—11 billion tonnes of carbon a year from fossil fuels. Even skeptics like the CO₂ Coalition echo this, leaning on guys like Ferdinand Engelbeen who do their maths by the consensus numbers on this issue of CO₂ origins.
But they might have it all back to front and be leaving out ocean chemistry and biology. In fact, I’m convinced they are.
The Keeling Curve—CO₂’s climb from 280 to 420 ppm—carries their blame. But what if the ocean’s fizzing more than they think? Their rock-solid evidence could be mostly myth.

I’ve been digging into this with Ivan Kennedy, my second guest for the webinar series ‘Towards a New Theory of Climate Resilience’. That was back in February and I’m still to process the audio from this discussion.
Instead, my focus has been on writing technical papers. Ivan and I are working through a hypothesis that could perhaps flip the climate script.
Engelbeen claims fossil fuels’ isotopic fingerprint—light ¹²C (isotope C12) dragging the air’s ¹³C-to-¹²C ratio from -6.5‰ (per mille)* to -8.5‰ since 1850—is proof of coal and oil’s guilt. Ocean CO₂, averaging 0‰ from deep waters, should nudge it up—not down. Case closed.
Except. That ¹²C/¹³C tale’s shakier than they admit. What if the ocean’s surface, warmed by the sun, fizzes CO₂ richer in ¹²C than the deep oceans 0‰?
Calcification—limestone forming in seawater—might churn out CO₂ at -10‰ or lower, diluting that delta 13 signal just like fossil fuels. It’s not the deep ocean I’m on about—it’s the top 65 meters, the mixed layer, where sunlight and warmth cause biological action. So much action that it has built the biosphere’s great carbonate deposits, even the White Cliffs of Dover.
Ivan and I talked some of this over—Great Barrier Reef, North Pacific—during our webinar (soon my first podcast—thanks for waiting!). Calcification’s no sleepy trick; it’s a biological buzzsaw—corals, algae, phytoplankton like coccolithophores churning limestone. In summer blooms, they might pump out tonnes of CO₂, light on ¹³C. Our Thermal Acid Calcification (TAC) hypothesis says nature’s pitching in more than you might think.
Ponder this next time you sip a soda: could the ocean be bubbling up a CO₂ twist?

TAC’s perhaps a second plank in my New Theory of Climate Resilience. Subscribe for irregular updates, and to know about next webinars.

This is Part 2 of How Climate Works. Part 1 was with Bill Kininmonth. I never properly processed the audio from Part 1, and I accepted the AI summary of our meeting click here.
************
When we say deep ocean carbon is 0‰ (per mille), we’re talking about its carbon isotope ratio, specifically the δ¹³C value. This is a measure of how much carbon-13 (¹³C) is present relative to carbon-12 (¹²C), compared to a standard reference.
In this case, 0‰ doesn’t mean there’s no carbon-13 in the deep ocean—it means the ratio of ¹³C to ¹²C in deep ocean dissolved inorganic carbon (DIC) is about the same as the standard reference, which is usually the Vienna Pee Dee Belemnite (VPDB). A δ¹³C of 0‰ indicates no enrichment or depletion of ¹³C relative to that standard.
Now, why is deep ocean carbon around 0‰? It’s because the deep ocean is a massive, well-mixed reservoir of carbon that’s been cycled through various processes over long timescales. Surface ocean carbon starts with a δ¹³C of about +1 to +2‰ due to photosynthesis, where phytoplankton preferentially take up ¹²C, leaving the surface water slightly enriched in ¹³C. But as organic matter sinks and decays, it releases carbon back into the deep ocean. This process, along with the mixing of water masses, balances out the isotopic signature. The deep ocean ends up with a δ¹³C close to 0‰ because it reflects a long-term average of all these inputs—biological, physical, and chemical—without much net fractionation.
In terms of carbon-13, this means the deep ocean has a pretty stable and “neutral” amount of ¹³C compared to the global carbon cycle. It’s not heavily skewed like surface waters or organic matter (which can be -20‰ or lower due to that photosynthetic preference for ¹²C). So, a δ¹³C of 0‰ tells us the deep ocean is kind of a baseline, a big pool where carbon isotopes have settled into equilibrium over thousands of years.
Addendum by Anthony: I was skeptical (gasp!) of Marohasy’s claims, so I decided to run it through Grok, but purposely asked the question in reverse of her premise, which I thought might prompt better results.. Grok does quite well with scientific questions. It seems to agree.
