1

Introduction

1.1 HUMAN PERTURBATIONS TO THE GLOBAL CARBON CYCLE

As of 2021, atmospheric carbon dioxide (CO₂) levels reached historically unprecedented levels, nearly 50 percent higher than preindustrial values only two centuries ago (Figure 1.1; GML, 2021). Present atmospheric CO₂ levels are higher than at any time in the past 800,000 years and likely several million years. The cause of the CO₂ increase over the 19th, 20th, and early 21st centuries is clearly and incontrovertibly identified as human activities including fossil-fuel burning, agriculture, and historical land-use change including deforestation (Figure 1.2; Friedlingstein et al., 2020). Human CO₂ emissions over the most recent decade were close to 35 billion tons of CO₂ per year (10⁹ t CO₂/yr = 1 Gt CO₂/yr).¹ The atmospheric accumulation of CO₂ would be even higher if not for land and ocean carbon sinks that together currently remove more than half the amount of human emissions from the atmosphere; however, these natural sinks may become less effective in a future high-CO₂, warmer world. The current level of human emissions greatly exceeds the ability of nature to remove CO₂, and a reduction on the order of 90 percent in human emissions is required to stabilize atmospheric CO₂ at some specified level, and approximately net-zero human CO₂ emissions is needed to stabilize climate because of inertia in the Earth system. In a number of climate scenarios, a period of net negative human CO₂ emissions occurs later in this century to compensate for other greenhouse gases (GHGs) and to address overshoots in atmospheric CO₂. An overall climate change response strategy will include climate mitigation approaches to reduce CO₂ and other GHG emissions. Carbon dioxide removal (CDR) approaches (Box 1.1) could be used in conjunction with emissions abatement to compensate for positive human emissions of CO₂ or

___________________

¹ For consistency, the unit metric tons (Mt) of CO₂ is used through most of this report when referring to CO₂ removal from the atmosphere. One billion Mt of CO₂ is equivalent to 0.128 parts per million (ppm) in global average atmosphere CO₂ mixing ratio, where, for reference, the 2021 CO₂ mixing ratio is approximately 415 ppm. But the reader should be aware that some sources in literature, policy documents, and press releases use a range of different units (e.g., 1 Gt CO₂ = 1015 g CO₂ = 1 Pg CO₂), and sometimes mass of carbon, C, is used rather than mass of CO₂ (1 Gt CO₂ = 0.273 Gt C). Ocean biological and geophysical carbon fluxes and stocks, for example, are often reported in carbon units rather than CO₂ units, reflecting the multiple chemical forms of carbon in biomass pools and in inorganic carbon dissolved in seawater.

contribute to net negative CO₂ emissions; this would require the durable storage of the removed carbon in some reservoir(s) away from the atmosphere for a sufficiently long period of time, typically taken as decades to centuries.

The rising level of atmospheric CO₂ is a major global concern because CO₂ is a key heat trapping gas, or GHG (USGCRP, 2017; IPCC, 2021), and elevated CO₂ levels are a major factor driving observed anthropogenic climate change that has already increased global average surface temperature by 1.12°C from preindustrial levels (NCEI, 2020) as seen in Figure 1.3. Although CO₂ is only one of several GHGs, contributing about 74 percent of the present total radiative imbalance leading to global warming (WRI, 2020; see also National Oceanic and Atmospheric Administration [NOAA] Earth System Research Laboratories²), the contribution of CO₂ to overall anthropogenic warming will likely grow in the future because of the long lifetime of excess CO₂ in the atmosphere (multidecadal and longer), ocean, and land biosphere system, compared with other GHGs (e.g., methane)³ (IPCC, 2021). While this report concentrates on removal of excess atmospheric CO₂, comprehensive climate mitigation strategies incorporate approaches to reduce human emissions of all GHGs and may even explore deliberate removal of gas beyond CO₂ such as methane.

The many impacts of climate change on managed and natural ecosystems and across human society are well documented in the scientific literature and in national and international assessment reports (e.g., USGCRP, 2018). In addition to climate change and associated ocean warming (Laufkötter et al., 2020) and decline in subsurface oxygen levels (Breitburg et al., 2018), marine ecosystems are also experiencing changes in seawater chemistry, termed ocean acidification, associated with the ocean uptake of excess CO₂ (Pershing et al., 2018). A wide range of marine organisms including shellfish and corals appear to be sensitive to ocean acidification, and the impacts extend to coastal human communities reliant on marine resources such as wild-caught fisheries, aquaculture, and marine tourism and recreation (Doney et al., 2020). Ocean acidification and climate change are closely linked because of a common underlying causal factor: large human CO₂ emissions to the atmosphere. CDR approaches for climate mitigation may also improve ocean acidification conditions, at least for some parts of the surface ocean.

The ocean holds great potential for uptake and longer-term sequestration of anthropogenic CO₂ for several reasons: (1) the ocean acts as a large natural reservoir for CO₂, holding roughly

___________________

² See https://www.esrl.noaa.gov/.

³ This statement should not be taken to suggest that other GHG emissions are unimportant or do not need to be reduced.

50 times as much inorganic carbon as the preindustrial atmosphere; (2) the ocean already removes a substantial fraction of the excess atmospheric CO₂ from human emissions; and (3) a number of physical, geochemical, and biological processes are known to influence air–sea CO₂ gas exchange and ocean carbon storage. The ocean covers 70 percent of Earth’s surface; it includes much of the global capacity for natural carbon sequestration, and it may be possible to enhance that capacity through implementation of an ocean-based CDR strategy.

