March 2023

Frontiers in Ocean Observing: Emerging Technologies for Understanding and Managing a Changing Ocean







2023 Oceanography Supplement

ON THIS PAGE. A Biogeochemical Argo float. For details, see

Boyd et al. (2023, in this supplement). Photo credit: David Luquet,

licensed under CC BY-SA 4.0 by Thomas Boniface

ON THE FRONT COVER. A depiction of some of the many

platforms that scientists use to sample and sense the ocean,

including (from left to right) moorings, autonomous underwater

vehicles, CTD rosettes, and towed vehicles (not to scale). See

Govindarajan et al. (2023, in this supplement) to read about how

these platforms can be used to sample environmental DNA for

studies of the mesopelagic ocean. Illustration by Natalie Renier,

Woods Hole Oceanographic Institution


• S. Kim Juniper, Ocean Networks Canada

• Sophie Seeyave, Partnership for Observation of the Global Ocean

• Emily Smith, NOAA’s Global Ocean Monitoring and Observing Program

• Martin Visbeck, GEOMAR Helmholtz Center for Ocean Research Kiel


• Ellen S. Kappel, Oceanography Editor

• Vicky Cullen, Oceanography Assistant Editor

• Johanna Adams, Layout & Design


Ocean-Climate Nexus: Observations for Marine Carbon Dioxide Removal

• Jens D. Müller, ETH Zürich

• Toste Tanhua, GEOMAR Helmholtz Center for Ocean Research Kiel

Ecosystems and Their Diversity: Patterns and Trends in Ocean

Biodiversity Under Climate Change

• Mark J. Costello, Nord University

• Qianshuo Zhao, Ocean University of China, Qingdao

• Charles Lavin, Nord University

• Cesc Gordó-Vilaseca, Nord University

Ocean Pollutants: Assessing the Damage Caused by Marine

Plastic Pollution

• Luisa Galgani, GEOMAR Helmholtz Center for Ocean Research Kiel

• Shiye Zhao, Japan Agency for Marine-Earth Science and Technology

Multi-Hazard Warning Systems: Ocean Observations for Coastal

Hazard Warning

• Soroush Kouhi, Ocean Networks Canada

• Benoît Pirenne, Ocean Networks Canada

Technology: Environmental DNA

• Annette Govindarajan, Woods Hole Oceanographic Institution

• Luke McCartin, Lehigh University


Support for this publication is provided by Ocean Networks Canada,

the National Oceanic and Atmospheric Administration’s Global Ocean

Monitoring and Observing Program, and the Partnership for Observation

of the Global Ocean.

This is an open access document made available under a Creative

Commons Attribution 4.0 International License, which permits use, shar-

ing, adaptation, distribution, and reproduction in any medium or format

as long as users cite the materials appropriately, provide a link to the

Creative Commons license, and indicate the changes that were made

to the original content. Users will need to obtain permission directly

from the license holder to reproduce images that are not included in the

Creative Commons license.

Published by The Oceanography Society

Single printed copies are available upon request from


Kappel, E.S., V. Cullen, M.J. Costello, L. Galgani, C. Gordó-Vilaseca,

A. Govindarajan, S. Kouhi, C. Lavin, L. McCartin, J.D. Müller,

B. Pirenne, T. Tanhua, Q. Zhao, and S. Zhao, eds. 2023. Frontiers in

Ocean Observing: Emerging Technologies for Understanding and

Managing a Changing Ocean. Oceanography 36(Supplement 1),



Introduction To Frontiers in Ocean Observing

by E.S. Kappel, M.J. Costello, L. Galgani, C. Gordó-Vilaseca, A. Govindarajan, S. Kouhi, C. Lavin, L. McCartin, J.D. Müller,

B. Pirenne, T. Tanhua, Q. Zhao, and S. Zhao



Operational Monitoring of Open-Ocean Carbon Dioxide Removal Deployments: Detection, Attribution,

and Determination of Side Effects

by P.W. Boyd, H. Claustre, L. Legendre, J.-P. Gattuso, and P.-Y. Le Traon


Assessing Changes in Marine Biogeochemical Processes Leading to Carbon Dioxide Removal with

