About the Marine CDR Tools

These Marine CDR tools are a joint project between CarbonPlan and [C]Worthy designed to help people understand how marine CDR works.

Both tools allow users to explore how differences in location and season affect how much carbon removal a marine CDR intervention can achieve over time. The first focuses on ocean alkalinity enhancement (OAE) — an approach that adds alkalinity to the surface ocean, increasing its ability to absorb CO₂ by improving chemical buffering capacity. The second focuses on direct ocean removal (DOR) — an approach that extracts CO₂ from seawater, creating space for the ocean to re-absorb atmospheric CO₂ to return to its original state, and storing or otherwise utilizing the extracted CO₂.

Both map tools are based on the same modeling framework, using a coarse-resolution global ocean circulation model from the Community Earth System Model. Briefly, we divided the surface ocean into 690 regions, and simulated an OAE or DOR intervention occurring in that region for 1 month (e.g., adding alkalinity or removing dissolved inorganic carbon at a rate of 10 mol m^–2yr^–1). We then ran the model for 15 years to simulate how this intervention led to atmospheric carbon removal.

You can use the tools to explore global patterns in carbon removal efficiency, or drill down to visualize how these interventions move through the ocean and result in carbon removal over time.

Map variables

The two mapping tools share many of the same variables.

Comparing OAE and DOR

Harmonizing efficiency metrics

We initially defined OAE efficiency (𝜂) as the amount of CO₂ removed from the atmosphere per unit of alkalinity added. However, to compare across DOR and OAE, we use a different efficiency metric (γ): the amount of CO₂ absorbed from the atmosphere per unit of removal potential generated by the intervention.

For DOR, the maximum potential removal is equivalent to the amount of CO₂ extracted from the surface ocean. For OAE, this comparison requires calculating the maximum removal potential associated with the simulated alkalinity additions. This is done as follows. Our original definition of efficiency, the amount of CO₂ removed from the atmosphere per unit of alkalinity added, can be written as

𝜂(t) = ∆DIC(t) / ∆Alk

where ∆Alk is the amount of alkalinity added to the ocean and ∆DIC(t) is the resulting change in the ocean carbon inventory over time. Under the condition of full re-equilibration, we have

𝜂_max = ∆DIC(t=∞) / ∆Alk

where ∆DIC(t=∞) is the amount of potential DIC uptake induced by the chemical effects of the alkalinity addition. Therefore, to compute an efficiency metric for OAE that is directly comparable to DOR, we divide 𝜂(t) by 𝜂_max, yielding

γ_OAE(t) = 𝜂(t) / 𝜂_max

To perform this conversion on the OAE dataset, we estimate 𝜂ₘₐₓ from the background ocean state, using the thermodynamic equations defining the seawater carbonate system equilibrium chemistry. Notably, 𝜂ₘₐₓ varies from about 0.7 in the tropical surface ocean to about 0.9 at high latitudes. This approach is not perfect, as we are using time-averaged fields and mixing along the trajectory of the alkalinity plume entrains waters with distinct chemical composition. In practice, however, we find that the errors introduced by these factors are very small (less than ±0.1) relative to the large-scale spatial and seasonal structure in the efficiency fields, so the approach provides a robust basis to compare the efficiency of OAE and DOR.

Choosing system boundaries

The goal of these tools is to illustrate how the efficiency of OAE and DOR interventions are constrained by ocean processes. But the ocean response alone doesn't capture the full impact of these interventions on the carbon cycle—and interpreting and comparing the reported efficiency values requires care.

First, it's important to emphasize that neither tool accounts for the upstream emissions associated with energy and material use necessary to effect the intervention. These upstream process emissions are critical to understanding the ultimate climate impact of these interventions.

Second, the DOR and OAE tools reflect different choices about system boundaries. In the DOR tool, we chose to include a “storage loss” slider to let users explore how efficiency depends on the fate of the stream of extracted CO₂ — whether it is stored completely, or returns to the atmosphere through unintentional leakage from storage systems or short-lived utilization pathways. We felt this was important for illustrating how DOR’s climate impact depends not just on ocean reuptake dynamics, but also on the fate of the extracted CO₂. Indeed, in some areas of the ocean, extracting CO₂ for utilization could amount to a net-emission.

However, this choice creates a potential asymmetry with the OAE tool. Some electrochemical OAE approaches generate alkalinity by removing acid from seawater, enabling atmospheric CO₂ uptake but generating an acid waste stream. If that acid is later returned to the ocean or requires energy inputs to handle and neutralize elsewhere, it can reduce or even negate the intervention’s effectiveness. Mineral-based OAE approaches do not share this issue. This inefficiency, while somewhat analogous to DOR’s storage losses, is not represented in the OAE tool. Because acid handling has an indirect connection with ocean carbon stores — and is specific to certain OAE pathways — we treated it as a process emission and excluded it from the system boundary explored in the tool.

Another consideration is that we do not treat potential inefficiencies such as mineral precipitation (e.g., brucite, calcite, etc.) or incomplete dissolution of mineral-based alkalinity treatments. We do not include these effects in the tool because the phenomena would be confined to the near-field area in the direct vicinity of the intervention and affect the amount of alkalinity actually added to (or removed from) the ocean. The OAE dataset can be interpreted to show the response to net alkalinity addition, inclusive of any potential near-field inefficiencies. Note that DOR interventions result in elevated pH levels and mineral saturation states in the near-field as well; these could yield mineral precipitation and hence concomitant changes in alkalinity associated with DIC removal — again, we do not treat these effects in the DOR tool.

Version history

Additional resources

Acknowledgements

CarbonPlan and [C]Worthy received funding from the Carbon to Sea Initiative (via Windward Fund) and the Environmental Defense Fund for the OAE tool and its explainer. The DOR dataset generation, explainer article, and interactive tool development were supported by grants to [C]Worthy from the Chan Zuckerberg Initiative, the Navigation Fund, and the Grantham Foundation, and funding to CarbonPlan from [C]Worthy.

This material is based on work by [C]Worthy and collaborators that was supported by the National Center for Atmospheric Research (NCAR), a major facility sponsored by the National Science Foundation (NSF) under Cooperative Agreement No. 1852977. Mengyang received support from the Advanced Study Program Graduate Visitor Program at the National Center for Atmospheric Research. The authors acknowledge high performance computing support from the Cheyenne supercomputer provided by NCAR's Computational and Information Systems Laboratory, sponsored by the NSF. This research also used resources of the National Energy Research Scientific Computing Center (NERSC), a Department of Energy Office of Science User Facility using NERSC award ALCC-ERCAP0034226.