1.2 CLIMATE MITIGATION, DECARBONIZATION, AND CARBON DIOXIDE REMOVAL

Without deliberate action to reduce human CO₂ emissions, continued rapid CO₂ accumulation in the atmosphere increases the expected magnitude of climate change and ocean acidification (UNEP, 2017; IPCC, 2021). Therefore, mitigation efforts to curb future climate change focus heavily on reducing the emissions of CO₂ (and other GHGs), which could potentially utilize CDR. On timescales of a few years, the atmosphere is relatively well mixed, and the global mean atmospheric CO₂ trend effectively reflects global net CO₂ fluxes (sources minus sinks). Reducing CO₂ emissions through abatement and removing CO₂ via CDR, therefore, are approximately equivalent on these timescales from the perspective of the atmospheric CO₂ budget, especially for the next several decades when emissions will remain well above net zero. The symmetry in the response of atmospheric CO₂ to emissions and removals may break down if there is a long delay (decades) in implementing CDR, such as may occur in atmospheric CO₂ overshoot scenarios, where land and ocean carbon processes may decrease the effectiveness of CDR (Zickfeld et al., 2021).

A climate target is needed to frame the amount and timing for emissions abatement and CDR. The international Paris Climate Agreement provides one such framing used widely in Intergovernmental Panel on Climate Change (IPCC) reports and the scientific literature. The Paris Agreement calls for limiting global warming to well below 2°C, preferably to 1.5°C, compared to preindustrial levels. In model simulations, meeting the Paris target requires reaching net-zero CO₂ emissions

globally well before the end of the century, with earlier mid-century net-zero targets for developed nations (NRC, 2015a; NASEM, 2019). Multiple abatement pathways will need to be pursued to rapidly decarbonize the U.S. and global economies including efforts to increase energy efficiency, switch to non-CO₂–emitting energy sources including renewables, capturing and sequestering CO₂ from point sources in geological reservoirs, and reducing CO₂ (and other GHG) emissions from land use (NASEM, 2021a).

While reaching net zero from current CO₂ emissions is primarily an effort in decarbonization and emissions abatement, essentially all of the climate mitigation scenarios that meet the Paris Agreement target include some form and amount of CDR (also commonly referred to as negative emission technologies or NETs) to balance residual CO₂ emissions from difficult-to-decarbonize sectors of economies (e.g., long-distance transportation, cement and steel production), to provide time for the development and implementation of different decarbonization approaches, and to compensate for short-term overshoots in emissions and atmospheric CO₂ and GHG levels. The reliance on, or even need for, CDR approaches varies considerably depending on the climate and socioeconomic scenarios and integrated assessment models used to evaluate human emissions and barriers to decarbonization. However, some form of CDR is common in climate/socioeconomic scenarios that attempt to limit global warming to well below 2°C, preferably to 1.5°C, compared to preindustrial levels in line with the international Paris Agreement (Figure 1.4; Allen et al., 2018; Fuhrman et al., 2019; Canadell et al., 2021; IPCC, 2021). The remaining carbon budget, or allowable cumulative future net human CO₂ emissions, is a particularly useful framing for assessing requirements for decarbonization and CDR. At current human CO₂ emission rates, the remaining carbon budgets would be expended in a little more than a decade to a few decades to stay below the Paris Agreement climate targets. The low-climate-warming scenarios used in the IPCC (2021) report reflect integrated assessment model simulations that often require substantial CDR, on the order of 10 Gt CO₂/yr or more by mid- to late-century, to stay within the remaining carbon budget constraints or to address atmospheric CO₂ overshoot where additional CDR is required to generate periods of net-negative human CO₂ later this century (Fricko et al., 2017; Riahi et al., 2017).

Historical and future global warming levels from now until mid-century (2050) approximately scale with cumulative CO₂ emissions, and CO₂ emissions reduction and CDR targets can be framed in terms of the remaining carbon budgets to meet specified warming targets (IPCC, 2021). For example, the estimated remaining carbon budget for a 1.5°C target is only about 300–900 Gt CO₂, a relatively small amount compared to current emissions of about 35 Gt CO₂/yr and the historical (1850 to 2019) cumulative human CO₂ emissions of 2,390 ± 240 Gt CO₂ (IPCC, 2021). Aggressive efforts to reduce emissions of methane and other GHGs would expand the remaining carbon budget but would need to be done in concert with decarbonization and CDR to stabilize global climate. This point is brought home well in a quote from the IPCC AR6 report:

There is considerable uncertainty, however, in the cumulative CDR requirement as noted in the IPCC (2018) Global Warming of 1.5°C report, where on the order of 100–1,000 Gt CO₂ of CDR over the 21st century is projected to meet the 1.5°C target with limited or no overshoot.