Autonomous Underwater Vehicles

by C. Garcia, B. Barone, S. Ferrón, and D. Karl


A Low-Cost Carbon Dioxide Monitoring System for Coastal and Estuarine Sensor Networks

by P.J. Bresnahan, E. Farquhar, D. Portelli, M. Tydings, T. Wirth, and T. Martz


PERSPECTIVE. The Growing Potential of Antarctic Blue Carbon

by C.J. Sands, N. Zwerschke, N. Bax, D.K.A. Barnes, C. Moreau, R. Downey, B. Moreno, C. Held, and M. Paulsen


PERSPECTIVE. Net Zero: Actions for an Ocean-Climate Solution

by A.M. Waite, M. Smit, E. Siegel, G. Hanna, and S. Leslie




Using Soundscapes to Assess Changes in Coral Reef Social-Ecological Systems

by T.-H. Lin, F. Sinniger, S. Harii, and T. Akamatsu


Functional Seascapes: Understanding the Consequences of Hypoxia and Spatial Patterning in Pelagic Ecosystems

by S.B. Brandt, S.E. Kolesar, C.N. Glaspie, A. Laurent, C.E. Sellinger, J.J. Pierson, M.R. Roman, and W.C. Boicourt


Recent Marine Heatwaves Affect Marine Ecosystems from Plankton to Seabirds in the Northern Gulf of Alaska

by S. Strom and the Northern Gulf of Alaska Long-Term Ecosystem Research Team


Modeling Aquaculture Suitability in a Climate Change Future

by A.L. Mackintosh, G.G. Hill, M.J. Costello, A. Jueterbock, and J. Assis


Monitoring Algal Blooms with Complementary Sensors on Multiple Spatial and Temporal Scales

by D.R. Williamson, G.M. Fragoso, S. Majaneva, A. Dallolio, D.Ø. Halvorsen, O. Hasler, A.E. Oudijk, D.D. Langer, T.A. Johansen,

G. Johnsen, A. Stahl, M. Ludvigsen, and J.L. Garrett


UAV High-Resolution Imaging and Disease Surveys Combine to Quantify Climate-Related Decline in

Seagrass Meadows

by L.R. Aoki, B. Yang, O.J. Graham, C. Gomes, B. Rappazzo, T.L. Hawthorne, J.E. Duffy, and D. Harvell


Long-Term Observations of Hypoxia off the Yangtze River Estuary: Toward Prediction and Operational Application

by X. Ni, F. Zhou, D. Zeng, D. Li, T. Zhang, K. Wang, Y. Ma, Q. Meng, X. Ma, Q. Zhang, D. Huang, and J. Chen



Exploring Methods for Understanding and Quantifying Plastic-Derived Dissolved Organic Matter

by L. Zhu, N. Gaggelli, A. Boldrini, A. Stubbins, and S.A. Loiselle


On the Potential for Optical Detection of Microplastics in the Ocean

by D. Koestner, R. Foster, and A. El-Habashi


Sediment Traps: A Renowned Tool in Oceanography Applied to New Marine Pollutants

by L. Galgani, H. Hepach, K.W. Becker, and A. Engel


Developing Realistic Models for Assessing Marine Plastic Pollution in Semi-Enclosed Seas

by J. She, A. Christensen, F. Garaventa, U. Lips, J. Murawski, M. Ntoumas, and K. Tsiaras




Ocean Monitoring and Prediction Network for the Sustainable Development of the Gulf of Mexico and the Caribbean

by J.C. Herguera, E.M. Peters, J. Sheinbaum, P. Pérez-Brunius, V. Magar, E. Pallàs-Sanz, S. Estrada Allis,

M.L. Aguirre-Macedo, V.M. Vidal-Martinez, C. Enriquez, I. Mariño Tapia, H. García Nava, X. Flores Vidal, T. Salgado,