Marohasy states:
Except. That ¹²C/¹³C tale’s shakier than they admit. What if the ocean’s surface, warmed by the sun, fizzes CO₂ richer in ¹²C than the deep oceans 0‰?
The Grok response to this question is below:
Grok question:
On the ocean surface, does carbon dioxide with carbon 13 isotope come out of solution and into the atmosphere faster than carbon dioxide containing carbon 12 isotope?
Grok response:
Let’s dive into this intriguing question about carbon dioxide isotopologues—specifically, whether CO₂ with carbon-13 (¹³CO₂) comes out of solution and enters the atmosphere faster than CO₂ with carbon-12 (¹²CO₂) at the ocean surface. The process we’re examining is the outgassing of CO₂ from seawater into the air, and the key is to determine if the isotopic difference between ¹³C and ¹²C influences this rate.
The exchange of CO₂ between the ocean and atmosphere involves several steps: dissolved CO₂ must diffuse from the ocean’s bulk to the surface, cross the air-sea interface, and enter the gas phase. This process can be affected by both physical and chemical factors, including diffusion rates and reaction kinetics. Since ¹³C is heavier than ¹²C (atomic mass ~13 vs. ~12), we need to consider the kinetic isotope effect (KIE), where the mass difference can alter the speed of these processes.
First, let’s look at diffusion. In water, dissolved CO₂ moves toward the surface based on concentration gradients. According to Graham’s law of diffusion, the rate of diffusion is inversely proportional to the square root of the molecule’s mass. For ¹²CO₂ (molecular mass ~44) and ¹³CO₂ (molecular mass ~45), the difference is small, but ¹²CO₂ diffuses slightly faster because it’s lighter. The ratio of diffusion rates is roughly √(45/44) ≈ 1.011, meaning ¹²CO₂ moves about 1.1% faster than ¹³CO₂. This suggests that, all else being equal, ¹²CO₂ would reach the surface and outgas more quickly.
Next, there’s the chemistry. In seawater, CO₂ exists in equilibrium with carbonic acid (H₂CO₃), bicarbonate (HCO₃⁻), and carbonate (CO₃²⁻). The conversion of dissolved CO₂ to these species (and back) involves hydration/dehydration reactions, which can also exhibit a kinetic isotope effect. Because ¹³C forms slightly stronger bonds (due to its higher mass and lower vibrational frequency), reactions involving ¹³CO₂ tend to be slower. Experimental studies confirm that the hydration of ¹²CO₂ to bicarbonate is faster than for ¹³CO₂, meaning ¹²CO₂ is more readily available to outgas as it spends less time locked in these dissolved forms.
At the air-sea interface itself, the transfer of CO₂ into the atmosphere is driven by the partial pressure difference and turbulence, but isotopic fractionation can still occur. Research on gas exchange, including models and measurements of carbon isotope ratios (e.g., δ¹³C), shows that ¹²CO₂ is preferentially released into the atmosphere. This is partly because its faster diffusion and reaction kinetics give it an edge in escaping the ocean. The overall fractionation during outgassing enriches the atmosphere in ¹²C relative to ¹³C, leaving the ocean slightly enriched in ¹³C, as detailed in studies like those from NOAA’s Global Monitoring Laboratory.
So, does ¹³CO₂ come out of solution faster than ¹²CO₂? No, the opposite is true. Due to its heavier mass, ¹³CO₂ diffuses more slowly, reacts more sluggishly in seawater, and crosses into the atmosphere at a slightly lower rate than ¹²CO₂. The difference is subtle—on the order of a few percent—but consistent across physical and chemical processes. Therefore, carbon dioxide with carbon-12 outgasses into the atmosphere faster than carbon dioxide with carbon-13 at the ocean surface.
References
- Kinetic Isotope Effect – LibreTexts Chemistry
- Graham’s Law of Diffusion – Britannica
- Kinetic Isotope Effects in CO₂ Hydration – PNAS
- Carbon Isotope Measurements – NOAA Global Monitoring Laboratory
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I have previously stated here (WUWT) that I believed the claim that the delta-13 ratio was not convincing evidence for the changes in the atmosphere being the result solely of burning fossil fuels. My position was basically that the role of isotopic fractionation had not been thoroughly investigated and we didn’t have good quantitative data on the processes. That is to say, as a general rule of thumb from consideration of energy requirements, we can expect lighter CO2 molecules to leave water more easily (consider how hydrogen behaves) and for light CO2 molecules to diffuse through water more rapidly, thus also preferentially being involved in chemical reactions and state changes. I’m glad to see some evidence presented by Grok that my intuitions were correct. We now need someone who is still on a payroll to do the tedious work of pulling everything together and audit all the CO2 Cycle changes that accumulate isotopic fractionations. Thank you Jennifer and Anthony.