As illustrated in Figure 1.4, the potential scale for peak CDR demand varies greatly from zero to a few Gt CO₂/yr up to tens of Gt CO₂/yr. The range in CDR demand reflects the climate warm-

ing target (a lower 1.5°C target results in more CDR demand than less aggressive warming targets such as 2° or 3°C), as well as assumptions about socioeconomic pathways, climate sensitivity, and the response of land and ocean carbon sinks to climate change. Even larger amounts of CDR would be required if the goal is not simply climate stabilization but rather to shift climate back toward preindustrial conditions as has been proposed by some as an even more challenging and not well-agreed-upon possible future objective to occur after climate stabilization has been achieved. The required annual scale of CDR is thus comparable to the amount of CO₂ that is absorbed by the global ocean currently, ~9 Gt CO₂/yr (Friedlingstein et al., 2020). Significantly, the required CDR scale is a substantial fraction of current fossil fuel CO₂ emissions, ~35 Gt CO₂/yr (Friedlingstein et al., 2020) and is also larger than the largest manufacturing industries that is cement production (Andrew, 2018; Cao et al., 2020; IEA, 2020). It is important, however, to keep in mind that these CDR estimates reflect model estimates using only a subset of possible approaches, typically land-based afforestation/reforestation and bioenergy with carbon capture and sequestration, and adding CDR approaches with different land, resource, energy, and cost constraints can result in a different estimates of total CDR, emissions abatement, and residual human CO₂ emissions for the same climate target (e.g., recent studies adding direct air capture CDR, Fuhrman et al., 2020). Studies highlight the challenge of accomplishing the projected level of CDR with land-based approaches alone, particularly for reforestation and bioenergy with carbon capture and sequestration CDR

methods that require substantial amounts of land with impacts on food and water supplies, energy use, and fertilizer demand (Fuhrman et al., 2020). A CDR portfolio approach that includes less land-intensive methods, such as direct air capture, (Fuhrman et al., 2021) along with ocean-based methods may be more appropriate.

A CDR objective of removing and durably storing tens of Gt CO₂/yr by mid-century will likely be quite challenging to achieve. Today, global CO₂ sequestration activities accomplish <0.1 Gt CO₂/yr storage (Liu et al., 2018; Page et al., 2020; Townsend and Gillepsie, 2020). IEA (2021a) estimates that the global capacity of carbon capture, utilization, and storage facilities for CO₂ capture in 2020 was only about 40 Mt CO₂/yr. At present, industrial-scale CDR approaches are even smaller scale than pilot and demonstration plants; for example, a new direct air capture facility in Iceland would, at fully planned capacity, remove 4 kt CO₂/yr (Gertner, 2021). While ocean-based CDR is an important direction for achieving large-scale CDR, the scale of the challenge is daunting. For context, consider the scale of one of the largest manufacturing sectors, the cement industry, which produces around 4.5 Gt/yr of cement clinker, resulting in the downstream production of more than 20 Gt/yr of concrete (i.e., a mixture of cement, stone, sand, and water). The required rapid ramp-up of CDR scale implied by integrated assessment model scenarios involves the creation of a new sector, de novo, that is of a size similar to the cement/concrete sector, albeit in 30 years. The construction and commissioning of large capital facilities needed for many CDR approaches would require time, even with expected advances along technology learning curves. For example, to plan, permit, build, and commission a cement plant that produces ~1 Mt/yr of cement clinker requires at least on the order of 4 years. Growing policy or market demand for CDR, alone, does not guarantee success in reaching adequate scale for CDR. Substantial investments likely would be needed at multiple stages of innovation for a technology, as described in Nemet et al. (2018), addressing factors on both the supply side (e.g., research and development, demonstrations, and scale-up) and the demand side (e.g., demand pull, niche markets, and public acceptance). For this sectoral CDR scaling, it is furthermore important to consider the operational cost reductions that could be achieved, in time, and the amount of capital investment that is needed to stand up a CDR industry. Investments in CDR sectors will also depend on overall demand and price for CDR.

End-to-end carbon removal, ranging from carbon capture to geological sequestration (e.g., from point sources or the atmosphere), currently costs in the vicinity of ~$70 to $700/t CO₂. It has been suggested that it is necessary that carbon removal from the atmosphere be achieved at a net cost less than $100 (net present value based on 2021 dollars) (Budinis et al., 2018; Pilorgé et al., 2020; IEA, 2021b). The NASEM (2019) CDR report adopted a value of <$100/t CO₂ as a rough guide for “economical” CDR approaches, and for consistency this report uses the same cutoff, acknowledging that more expensive CDR approaches may also be considered because of other factors such as low resource demands and co-benefits. The requisite level of cost reduction, beyond technology improvements, requires a variety of actions including achieving (1) economies of scale, (2) massive replicability in manufacturing and deploying technology components and capital assets, and (3) the use of abundant, cost-effective, and accessible materials and components in technological systems. What remains prerequisite, however, are clear and consistent approaches for environmental and construction permitting, which limits overhead costs and restricts escalations of the overnight cost of construction.

Yet another important aspect related to the deployment of CDR technologies, however, implies the integration of carbon management solutions with existing industrial operations. For ocean-based CDR this could include coupling with marine aquaculture, shipping, and transportation systems, and connecting with and learning from coastal industrial operations such as desalination and chemical production. Some shore-based industrial facilities, for example, are already supplied with seawater intakes that could be utilized in some ocean-based CDR approaches; these intakes can account for a substantial fraction of the cost of capital construction. Such synergistic integra-

tion of engineering with existing technologies (e.g., particularly for electrochemical approaches) may provide opportunities to accelerate demonstration projects and scaling up of some ocean-based CDR technologies. Further, some electrochemical CDR approaches produce hydrogen gas as a byproduct, and assuming that the initial electrical energy source for the CDR comes from low-CO₂ sources, the hydrogen gas could be used as a clean fuel (or temporary energy storage) to reduce overall costs and carbon intensity of the CDR approach or other industrial activity.