R. Romero-Centeno, J. Zavala-Hidalgo, E.A. Cuevas Flores, A. Uribe Martínez, and L. Carrillo


An Experimental Platform to Study Wind, Hydrodynamic, and Biochemical Conditions in the Littoral Zone During

Extreme Coastal Storms

by B.M. Phillips, F.J. Masters, B. Raubenheimer, M. Olabarrieta, E.S. Morrison, P.L. Fernández-Cabán, C.C. Ferraro, J.R. Davis,

T.A. Rawlinson, and M.B. Rodgers


Probabilistic Approaches to Coastal Risk Decision-Making Under Future Sea Level Projections

by T. Spencer, M. Dobson, E. Christie, R. Eyres, S. Manson, S. Downie, and A. Hibbert


The Texas A&M – University of Haifa Eastern Mediterranean Observatory: Monitoring the Eastern Mediterranean Sea

by G. Milton, S.F. DiMarco, A.H. Knap, J. Walpert, and R. Diamant


Subsea Cables as Enablers of a Next Generation Global Ocean Sensing System

by E. Pereira, M. Tieppo, J. Faria, D. Hart, P. Lermusiaux, and the K2D Project Team


Integrating Topographic and Bathymetric Data for High-Resolution Digital Elevation Modeling to Support Tsunami

Hazard Mapping

by C. Bosma, A. Shumlich, M. Rankin, S. Kouhi, and R. Amouzgar


Assessment of Tsunami Hazard Along British Columbia Coastlines from Coseismic Sources

by S. Kouhi, R. Amouzgar, M. Rankin, C. Bosma, and A. Shumlich


Detection of Landslides and Tsunamis in Douglas Channel and Gardner Canal, British Columbia

by F. Nemati, L. Leonard, G. Lintern, C. Brillon, A. Schaeffer, and R. Thomson


EASTMOC: Environmental Alert System for Timely Maintenance of the Coastal Zone

by V. Kondrat, I. Šakurova, E. Baltranaitė, and L. Kelpšaitė-Rimkienė



Advances in Environmental DNA Sampling for Observing Ocean Twilight Zone Animal Diversity

by A.F. Govindarajan, A. Adams, E. Allan, S. Herrera, A. Lavery, J. Llopiz, L. McCartin, D.R. Yoerger, and W. Zhang


Detecting Mediterranean White Sharks with Environmental DNA

by J.F. Jenrette, J.L. Jenrette, N.K. Truelove, S. Moro, N.I. Dunn, T.K. Chapple, A.J. Gallagher, C. Gambardella, R. Schallert,

B.D. Shea, D.J. Curnick, B.A. Block, and F. Ferretti


Toward Identifying the Critical Ecological Habitat of Larval Fishes: An Environmental DNA Window into

Fisheries Management

by E.V. Satterthwaite, A.E. Allen, R.H. Lampe, Z. Gold, A.R. Thompson, N. Bowlin, R. Swalethorp, K.D. Goodwin, E.L. Hazen,

