NASEM mCDR update
This week the National Academies of Sciences, Engineering, and Medicine published the first installment of its updated research strategy for marine carbon dioxide removal (mCDR). I served on the standing committee that produced it, focusing my work on the module on ocean alkalinity enhancement, or OAE.
The idea behind OAE is simple enough. Add alkalinity to seawater and the carbonate chemistry shifts: dissolved CO2 converts to bicarbonate and carbonate ions, allowing surface waters can take up more CO2 from the atmosphere. Since the ocean’s chemistry is permanently shifted, carbon stored this way stays put for millennia. The geochemistry is well established, essentially an acceleration of the natural rock weathering that has regulated atmospheric CO2 over geologic time.
The new module is a substantially expanded treatment of OAE compared to the chapter in the 2022 report, reflecting how much has happened in four years: in-water field trials, mesocosm and laboratory programs across many institutions, a fast-growing modeling literature, and companies selling carbon credits. Important research is being done and we have learned a great deal.
The central message of our assessment is that OAE remains promising and still uncertain, and these uncertainties get sharper the larger you imagine the deployment. Open ocean field trials, such as LOC-NESS in 2025, are answering important questions, but these are still quite small scale. Removing carbon at a scale that measurably changes atmospheric CO2 will present an array of challenges yet to be addressed. Four stand out to me.
First, we do not yet understand the processes that follow an intervention well enough to model them. When alkalinity enters seawater, a chain of physical, chemical and biological processes determines how much CO2 is actually drawn down. Minerals dissolve at rates that are poorly constrained for most candidate products, and if particles sink before dissolving, they are lost. If the local saturation state gets too high, calcium carbonate precipitates and consumes the alkalinity just added. Air-sea CO2 exchange is slow: the classic estimate is six months for a surface mixed layer to equilibrate, during which currents carry that water away and may subduct it out of contact with the atmosphere altogether. Uptake is therefore both delayed and spread over a wide area. Models will be essential tools to quantify these processes and the resulting impact on atmospheric CO2. Models have advanced substantially in the last few years, but we need much more work, including rigorous validation against in-water observations.
Second, distinguishing intervention carbon from everything else is hard. The ocean already absorbs roughly one-third of anthropogenic CO2 emissions, and this natural sink will remain far larger than any mCDR signal for the foreseeable future. Layered on top is enormous natural variability in space and time. Demonstrating additionality means detecting a small, deliberate perturbation against a large, noisy, shifting background. My own group works on measuring and modeling this sink, and uncertainties remain large, particularly at the local scale most relevant to mCDR.
Third, a large deployment must be net negative, and that is not automatic. Mining, grinding, transporting and dispersing alkaline material all consume energy and emit carbon. Finer grinding speeds dissolution but raises electricity demand; electrochemical approaches need large seawater throughput and low-carbon power. Transparent, consistent life-cycle accounting (LCA) has to be part of any carbon removal claim, and the report finds current practices inconsistent enough to make approaches hard to compare.
Fourth, the proposed scale-up rates and the raw resource requirements are enormous. Gigaton-scale removal by mid-century would require technology scale-up rates faster than any historical analogue. Global limestone resources are probably not the binding constraint, but achievable rates of quarrying, kiln capacity, processing, transport and at-sea dispersal may well be. Resources are also unevenly distributed, creating national and regional logistical problems that a global average hides.
None of this is an argument against OAE research. Research that keeps pace with real-world deployments is essential.
There are also significant policy and governance challenges. My Columbia University colleague Romany Webb, who served on both the 2022 committee and this one, has written an excellent summary of the governance findings on the Sabin Center’s Climate Law Blog. As she notes, governance gaps directly constrain how quickly OAE could scale: international agreements written for other purposes fit awkwardly around an activity that proposes to perturb the ocean to benefit the climate, and most countries have no legal framework tailored to mCDR at all. Her post is well worth reading in full.
We need every viable tool for addressing climate change, and the ocean is too large a part of the carbon cycle to ignore. But the case for mCDR at scale has to be built on evidence we do not yet have. Building that evidence starts with the research agenda proposed in this report.