1.3 SEAWATER CO₂ AND CARBONATE SYSTEM CHEMISTRY

The addition or removal of CO₂ from the ocean alters the acid-base chemistry of seawater. CO₂ gas dissolved in seawater, aqueous CO₂ (CO₂ (aq)), can react with water to form carbonic acid (H₂CO₃), a weak acid:

CO₂ (aq) + H₂O ↔ H₂CO₃

The partial pressure of CO₂ gas, pCO₂, varies proportionally with the concentration of CO₂ (aq) and is also influenced by temperature and salinity. The hydration reaction of CO₂ is relatively rapid and for most applications can be assumed to be at equilibrium. At seawater pH (~8), H₂CO₃ decomposes into bicarbonate (HCO₃⁻), an inorganic carbon ion, and a hydrogen ion (H⁺):

H₂CO₃ ↔ HCO₃⁻ + H⁺

Similarly, a bicarbonate ion can decompose into a carbonate ion (CO₃²⁻):

HCO₃⁻ ↔ CO₃²⁻ + H⁺

The seawater inorganic carbon system acid-base reactions are also in equilibrium as a function of temperature, salinity, and pressure. The dissolved inorganic carbon (DIC) is the sum of aqueous CO₂, H₂CO₃, HCO₃⁻, and CO₃²⁻, with HCO₃⁻ dominating at seawater pH. The addition of CO₂ increases seawater DIC, and the resulting production of H⁺ ions increases the acidity (lowers the pH) of seawater, where pH is defined on a logarithmic scale from the H⁺ ion concentration, pH = −log₁₀ [H⁺]. The seawater concentration of DIC is much higher than that in freshwater because of the high seawater alkalinity, a measure of the acid buffering capacity and a reflection of the balance of inorganic ions from rock weathering and other processes.

The addition of CO₂ gas to seawater from either physical or biological processes (e.g., respiration of organic matter) increases the DIC concentration but does not affect alkalinity. Because the CO₂ hydration reaction produces a weak acid H₂CO₃, CO₂ addition also makes the seawater more acidic (lowers pH) and shifts the partitioning of the inorganic carbon ions that make up DIC, increasing CO₂ and HCO₃⁻ and lowering CO₃²⁻. Increasing seawater alkalinity, for example, by adding a base (e.g., sodium hydroxide [NaOH]), shifts the inorganic ion partitioning in the opposite sense, lowering CO₂ and increasing pH and CO₃²⁻. The dissolution of calcium carbonate (CaCO₃) minerals into seawater increases both DIC and alkalinity, with the alkalinity increasing to twice that of DIC, and also results in an increase in the pH and CO₃²⁻.

Air–sea CO₂ flux is controlled thermodynamically by the difference between the partial pressure of CO₂ (pCO₂) between the surface ocean and atmosphere. Thus, the addition or removal of CO₂ and alkalinity can affect air–sea exchange by altering CO₂ (aq) and pCO₂. For example, the formation of organic matter by photosynthetic organisms in the surface ocean involves the uptake of CO₂, which acts to lower seawater pCO₂ and enhance the downward flux of CO₂ from the atmosphere into the surface ocean. The kinetics of air–sea CO₂ gas exchange are relatively slow because

the gas CO₂ (aq) is a small fraction of the large seawater DIC reservoir, and the inorganic carbon ions do not directly exchange with the atmosphere. Typical gas exchange equilibration timescales for the surface ocean are on the order of a year, and surface water pCO₂ can exhibit larger differences (disequilibrium) from atmospheric pCO₂.

1.4 OCEAN CARBON CYCLE AND OCEAN ANTHROPOGENIC CO₂ UPTAKE

The ocean geographic patterns and seasonal cycle of air–sea CO₂ vary substantially because of the interplay of biological and physical processes, resulting in ocean regions with both outgassing and ingassing from the atmosphere. Rising atmospheric CO₂ levels from human emissions shift the balance toward enhanced downward CO₂ flux from the atmosphere and ocean, and the anthropogenic CO₂ perturbation flux overlays the natural, preindustrial patterns. Globally the current net air–sea flux of anthropogenic CO₂ is roughly a quarter of emissions for fossil fuel consumption or about 9 Gt CO₂/yr (Friedlingstein et al., 2020), after accounting for the effects of river carbon inputs. The rate of anthropogenic CO₂ uptake is constrained by multiple observational approaches including global surveys of air–sea CO₂ flux, temporal changes in the ocean inventory of DIC, proxy methods based on other transient tracers such as chlorofluorocarbons, and numerical ocean models. The rate of ocean uptake of anthropogenic CO₂ is primarily controlled by physical circulation and the rate at which surface waters are exchanged into the thermocline and deep ocean.

The ocean uptake of anthropogenic CO₂ is a relatively small perturbation that occurs on top of the large natural background cycling and storage associated with the marine carbon system. The detection and quantification of anthropogenic CO₂ uptake because of rising atmospheric CO₂ has been a decades-long challenge for marine chemists and oceanographers, involving a combination of work to improve the accuracy and precision of seawater pCO₂, DIC, and alkalinity measurements and extensive field surveys of surface- and deep-ocean chemical and physical properties and their change over time. Detection and attribution of large-scale changes in ocean carbon storage due to ocean CDR approaches will have similar challenges. For comparison the cumulative ocean uptake of anthropogenic CO₂ from 1850 to 2019 is estimated to be about 591 Gt CO₂ or equivalently 161 Gt C (Friedlingstein et al., 2020), while the natural stock or reservoir of ocean inorganic carbon is about 38,000 Gt C (Figure 1.5). The large background ocean carbon inventory reflects a number of factors, in particular, the large seawater alkalinity that results in large DIC concentrations in equilibrium with any particular atmospheric CO₂ level and the transfer of carbon from the surface ocean–atmosphere reservoirs to the deep ocean by the long-term action of the biological carbon pump (Figure 1.6).