S.J. Bograd, S.A. Matthews, and B.X. Semmens


The Use of eDNA to Monitor Pelagic Fish in Offshore Floating Wind Farms

by T.G. Dahlgren, J.T. Hestetun, and J. Ray


Deep-Sea Predator-Prey Dynamics Revealed by Biologging and eDNA Analysis

by V.J. Merten, F. Visser, and H.-J.T. Hoving

100 Evaluating Connectivity of Coastal Marine Habitats in the Gulf of Maine by Integrating Passive Acoustics

and Metabarcoding

by G. Milne, J. Miksis-Olds, A. Stasse, B.-Y. Lee, D. Wilford, and B. Brown

102 Authors

106 Acronyms

In this second supplement to Oceanography on frontiers

in ocean observing, articles describe the many creative

and promising ways in which scientists are now sampling

and studying the ocean and its constituents, from carbon

dioxide and oxygen, environmental DNA (eDNA), plastics,

and microplastics, to coral reefs, fish, and whole ecosys-

tems. These papers show, for example, how acoustic tech-

niques are critical components for early warning of natural

hazards such as earthquakes, tsunamis, and landslides

and how they play significant roles in investigations such

as determining how ocean soundscapes may be used to

monitor coral reef ecosystem health. Several articles

demonstrate the application of long-used technologies for

new and important data-gathering purposes, while others

describe how data collected by multiple technologies

deployed simultaneously have improved monitoring of

threats such as harmful algal blooms and oil spills. Some

authors detail the application of eDNA analyses, especially

to midwater environments, where they are providing new

insights, in combination with other sampling and sensing

methods. Modeling is a key aspect of several of the stud-

ies, where ocean observing data serve as critical inputs

and for validation. These innovative observing technolo-

gies and analytical techniques are advancing our under-

standing of the world ocean and supporting its sustainable

use and management.

Similar to last year’s supplement, we invited potential

authors to submit letters of interest aligned with the pri-

orities of the UN Decade of Ocean Science for Sustainable

Development (2021–2030), though topics were further

narrowed into specific themes. In addition, this year each

topic had guest editors, some of whom are early career sci-

entists. The idea was that this supplement would provide

an opportunity for a senior scientist to mentor one from

the next generation and for early career scientists to gain

some experience as guest editors for a journal.


THEME: Observations for Marine Carbon Dioxide Removal

GUEST EDITORS: Toste Tanhua and Jens Daniel Müller

How ocean observations and emerging technologies

are helping to identify potential marine carbon dioxide

removal (mCDR) opportunities and to measure the effec-

tiveness of mCDR actions, including studies evaluating the

potential adverse effects of human interventions on bio-

diversity and ecosystem function. 


THEME: Patterns and Trends in Ocean Biodiversity Under

Climate Change

GUEST EDITORS: Mark John Costello, Qianshuo Zhao,

Charles Lavin, and Cesc Gordó-Vilaseca

Ocean observing efforts that record environmental and

related biodiversity changes occurring in different eco-

systems, from the coasts to the deep ocean and from

the tropics to the high latitudes, and that address the link

between observations and policy.


THEME: Assessing the Damage Caused by Marine

Plastic Pollution

GUEST EDITORS: Luisa Galgani and Shiye Zhao

Both classical oceanographic approaches and new tech-

niques for quantifying marine plastic pollution as well as

its long-term consequences on ecosystems and climate

through their interaction with natural elements of the

marine environment. 


THEME: Ocean Observations for Coastal Hazard Warning

GUEST EDITORS: Benoît Pirenne and Soroush Kouhi

Observations and monitoring, forecasting, alerting, and

hazard research together with the systems developed

around them and their applications in coastal communities. 


THEME: Environmental DNA Technology

GUEST EDITORS: Annette Govindarajan and Luke McCartin

Recent developments in all aspects of eDNA technology

and interpretive approaches relevant for observing and

studying animal biodiversity, especially in the ocean mid-

water environment.

We thank Ocean Networks Canada, the US National

Oceanic and Atmospheric Administration’s Global Ocean

Monitoring and Observing Program, and the Partnership

for Observation of the Global Ocean for generously sup-

porting publication of this supplement to Oceanography.



By Ellen S. Kappel, Mark John Costello, Luisa Galgani, Cesc Gordó-Vilaseca,

Annette Govindarajan, Soroush Kouhi, Charles Lavin, Luke McCartin, Jens Daniel Müller,

Benoît Pirenne, Toste Tanhua, Qianshuo Zhao, and Shiye Zhao


Observations for

Marine Carbon Dioxide


Operational Monitoring of Open-Ocean Carbon Dioxide Removal Deployments:

Detection, Attribution, and Determination of Side Effects

By Philip W. Boyd*, Hervé Claustre*, Louis Legendre*, Jean-Pierre Gattuso, and Pierre-Yves Le Traon (*equal first authors)

Human activities are causing a sustained increase in the

concentration of carbon dioxide (CO2) and other green-

house gases in the atmosphere. The resulting harmful

effects on Earth’s climate require decarbonizing the econ-

omy and, given the slow pace and inherent limitations of

decarbonization of some industries such as aviation, also

the active removal and safe sequestration of CO2 away

from the atmosphere (i.e., carbon dioxide removal or CDR;

NASEM, 2022). Limiting global warming to 1.5°C—a target

that may already have been exceeded—would require

CDR on the order of 100–1,000 Gt CO2 over the twenty- first

century (IPCC, 2018).