Clyde, what Grok and Jennifer said was basically right: there is a lot of discrimination between 12C and 13C in a lot of chemical and physical reactions and even more in biological reactions.
Where she failed is mainly in a following points:
That calcification of living creatures is the main cause of the sharp δ13C decline since about 1850 is very unlikely, as there is already a certain cause available. Nature that exactly mimics human emissions? Would be very interesting…
cant Jennifer just measure or do experiments to prove her hypothesis?
seems relatively easy. i thought she was a practical marine biologist.
i am quite a fan of jennifer so dont understand why she allows her argument to be diluted by lack of experimental or other evidence.
Thanks for the suggestions. I’m actually building on the work of Ivan Kennedy et al. https://www.mdpi.com/2673-7264/2/4/28 with plans to get a research program up and going. :-).
Marohasy has done her usual trick of tossing-in some half-baked theory then walking away from the mosh-pit.
Most of what I have read has nothing to do with Marohasy’s hypothesis and I’m surprised that so many commentators have contributed almost no understanding of the issues.
Yours sincerely,
Dr Bill Johnston
http://www.bomwatch.com.au
Firstly, much thanks to Anthony for republishing. :-).
Somewhere in this discussion thread the suggestion has been made that I have the equation the wrong way around, and that calcification is always a biologically driven process.
If I can make a few comments. The objective of the original work by Ivan Kennedy et al. was to explain the seasonal oscillation as measured at Mauna Loa.
We are still attempting a model based on ocean chemistry that explains maximum degassing in April-May at measured at Mauna Loa, following winter in the Northern Hemisphere.
For sure there are abiotic and, also, biotic processes. For sure there is precipitation of calcite into seawater as a thermodynamic abiotic response. Enzymes, specifically carbonic anhydrase speed up the reaction of carbon dioxide reacting with water to produce hydrogen ions and bicarbonate ions important for the biotic process of photosynthesis for coral building, and there are phytoplankton that are also photosynthesising and calcifying.
Please read the associated ocean chemistry as explained by Kennedy et al. 2022 in the following research paper: https://www.mdpi.com/2673-7264/2/4/28 .
Note in the results section of this paper that the Aloha data shows maximum concentrations of C02 in seawater and lowest pH values in late summer. This is at the same time as the atmospheric oscillation pC02 is minimal, clearly not equilibrated.
Thanks for your interest.
Dear Jennifer,
That indeed gives a quite good answer about the reasoning behind the new theory.
Still the arrows are at least half wrong: CO2 is released by higher ocean temperatures and absorbed by lower ocean temperatures. That contradicts half of the picture…
Moreover, the yearly ocean surface pCO2 is slightly below the atmospheric pCO2, thus the net flux over the (at least 67 past) years is from the atmosphere into the oceans, no matter what within a year over the seasons happens.
Then the question about the biological or abiotic origin of the seasonal and long term changes: there are several measurable differences:
In short: there is no proof of an abiotic process that influences both the seasonal and long-term CO2, O2 and δ13C processes in the atmosphere.
Some graphics that support what I have written:
Point 1: The δ13C level in ice cores since the LGM from Taylor Dome (TD), the past 1000 years from Law Dome (LD) and direct measurements from Point Barrow (PB) since 1980.
Point 2: The huge opposite change in CO2 and δ13C for Barrow and Mauna Loa over the seasons, pointing to organic CO2 changes, not abiotic.
Point 4: the opposite CO2 and O2 changes at different stations.

It’s an excellent question and exactly the same reasoning is used as a proxy for temperature using oxygen isotope concentrations in the ice cores to determine past temperatures on the assumption of preferential evaporation.
The change in isotopic composition indeed is exactly what is used to calculate the temperature of the oceans where the evaporation did take place and, even more important, the temperature where the snow was formed: directly out of water vapor near Antarctica which made up the ice.
The ratio change of deuterium (δD) and δ18O in the water molecules depends of the temperature of where the transition happened. Because the influence is largest for the vapor-snow transition, Antarctic temperatures weigh more in than ocean temperatures.
The result is some 8 ppmv/K for Antarctic temperatures, somewhat higher for global temperatures.
Anyway much smaller than the recent over 100 ppmv/K in the past 170 years, which can’t be caused by warmer oceans…