The biological pump consists of two complementary components: biological production of organic matter and the formation of biominerals (i.e., shells and skeletons made from CaCO₃). These processes in the surface ocean result in the surface uptake of DIC and export of carbon to depth via several physical and biological pathways (e.g., gravitational particle sinking, physical mixing of organic carbon into the mid-waters, and active biological transport via vertical migration). The majority of the exported organic carbon and biomineral CaCO₃ is respired (or remineralized for CaCO₃) back into DIC in the upper 1,000 meters of water column, and respiration and remineralization continue in the deep-water column and at the sediment surface, with only a small fraction (<1 percent for organic carbon, 10–20 percent for inorganic carbon; Andersson, 2014) buried in marine sediments. Ocean physical circulation then returns the respired and remineralized DIC and associated nutrients to the surface ocean on timescales of years to centuries.

Over time the biological pump acts to lower surface DIC and increase deep-ocean carbon storage, generating a vertical gradient of DIC in the ocean, until a steady state is reached where the downward flux of organic matter and carbonate is balanced by the upward flux of excess DIC (Sarmiento and Gruber, 2013). On the timescale of the ocean overturning circulation of centuries to

a millenia, the biological pump acts to maintain a much lower atmospheric CO₂ level than would occur in its absence. Variations in ocean circulation and the biological pump are also indicated as the most likely cause of large variations of ~100 ppm in atmospheric CO₂ documented in ice-core records for the past ~900,000. Cold glacial periods exhibited lower atmospheric CO₂ than warm interglacial periods including preindustrial conditions, with shifts between the two states occurring on timescales of 10,000 years or more. The effect of the biological pump on drawing down surface pCO₂ depends on the ratio of organic to CaCO₃ production and export because the effect of the alkalinity decline from CaCO₃ formation increases surface water pCO₂, opposing the effect of organic matter production and declining DIC. Ocean carbon storage is also sensitive to the fraction of sinking organic matter that reaches the deep sea prior to being respired (often termed the remineralization length scale) and to the extent of biological nutrient utilization in high-latitude surface ocean, especially in the Southern Ocean. The preindustrial ocean appears to have been in a steady state, with only small variations in atmospheric CO₂, for hundreds to thousands of years prior to the sharp growth of atmospheric CO₂ following the industrial revolution in the 1800s (see Figure 1.7).

From the perspective of anthropogenic CO₂ uptake by the ocean, the biological pump only contributes to the extent that it has been perturbed away from this steady state (Broecker, 1991), and even then many processes that may alter the rate of the biological pump have a relatively small impact on net ocean carbon storage because there is partial to nearly complete cancelation between the changes in the biological export flux and compensating changes in physical transport of DIC. For example, enhanced upwelling of nutrient-rich subsurface water can enhance biological productivity and export flux, but the physical upwelling supplying the extra nutrients also brings up excess

DIC; both the nutrients and excess DIC come from prior respiration of organic matter in the subsurface ocean. Substantial biological perturbations in global-scale ocean carbon storage require either shifts in the depth patterns of export and remineralization, alteration in the fraction of nutrients in the deep ocean supplied by respiration versus physical transport from the surface ocean, primarily the Southern Ocean with abundant surface macronutrients, or decoupling of carbon–nutrient relationships (Boyd and Doney, 2003; Sarmiento and Gruber, 2013). As documented in the 2021 IPCC report (Canadell et al., 2021), at least so far, there is only weak evidence of detectable changes in the global-scale ocean biological pump affecting net ocean carbon storage due to climate change or ocean acidification on a large scale and low confidence in our understanding of the magnitude and sign of ocean biological feedbacks to CO₂ storage and climate; the ocean uptake in anthropogenic CO₂ is attributed almost wholly to physicochemical processes and trends in human CO₂ emissions. However, given the large carbon fluxes associated with the biological carbon pump, with 5–12 Gt C/yr leaving the surface ocean annually (Siegel et al., 2014), relatively small variations in the function of the biological pump (e.g., carbon-to-nutrient ratios in organic matter; organic-to-inorganic carbon ratios; fraction of organic matter reaching the deep sea) have the potential to modify the vertical partitioning of ocean DIC, ocean carbon storage, and atmospheric CO₂ level.

1.5 OCEAN-BASED CARBON DIOXIDE REMOVAL

The objective of any CDR approach is to remove excess CO₂ from the atmosphere and store or sequester this carbon in some other reservoir away from the atmosphere for some time period, typically decades or longer. For ocean CDR, the removal from the atmosphere is indirect via an enhancement of the downward air–sea flux of CO₂ from the atmosphere to the surface ocean. This can occur through a variety of mechanisms, depending on the particulars of the ocean CDR method, including increasing the alkalinity and thus the DIC holding capacity of surface seawater; removing CO₂ from seawater for storage in some nonmarine or geological reservoir and thus creating a CO₂ or pCO₂ deficit in surface waters; removing surface CO₂ by increasing the storage of organic carbon in biomass, detritus, and dissolved organic carbon pools; directly injecting CO₂ into the deep ocean; or enhancing the biological transport of organic carbon from the surface ocean to the deep sea. As detailed in subsequent chapters, current scientific understanding of these ocean CDR approaches is insufficient to inform societal decision making and also differs substantially across the range of possible approaches (Box 1.2).