Natural terrestrial and ocean processes already remove

about half of human CO2 emissions from the atmosphere,

with half of this amount (i.e., a quarter of the total) ending

up in the ocean. These natural processes slow down global

warming; without the continuous removal of atmospheric

CO2 since the beginning of the Industrial Era (1750), the

present (2022) level of 420 ppm would have been reached

in the 1980s. In the ocean, CO2 combines with water (H2O)

to form dissolved inorganic carbon (DIC: CO2 gas, H2CO3,

and HCO3

– and CO3

2– ions), and photosynthetic organisms

use some of the DIC to synthesize the organic matter that

is the basis of pelagic marine food webs. Marine organic

carbon exists in both particulate and dissolved forms (POC

and DOC, respectively). A number of physical, chemical,

and biological processes, collectively called ocean carbon

pumps, transfer carbon from surface waters downward

and store it in the ocean as DIC and refractory (i.e., long-

lived) DOC and POC in the ocean. Some of this storage takes

place on climatically significant timescales and is called car-

bon sequestration. Sequestration of DIC can occur at any

depth, but its potential is higher at greater depths.

There is increasing discussion of implementing marine

CDR (mCDR) approaches, which range from methods

based on natural processes to more industrial tech-

niques (NASEM, 2022). Here, we focus on open- ocean

mCDR approaches, including alkalinization (i.e.,  adding

alkaline substances, such as olivine or lime, to seawater

to enhance the ocean’s chemical uptake of CO2 from the

atmosphere) and nutrient fertilization (i.e., adding a nutri-

ent that limits phytoplankton photosynthesis, such as iron,

to surface waters to enhance the photosynthetic uptake

of DIC), which aim to enhance DIC sequestration resulting

from increased CO2 influx from the atmosphere.

There is a growing body of literature on various aspects

of mCDR approaches. Published mCDR studies have

addressed the appropriateness of implementation, testing

the efficiency of sequestering CO2 and/or assessing det-

rimental ecological effects (laboratory/mesocosm studies,

field trials), and identifying potential deployment sites.

Such pilot studies are precursors to possible future mCDR

deployments (NASEM, 2022), which should only occur in

cases where the pilot studies indicate that mCDR would not

unduly disrupt marine ecosystems. In contrast, this paper

addresses the situation where mCDRs are to be deployed

at scales commensurate with the target of removing giga-

tons of atmospheric carbon.

Here, considering information from satellites and auton-

omous platforms combined with artificial intelligence (AI)

and models (Figure 1), we describe a future operational

monitoring system for the detection, attribution, and deter-

mination of side effects of open- ocean mCDR deployments.

We mainly address the monitoring challenge described

in NASEM (2022), based upon the current and expected

readiness of observational platforms and sensors. This

approach ensures that the proposed monitoring system

would be tractable and deployable. The assessment of

future mCDR deployments will include three components,

together referred to as MRV: measurement or monitoring

FIGURE 1. The three main components of a marine carbon dioxide removal (mCDR) monitor-

ing system are the tools to be used, field implementation, and reporting and verification. These

would interact, and thus progressively improve, prior to and during the long-term mCDR deploy-

ment. The development of tools would be accelerated by the urgency created by the mCDR

deployments. Connections between the three objectives of the monitoring system (i.e., detection,

attribution, and determination of side effects) and the monitored variables are described under

“Monitoring mCDR Deployments” in the text. OSSE = Observation System Simulation Experiment.

