Ocean alkalinity enhancement removes carbon without planting anything. How is that measured? The chemistry, models and sensors behind OAE's proof problem.
Plant a hectare of mangroves and a verifier can, at least in principle, walk the plot. Add alkalinity to seawater and there is nothing to walk: the carbon you claim is invisible chemistry — a shift in dissolved CO₂ in a volume of moving ocean, somewhere down-current of a pipe or a beach of dissolving rock. Ocean alkalinity enhancement (OAE) has some of the fastest-dropping theoretical costs of any carbon-removal approach on the table — as our introduction to ocean alkalinity enhancement sets out — and exactly one mountain still left between it and scale: proving, creditably, how much carbon was removed. That proof problem has its own name in the field — mMRV, for measurement, monitoring, reporting and verification of marine CDR — and it's one of the most active scientific contests in climate today. Here's the shape of the problem and the toolkit being assembled to solve it.
Terrestrial and even blue-carbon MRV has a comfortable structure, rehearsed in our MRV explainer and its kelp-and-mangrove kin: measure a stock in a bounded place, compare it to a control, verify a delta on a defined map. OAE breaks every clause of that:
mMRV is, in the end, the price OAE pays for its biggest advantage: it works through the ocean's own physics rather than by building a new standing stock.
Out of those constraints has emerged a verifiable account with a three-layer architecture — each layer a different discipline, none sufficient alone.
The empirical layer measures carbonate chemistry along gradients away from the addition point: total alkalinity (TA), dissolved inorganic carbon (DIC), pH and pCO₂, temperature and salinity — the variables of seawater's carbon system that decide how much CO₂ the water can hold, as laid out in our seawater chemistry explainer. Ships, autonomous gliders and moored sensors make these measurements across the plume and past it; the change in TA along the dispersal path is the observational core of the removal estimate.
Its limits are the honest heart of mMRV: sensors carry drift, weather gates shipping, spatial coverage under-samples any given plume, and — the subtle one — alkalinity can change for non-carbon reasons, from calcification by biology to dissolution pulses from sediments. Direct measurement is therefore evidence about the plume, not a proof of the plume: it calibrates the model layer more than it stands alone.
Between the addition and the atmosphere lies the model layer: ocean circulation and biogeochemistry models that take measurements of the plume, the mixing physics of the local sea, and gas-exchange kinetics, and simulate where the alkalinity went and how much CO₂ it pulled from the air while it travelled. The model produces the removal estimate with its uncertainty and maps the fraction of the claim that occurred inside the monitored footprint versus down-stream beyond it.
Models are not a contravention of rigor — they are the only honest shape rigor can take here. (Even grassland soil carbon credits rely on models of what fields would have done unfarmed; OAE just strips this architecture naked.) But they raise the verification bar: a credit anchored in a model must show that the model has been validated against real plumes, run under the project's actual conditions, and that its uncertainty bounds are stated and priced into the claim.
The least glamorous layer is the most audit-proof: what was physically added? Ground mineral weighed out of a quarry and onto a beach, dissolved alkaline solution metered through a pipe — the addition side of the ledger is a normal inventory, and it puts a hard ceiling on what can ever be claimed, because the ocean can never remove more CO₂ than the alkalinity's chemistry permits. Mass-balance accounting doesn't tell you what the ocean did with the addition — losses to burial, bio-calcification, or mineral precipitation in the mixing zone all cash out below the ceiling — but it gives every verification an inarguable "not more than this" bound, and it makes deliberate overclaiming structurally difficult to hide.
A credible mMRV statement doesn't hide the things that still wobble; it prices them. The major sources:
Into this scientific scaffolding, the carbon-market machinery is laying its own layers: public protocols and methodologies from emerging ocean-CDR registries; registries' requirements that pilots run public data and third-party evaluation before commercial issuance; and the slow accretion of regulatory attention to who permits an alkalinity release into whose ocean — the topic of our piece on who governs ocean carbon removal. The drift of this governance is one of the encouraging signals in CDR: the community is converging on public, peer-auditable mMRV as the price of admission, not a premium-tier add-on.
It matters to kelp, too. OAE deployments are being proposed in exactly the coastal bands where kelp forests live, and the interactions extend in both directions — explored in our piece on alkalinity enhancement near kelp forests. The same reefs that stand to gain from locally moderated acidification also serve as an ecosystem-effects observatory for the measurement systems — the same sensor arrays, isotope work and biological survey infrastructure thinking built for kelp MRV is part of the foundation mMRV is being raised on. The ocean does not know which credit column an instrument belongs to; good measurement infrastructure, like good stewardship, is community property.
Not directly: no instrument can follow a specific parcel of water long enough for the drawdown to be metered as it happens. The removal is reconstructed from measured seawater chemistry plus modelled gas exchange — hence the three-layer architecture instead of a gauge.
Different, not weaker — and the rest of climate accounting is more model-based than it looks. What matters is that the model is validated against observed plumes, run under the project's actual conditions, and that its uncertainty is stated and priced into the credit, not laundered out of the public record.
The definitions: robust protocols issue credits only for removal that occurred within a defined and monitored footprint, and track water masses so claimed tonnage has a stated home. Beyond the footprint, uptake may be real but is unclaimed until governance — via mechanisms like Article 6 accounting — catches up to attribute it.
Increasingly, yes: mMRV frameworks pair the carbon ledger with environmental monitoring — biology, turbidity, sediment mineralogy — because a carbon credit from a technique that harms the system that issues it is not a serious credit. The ecological layer of mMRV is one of its genuinely novel contributions to marine monitoring.
The honest answer: when replicated pilots have shown measurement, model and mass-balance layers agreeing within stated uncertainty across enough sites and seasons for registries and buyers to treat the system as proven rather than promising. That work is underway now; watch the public mMRV results, not the press releases.
Ocean alkalinity enhancement is a leading candidate for durable ocean carbon removal. How it works, what it might store, and the questions still open.
Dr. Lena Fischer ·
Beyond mangroves and kelp there is a fast-growing field of ocean carbon dioxide removal. Here is a clear map of the methods, their evidence, and their risks.
Priya Raman ·
Measurement, Reporting and Verification stripped to plain language: where MRV came from, what each step demands, and how to read the quality of any carbon project.
Priya Raman ·