The timescale of carbon sequestration driven by ocean CDR will depend on the location and form of the excess carbon. Relevant questions include the permanence of changes in seawater DIC holding capacity, timescales for conversion of excess organic matter back to DIC, water column physical transport pathways of excess DIC back to the surface, and the leakage rate to the water column or atmosphere of geological or sediment sequestration. Ocean circulation pathways and rates are key to sequestration timescales for CDR approaches that deposit carbon in the water column or at the seafloor. The ocean thermocline covering roughly the upper 1,000 meters of the water column exhibits relatively rapid ventilation timescales of years to a few decades, and carbon must be transported into the deep sea (depth > ~1,000 meters) to achieve century-long sequestration times (Siegel et al., 2021a). This can be problematic for CDR approaches that enhance the ocean biological carbon pump because typically only a small fraction of sinking organic matter passes 1,000 meters, the remainder being respired back to CO₂ in the upper ocean and thermocline.

Here we consider ocean CDR techniques with sufficiently long sequestration permanence or durability to contribute to the portfolio of climate mitigation approaches in development to reduce atmospheric CO₂ this century and beyond and to buy time in the short term for deployment of other mitigation approaches. Although there is no uniform agreement in the literature or policy discussions on a specific sequestration threshold, the committee focused on methods that could potentially deliver durable CO₂ sequestration on timescales of several decades to a century or longer. Sequestration shorter than a decade is likely too short to be an effective policy tool. No sequestration method is foolproof, and probabilistic approaches will be warranted, along with monitoring, to evaluate the expected risk of CO₂ release back to the surface ocean and atmosphere over time.

The efficacy of CDR methods typically is evaluated as the near-term (weeks to months) removal of CO₂ from the atmosphere. Simply shifting carbon from one ocean carbon reservoir to another or to a geological reservoir is insufficient for climate mitigation if one cannot demonstrate the actual removal of CO₂ from the atmosphere. On longer timescales, the effectiveness of any CDR technique, land- or ocean-based, depends on the response of the full Earth system that will tend to dampen the atmospheric CO₂ response (Canadell et al., 2021). Lowering atmospheric CO₂ by any form of CDR reduces the growth rate of atmospheric CO₂ and thus slows the physicochemical uptake of anthropogenic CO₂ by the ocean, even in some cases possibly causing a small CO₂ outgassing. The same is true for human CO₂ emissions, where the growth rate in the atmospheric CO₂ inventory is only slightly less than half of human emissions (the airborne fraction) because of land and ocean carbon sinks.

The durability of ocean-based CDR must also take into consideration the impacts of ongoing and future climate change and ocean acidification. Ocean acidification and elevated CO₂ reduce the buffer capacity of seawater, lowering the effectiveness of CO₂ uptake. Ocean warming, for example, is expected to decrease CO₂ solubility, increase vertical stratification in the ocean, and alter ocean circulation patterns and marine ecosystem dynamics. In model simulations, climate carbon-cycle feedbacks reduce ocean CO₂ uptake somewhat, but the dominant factor governing the magnitude of the ocean sink remains strongly dependent on the CO₂ emissions scenario.

Ocean CDR approaches must also be assessed against the consequences of no action. Without substantial decarbonization, emissions abatement, and potential options such as CDR, atmospheric CO₂ growth will continue unabated with associated rising impacts from climate change and ocean acidification. Marine ecosystems and ocean resources are vulnerable to both climate change and ocean acidification (Pershing et al., 2018), and these impacts should be considered when evaluating the environmental impacts of ocean CDR. Even excluding deliberate ocean CDR, the ocean will continue to act as a sink for anthropogenic CO₂ because of physicochemical uptake; this process will continue into the future even if other mitigation options stabilize atmospheric CO₂ at some elevated level. In fact, over time the ocean will naturally sequester a larger and larger fraction of anthropogenic CO₂ (Archer et al., 2009) with the rate controlled on decadal to millennial timescales by ocean physical circulation and overturning and on longer timescales by adjustments in the marine cycling of CaCO₃ and the rate of marine CaCO₃ sedimentation (Figure 1.8).

1.6 ORIGIN AND PURPOSE OF THE STUDY

In 2013, the Board on Atmospheric Sciences and Climate convened the Committee on Geoengineering Climate: Technical Evaluation and Discussion of Impacts. The committee produced two reports, one on CDR, Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration (NRC, 2015a), and a second on solar radiation management (NRC, 2015b). Since the publication of the CDR report, interest in developing strategies for carbon sequestration has increased in concert with the increasing recognition of the potential need to employ CDR to prevent the more dire consequences associated with past and current GHG emissions. For CDR approaches to meaning-

fully contribute to a portfolio of responses to climate change, they need to “occur at a truly massive scale” (NRC, 2015a). It will be challenging to develop technologies to remove significant amounts of carbon from the atmosphere at a scale and cost that can be adopted in time to meet global targets for limiting warming “to well below 2 degrees Celsius, while pursuing efforts to limit the increase to 1.5 degrees” (Conference of Parties to the U.N. Framework Convention on Climate Change, December 2015).