TRL = Technological Readiness Level. Licensed under CC BY-SA 4.0 by Thomas Boniface

(M) as described in this study, reporting (R) of the result-

ing data to a certified authority, and verification (V) by this

authority, using data and models, that any deployment is

successful at increasing CO2 influx from the atmosphere

and enhancing its sequestration in the ocean. Successful

verification of removal and sequestration will result in cer-

tification of the mCDR. The last two MRV components are

mentioned in the last section of the study.

Our study for this “Frontiers in Ocean Observing” sup-

plement of Oceanography focuses on the observational

aspects of the monitoring of open- ocean mCDR deploy-

ments, with less emphasis on the corresponding, essential

modeling components. We nevertheless briefly describe

the latter where necessary.


Monitoring is essential in order to quantify the effective-

ness (removal) and durability (sequestration) of carbon

storage resulting from open- ocean mCDR deployments

and to identify environmental impacts (NASEM, 2022).

Here, we examine three objectives of a future open- ocean

mCDR monitoring system. Our definitions of detection and

attribution are consistent with those in the Glossary of the

Intergovernmental Panel on Climate Change (IPCC, 2021).

Detection. To quantify the amount of carbon sequestered

as DIC. This will require quantification of metrics that docu-

ment both the amount of carbon removed, based on mod-

els that assimilate accurate in situ measurements of carbon

system variables, and the durability of its removal (i.e., long-

term [decadal] estimates of air- sea CO2 exchanges).

Attribution. To assign the detected carbon sequestra-

tion solely to a particular mCDR deployment. Attribution

requires an understanding of the processes that jointly

determine the success or failure of the given mCDR

deployment and must thus address the influence of com-

plex drivers in the carbon cycle to demonstrate additional-

ity (see next section). Attribution addresses the proportion

of carbon sequestration that can be attributed to an mCDR

deployment, even if there are contributions by other

drivers within the carbon cycle. This will involve advanced

modeling capabilities that simulate the state of the cou-

pled physical and biogeochemical ocean and its modifica-

tion by the mCDR deployment.

Determination of Side Effects. To identify and quantify

ecological impacts of the mCDR and ensure that they do

not exceed the impacts expected from the pilot studies.

The mCDR deployments will necessarily modify the ocean,

but intended and unintended ecological impacts are poorly

known. This will require monitoring ecological variables in

at least the upper 1,000 m as well as deeper, including at

the seafloor where benthic systems could be affected. Side

effects would be assessed through modeling studies of the

impacts of biogeochemical changes on marine ecosystems.

Unacceptable ecological side effects, acute or chronic, may

lead to the termination of an mCDR deployment.



To achieve these three objectives, model simulations will

be run, with and without mCDR. In order to fulfill this role,

simulations will have to capture many ongoing changes

in the ocean that include those due to climate change,

the hysteresis effects from climate change, the effects

of other mCDR (and terrestrial CDR), and the effects of

emissions reductions on the ocean carbon cycle. It will

require advanced modeling capabilities that could effec-

tively simulate the state of the coupled physical and bio-

geochemical ocean and its changes under the different

mCDR scenarios. Such models pose many scientific and

technological challenges that impede the development of

Digital Twins of the Ocean (DTO). The DTO will combine

next- generation ocean modeling, artificial intelligence, and

high- performance computing to create digital replicas of

the ocean that are regularly informed and improved with

observations. Some of the observations will be used in the

models, and others will be kept for model validation.

Extensive validation of these models and their improve-

ments (e.g.,  optimization of model parameters) will be

needed well in advance of and during mCDR deploy-

ments. Confirming that the simulations are consistent

with observations will require the initiation of monitoring

a long time before any mCDR deployment. During this pre-

deployment period, no mCDR could be undertaken in the

mCDR- intended region.

