carbon sunk in

the ocean, uncounted

Problem statement

Ocean-based carbon dioxide removal (mCDR) — including biomass sinking, ocean alkalinity enhancement, and artificial upwelling — lacks a reliable system for monitoring, reporting, and verifying (MRV) that carbon has been durably sequestered. Running Tide — one of the largest carbon-removal startups operating in the ocean, if not the largest — admitted before its 2024 shutdown that it could not observe sunk biomass for more than about three hours after release and could not distinguish its alkalinity signal from natural ocean variability. Without credible MRV, carbon credits from ocean-based approaches cannot be trusted, buyers cannot assess what they're purchasing, and the entire market mechanism fails.

Why this matters

The ocean absorbs roughly 25% of anthropogenic CO2 emissions (Global Carbon Budget 2023: an ocean sink of 2.8 GtC against total emissions of 11.1 GtC in 2022) and has vastly more sequestration capacity than terrestrial systems. Carbon removal at scale is not optional — the IPCC's AR6 WGIII assessment states that "the deployment of carbon dioxide removal (CDR) to counterbalance hard-to-abate residual emissions is unavoidable if net zero CO2 or GHG emissions are to be achieved," and names "development of agreed methods for measurement, reporting and verification of carbon flows" among the key enablers of upscaling. Buyers including Microsoft, Stripe, and Shopify have committed heavily to carbon removal purchases — the Stripe-run Frontier coalition alone is an advance market commitment to buy $1.8 billion of permanent carbon removal by 2040 (frontierclimate.com). But without verifiable measurement, ocean-based approaches remain scientifically unproven at scale. Running Tide's collapse — despite more than $50 million raised (including a $54 million Series B) and offtake deals with Microsoft and Shopify — demonstrates that the market will not sustain companies that cannot prove their climate impact.

What’s been tried and why it hasn’t worked

Running Tide sank wood biomass coated with limestone into deep ocean waters off Iceland. Their monitoring approach relied on short-term tracking of material as it descended, but signal was lost within hours. The company's carbon accounting relied on theoretical models of what happens to submerged biomass rather than empirical measurement. Climate economist Danny Cullenward criticized the company's short-cycle-to-long-cycle carbon framing as "a vast oversimplification that substantially overstates the net climate benefits when you apply that to ocean-based processes" (Latitude Media). Ship-based monitoring is prohibitively expensive for tracking dispersed material across thousands of square kilometers of ocean. Satellite remote sensing can detect surface changes but cannot observe processes in the deep ocean where sequestration occurs. Existing oceanographic sensor networks (Argo floats, moored buoys) are not designed to detect the chemical signatures of deliberate carbon sequestration against the background variability of ocean chemistry. The fundamental challenge is that the ocean is vast, opaque to most remote sensing, and chemically noisy.

What would unlock progress

A breakthrough in low-cost, persistent deep-ocean chemical sensing could transform marine CDR from speculative to verifiable. This might involve: autonomous underwater vehicles with carbon-isotope sensors that can distinguish anthropogenic from natural carbon signals; degradation-resistant tracer compounds that co-sink with biomass and can be detected months later; or distributed sensor networks (building on existing Argo float infrastructure) calibrated for alkalinity and dissolved inorganic carbon at the precision needed to detect CDR signals. Acoustic monitoring of biomass descent trajectories could extend the tracking window. Adjacent fields with relevant approaches include deep-sea mining environmental monitoring, submarine hydrothermal vent sensing, and radiocarbon tracing in paleoceanography.

Entry points for student teams

A student team could: (1) design and prototype a low-cost chemical tracer system that could be embedded in sinkable biomass and detected at depth weeks or months later, focusing on tracer selection and detection limits; (2) model the minimum sensor density and precision needed to distinguish a CDR alkalinity signal from natural ocean variability in a specific ocean region; (3) prototype an acoustic tracking system for monitoring biomass descent and dispersal in a controlled water column (tank or coastal environment). Relevant disciplines include oceanography, environmental engineering, sensor design, chemical engineering, and data science.