The NRC (2015a) CDR report advised that “if carbon dioxide removal technologies are to be viable, it is critical now to embark on a research program to lower the technical barriers to efficacy and affordability while remaining open to new ideas, approaches, and synergies.” In 2019, the National Academies published a report that advances this goal, Negative Emissions Technologies and Reliable Sequestration: A Research Agenda (NASEM, 2019). The study found that, to meet climate goals, some form of CDR will likely be needed to remove roughly 10 Gt CO₂/yr by mid-century and 20 Gt CO₂/yr by the end of the century. To help meet that goal, four land-based CDR approaches were ready for large-scale deployment: afforestation/reforestation, changes in forest management, uptake and storage by agricultural soils, and bioenergy with carbon capture and storage, based on the potential to remove carbon at costs below $100/t CO₂. The study included a detailed research agenda to assess the benefits, risks, and sustainable scale potential for those four land-based approaches to CDR. The committee also examined approaches described as coastal blue carbon, limited to nearshore coastal land management strategies (e.g., seagrasses and wetlands), concluding that the potential for removing carbon is lower than other approaches but continued research is warranted to understand how future uptake of carbon may be affected by climate change and coastal management practices.

The 2019 report did not examine the more global ocean-based approaches but did recognize the potential for ocean-based CDR and the need for a research strategy to explore these options. To address this gap in understanding and the need for further exploration into CDR options that could feasibly contribute to a larger climate mitigation strategy, the National Academies convened the Committee on A Research Strategy for Ocean-Based Carbon Dioxide Removal and Sequestration. Specifically, this committee was assembled to develop a research agenda to assess the benefits, risks, and potential for responsible scale-up of a range of ecosystem-based and technological ocean-based CDR approaches. The committee’s Statement of Task is presented in Box 1.3. The six approaches in the Statement of Task, as defined by the study sponsor, are representative of the range of proposed ocean-based CDR approaches; they should not be taken as a comprehensive list of all ocean-based approaches. Additionally, this report does not repeat the work on blue carbon in vegetated coastal ecosystems covered in the 2019 National Academies report.

The intended audience for this report is wide ranging, including those interested in incorporating ocean-based CDR as part of a larger climate mitigation strategy. The committee’s task (Box 1.3) and focus was on identifying research and development needs within the ocean-based CDR space that could supply information to decision makers considering next steps involved in the scale-up of promising ocean-based CDR solutions.

Funding for the study came from the ClimateWorks Foundation, a nonprofit organization serving as a philanthropic platform for advancing climate solutions. As part of ClimateWorks Ocean CDR Portfolio, this task included examination of six groups of ocean-based CDR approaches, to identify key scientific and technological questions, including questions surrounding governance and societal dimensions that could increase the viability of responsible use of the ocean as a mechanism for carbon removal from Earth’s atmosphere.

1.7 STUDY APPROACH AND FRAMEWORK FOR ASSESSMENT

The study is organized around the six groups of ocean-based CDR approaches identified in the Statement of Task: nutrient fertilization, artificial upwelling and downwelling, seaweed cultivation, ecosystem recovery, alkalinity enhancement, and electrochemical approaches, illustrated in Figure 1.10. Chapter 2 of the report covers a series of crosscutting issues—legal, regulatory, and governance issues, social dimensions and justice considerations, and economic and funding considerations—foundational to all ocean-based CDR approaches. Chapter 2 also includes a subsection on common ocean monitoring requirements that will be needed for both CDR verification and assessment of environmental impacts. Chapters 3–8 then document the six ocean-based CDR groups followed by a synthesis chapter (Chapter 9):

• Nutrient fertilization (Chapter 3): Addition of micronutrients (e.g., iron) and/or macronutrients (e.g., phosphorus or nitrogen) to the surface ocean may in some settings increase photosynthesis by marine phytoplankton, and can thus enhance uptake of CO₂ and transfer of organic carbon to the deep sea where it can be sequestered for timescales of a century or longer. As such, nutrient fertilization essentially locally enhances the natural ocean biological carbon pump using energy from the sun, and in case of iron, relatively small amounts of iron are needed.
• Artificial upwelling and downwelling (Chapter 4): A process where water from depths generally cooler and more nutrient and carbon dioxide rich than surface waters is pumped into the surface ocean. Artificial upwelling has been suggested as a means to increase localized primary production and ultimately export production and net CDR. Artificial downwelling is the downward transport of surface water; this activity has been suggested as a mechanism to counteract eutrophication and hypoxia in coastal regions by increasing ventilation below the pycnocline and as a means to carry carbon into the deep ocean.
• Seaweed cultivation (Chapter 5): The process of producing macrophyte organic carbon biomass via photosynthesis and transporting that carbon into a carbon reservoir removes CO₂ from the upper ocean. Large-scale farming of macrophytes (seaweed) can act as a CDR approach by transporting organic carbon to the deep sea or into sediments.
• Recovery of ocean and coastal ecosystems (Chapter 6): CDR and sequestration through protection and restoration of coastal ecosystems, such as kelp forests and free-floating Sargassum, and the recovery of fishes, whales, and other animals in the oceans.
• Ocean alkalinity enhancement (OAE) (Chapter 7): Chemical alteration of seawater chemistry via addition of alkalinity through various mechanisms including enhanced mineral weathering and electrochemical or thermal reactions releasing alkalinity to the ocean, with the ultimate aim of removing CO₂ from the atmosphere.
• Electrochemical approaches (Chapter 8): Removal of CO₂ or enhancement of the storage capacity of CO₂ in seawater (e.g., in the form of ions or mineral carbonates) by enhancing its acidity or alkalinity, respectively. These approaches exploit the pH-dependent solubility of CO₂ by passage of an electric current through water, which by inducing water splitting (“electrolysis”) changes its pH in a confined reaction environment. As one example, OAE may be accomplished by electrochemical approaches.