However, there may be cases where model validation

could not be undertaken well in advance of the mCDR

deployments, for example, where a deployment would

take place without prior, long- term consultation with the

authority to which monitoring would be reported. In such

a case, the results of model simulations run with and with-

out mCDR could be compared with observations made in

the mCDR- deployment region and one or more control

regions (with long- term time- series observations) where

conditions would be comparable to those in the selected

mCDR region and where no mCDR of any type would

be deployed. Of course, control regions will eventually

become “contaminated” by the spread of DIC—via ocean

circulation—from mCDR deployed elsewhere (e.g.,  Boyd

and Bressac, 2016), but they could be used for model vali-

dation before this occurs and even after.

Failure to implement either pre- deployment periods

or control regions would make attribution impossible

and therefore compromise the monitoring of all mCDR

deployments in a given ocean region. In addition, an open

registry or metadatabase of the mCDR pilot studies and

deployments would be very useful in this context. It would

provide information, in particular for modelers, on the

location and depth of each activity and key information

on its technical aspects (e.g., for alkalinization, the mineral

type, timing, and amount of alkalinity added).

Furthermore, we assume here the desirable exclusion

principle, whereby the deployment of one type of mCDR

in a given ocean region excludes the possibility of deploy-

ing other types there. This principle stems from the like-

lihood that multiple- type deployments in a given region,

especially as mCDR deployments aim to sequester carbon

at the gigaton scale,1 would make attribution of individ-

ual deployments impossible (Boyd and Bressac, 2016).

Exclusion is also important because multiple-type deploy-

ments could potentially cause interactive side effects that

were not anticipated by single-type mCDR pilot studies.

Pre- deployment periods or control regions—in situ or

in models—are also needed to assess the additionality of

mCDR deployments, defined in IPCC (2022) as: “The prop-

erty of being additional. Mitigation is additional if the

greenhouse gas emission reductions or removals would

not have occurred in the absence of the associated policy

intervention or activity.” Thus, additionality is the require-

ment that the net increase in the air- to- sea CO2 flux due to

an mCDR deployment (i.e., based on detection and attribu-

tion) exceeds the flux in the absence of this mCDR.

Monitoring for additionality will be especially challeng-

ing as the change in net carbon flux into the ocean and the

magnitude of carbon sequestration caused by an mCDR

deployment will be very small compared to natural air-

sea carbon fluxes and the magnitude of the ocean carbon

sink. In addition, the models will need to take into account

the inherent uncertainties of field measurements. This will

pose challenges for both observation and modeling.

1 The intended gigaton-scale magnitude of mCDR deployments would be much larger than existing multiple overlapping uses and perturbations.

For example, global marine capture fisheries and aquaculture harvested 112 Mt of animals and 36 Mt of seaweeds (fresh weight) in marine waters

in 2020 (FAO, 2022), which represented ~0.02 GtC yr–1.



Guidelines for monitoring were set for the Global Ocean

Observing system (GOOS) in the context of the Framework

for Ocean Observation (FOO; Tanhua et al., 2019). A key

element of the FOO is its organization and coordination

around essential ocean variables (EOVs) rather than spe-

cific observing systems. The mCDR operational system

described here differs from the GOOS observations, but

also focuses on EOVs.

We advocate that “essential mCDR variables” would

include most current EOVs (see GOOS, 2021), along with

more detailed data on lower atmosphere CO2 concentra-

tion and oceanic DIC used to estimate air- sea CO2 flux.

Variables would also include wind speed, which has a

strong influence on air- sea gas exchange.

The effects of mCDR deployments on carbon capture and

sequestration will accumulate over time. Consequently,

meeting the three objectives discussed above will require

long- term monitoring.

Given the remote nature and carbon- sequestration tar-

get of open- ocean mCDR deployments, monitoring their

effects will require systems with at least the following char-

acteristics to efficiently address the objectives of detection,

attribution, and determination of side effects:

• Calibrated sensors on autonomous platforms, that is,

satellites (Figure 2) and in situ robots (see Box 1)

• Sampling over large surface areas to address horizontal

eddy diffusion and transport

• Recurrent long- term measurements, commensurate

with the duration of mCDR deployments

• Quasi- simultaneous estimates of air- sea CO2 exchange

and concentrations of DIC, particulate inorganic carbon

(PIC), DOC, and POC from surface to depth to monitor

the fate of the additional carbon

Air- sea CO2 flux cannot be measured directly over large

areas; it would be estimated by modeling. To do so, at

least two parameters of the carbonate system should be

measured in the water column. These parameters include

pH, total DIC, total alkalinity (TA), and CO2 partial pressure

(pCO2). Detecting changes in the carbonate system is chal-

lenging, and detecting a superimposed mCDR effect would

be very difficult.