For clarity in the report, we refer to carbon dioxide removal or CDR as intentional efforts to remove CO₂ from the atmosphere and store or sequester that carbon in some reservoir isolated from the atmosphere for some extended period of time, typically multiple decades or longer. Natural biotic and abiotic processes also act to sequester carbon away from the atmosphere—the atmosphere holds only a fraction of the amount of carbon in the ocean and land biosphere, let alone

more slowly evolving geological reservoirs. Where possible, we attempt to keep distinct deliberate human actions that enhance carbon storage away from the atmosphere; they are often perturbations on the much larger natural carbon fluxes and storage reservoirs.

The committee used a variety of information sources to inform and enrich deliberations and conduct their assessment, including a review of the scientific literature and a series of public meetings held in the virtual setting including four workshops and two additional public sessions. More than 65 experts from academic, governmental, and nongovernmental communities (see Appendix B for a list of experts invited to speak to the committee) were invited to present to the committee to assist the committee in better understanding stakeholder interest and exploring the current state of knowledge, potential, and limitations of ocean-CDR approaches. Workshop and meeting programs were developed to encourage discussion from diverse perspectives on ocean CDR feasibility and included presentations, made publically available, as well as moderated panel discussions incorporating questions from the committee and the online audience.

Each of the six groups of ocean-based CDR approaches was evaluated against a common set of criteria, where feasible. The criteria were developed by the committee based on specific elements included in the Statement of Task and from a review of previous planning and synthesis documents on CDR (e.g., NRC, 2015a; GESAMP, 2019; NASEM, 2019). The criteria were also used as prompts for invited speakers for the committee’s public sessions. These criteria together inform discussion on the viability (or feasibility) of responsible CDR and sequestration as highlighted in the Statement of Task. The criteria investigated include the following:

Knowledge base: What is the current state of scientific and technical understanding and readiness? How much of current understanding is based on theory and models and laboratory-scale experiments versus contained field experiments (i.e., mesocosms or similar approaches) and uncontained ocean perturbation experiments? What are the main knowledge gaps, and what are the uncertainties and/or confidence in this knowledge?

Efficacy: Can effective CDR from the atmosphere be demonstrated? Does the approach meet additionality? That is, on the system-level scale, what is the expected net CDR from the atmosphere and are there any compensating climate mitigation effects such as release of other GHGs? What, if any, downstream effects will occur and how does this influence efficacy?

Durability or permanence: Where is the excess carbon stored? On what timescale(s) will the carbon be released back into the atmosphere? What are the risk factors, both natural and social, associated with CO₂ release? Until widely accepted methods are developed to equate varied durability terms, longer and more durable storage terms have greater value.

Monitoring and verification: What are the monitoring and verification activities needed to quantify CDR efficacy (carbon accounting of the CDR from atmosphere, the increase in carbon stored in some non-atmosphere reservoir, and timescale of loss of sequestered carbon back to atmosphere)? Similarly, what are the monitoring needs to identify environmental and social impacts? Are there potential synergies with other ocean and environmental/climate observing systems?

Scale: What is the potential scale of the CDR technique, in terms of annual CDR, at partial up to full deployment? Are there geographic constraints on efficacy and total scale? Is there information on the temporal ramp-up rate to deploy the approach at scale? To facilitate comparisons across methods, a nominal annual scale of 0.1 Gt CO₂/yr is used. While smaller than the possible total CDR demand of up to tens of Gt CO₂/yr, the nominal scale may be sufficient to contribute to a portfolio of CDR approaches.

Viability and barriers: What is the potential viability of the CDR approach for deployment, taking into consideration a full suite of technical, scientific, economic, safety, and sociopolitical factors? What are the possible environmental and social impacts of the CDR approach, considering both intended and unintended consequences? Are the impacts localized to the marine environment, or do they extend into coastal and terrestrial regions? Are there possible co-benefits of the CDR approach, or is the CDR approach a co-benefit for some other environmental or conservation goal? What are the costs of the CDR approach ($/t CO₂) including the CO₂ removal/sequestration and the monitoring and verification costs for carbon accounting and environmental or social impacts? What are the energy, resource, infrastructure, land, and ocean-space requirements for the CDR approach?

Governance and social dimensions: What is the governance landscape for research on and possible future deployment of the CDR approach? Here governance is defined broadly to mean the legal, policy, and social context in which activities relating to ocean-based CDR and sequestration take place. It encompasses the laws and rules applying to activities, as well as the policies, processes, and institutions by which decisions about activities are made, including the role of various stakeholders and the public in decision-making. What are the social dimensions and environmental justice issues associated with the CDR approach?

Research and Development (R&D) opportunities: What are the R&D opportunities for the CDR approach over the next decade with the objective that research investments in the near term should better inform societal decisions in the future about potential deployment or not of a CDR approach? How can CDR research programs be framed in terms of “responsible innovation,” defined as “taking care of the future through collective stewardship of science and innovation in the present” (Stilgoe et al., 2013)? Are there best practices for CDR research that include transparency, adequate monitoring (for accounting and for environmental and social impacts) that limit any potential negative impacts of the research, and include engagement of coastal communities and the public? Are there research opportunities for expanding knowledge by moving research from modeling and laboratory scale to carefully constructed field experiments? Will CDR research have co-benefits of improving ocean science understanding? What are the possible funding mechanisms for CDR research?