The monitoring system would combine satellite remote

sensing (Figure 2) and long- term regional in situ measure-

ments. The latter would be performed with autonomous

FIGURE 2. One of the numerous satellites used for mCDR monitoring,

PACE (Plankton, Aerosol, Cloud, ocean Ecosystem, to be launched in

early 2024) will be equipped with a hyperspectral spectrometer that

could be used for assessing possible ecological side effects of mCDR

deployments (e.g.,  changes in phytoplankton community composi-

tion). This figure is a derivative of

File:PACE_Spacecraft_beauty2.jpg by NASA, in the public domain.

robots (see Box 1), as described by Chai et  al. (2020).

Biogeochemical data would subsequently be analyzed using

AI and assimilated in models. The integration of these plat-

forms, analyzing their data with AI, and combining the data

with models is already partly implemented in open- ocean

research (e.g., Claustre et al., 2021) and could be readily

applicable to monitor open- ocean mCDR deployments.

Ideally, the observational and modeling components of

the mCDR monitoring system should be in place prior to

an mCDR deployment. If this is impossible, data collected

by the global networks of Biogeochemical- Argo (BGC- Argo)

floats (Figure 3) and ocean color satellites (Figure 2) could

document natural variability and contribute to the valida-

tion/calibration of models required for attribution.



Satellites and underwater robots operate autonomously,

making relatively high- frequency measurements over sev-

eral years. Present- day BGC- Argo floats (Figure B1) can

achieve 300 profiles of up to 13 different variables2 during

their lifetimes, for example, profiling every five days over

four years. Horizontally, satellites (Figure 2) cover large

surfaces (with a spatial resolution of up to 4 × 4 km), and

2 Temperature, salinity, dissolved O2, pH, dissolved NO3, chlorophyll a, particulate backscattering coefficient (bbp), colored dissolved organic

matter (CDOM), downwelling irradiance, upwelling radiance, particle size spectra, particles and plankton 100–2,00 µm, optical sediment trap





The word “robots” refers here to autonomous vehicles and

platforms, which include:

• Buoyancy- driven robots, encompassing

° Biogeochemical (BGC)- Argo floats (Figures B1 and B2)

° BGC- gliders (Figure B3)

• Uncrewed surface vehicles (USVs; Figure B4)

Floats and USVs could rendezvous for data intercomparison

and transfer (Figure B5). In the future, USVs could intercept

and reposition floats to maintain them in the best monitoring

locations or create “virtual moorings.”

FIGURE B1. A jumbo biogeochemical profiling float (BGC- Argo

float, REFINE type NKE CTS5) can provide profiles every five

days over four years. Sensors are identified that contribute

to the three mCDR monitoring objectives (detection, attribu-

tion, and determination of side effects). Photo credit: David

Luquet, used with his permission. Licensed under CC BY- SA

4.0 by Thomas Boniface

FIGURE B2. BGC- Argo floats record variables from 2,000 m

depth to the surface, where data are rapidly transmitted to a sat-

ellite. The floats then descend to a parking depth (e.g., 1,000 m)

where they stay for 10 days before initiating the next vertical

profile. During the parking phase, floats have the potential to

monitor key properties for mCDR such as wind (passive acous-

tic) and particle flux (transmissometer used as optical sediment

trap). Photo credits: (top) David Luquet and (bottom) Thomas

Boniface, used with their permissions. Licensed under CC BY- SA

4.0 by Thomas Boniface.

BGC- Argo Float

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112