March 2026

IN THIS ISSUE. Under-Ice Hyperspectral Imaging of Nearshore Antarctica; Observations of Biophysical Interactions by Combining SWOT and PACE; Turning Forecasts into Actions: 2024/25 Heatwaves in Australian Waters; And more…

IN THIS ISSUE

UNDER-ICE HYPERSPECTRAL IMAGING OF NEARSHORE ANTARCTICA

OBSERVATIONS OF BIOPHYSICAL INTERACTIONS BY COMBINING SWOT AND PACE

TURNING FORECASTS INTO ACTIONS: 2024/25 HEATWAVES IN AUSTRALIAN WATERS

AND MORE...

Oceanography

THE OFFICIAL MAGAZINE OF THE OCEANOGRAPHY SOCIETY

VOL. 39, NO. 1, MARCH 2026

POWERING SCIENCE-BASED DECISIONS

FOR A BETTER OCEAN.

March 2026 | Oceanography

contents VOL. 39, NO. 1, MARCH 2026

5 QUARTERDECK. Four Reasons (Among Many) Why Early-Career Professionals Should Go to the Office

By E.S. Kappel

6 BREAKING WAVES. Advancing Monitoring of Nearshore Antarctic Sea Ice and Benthic Ecosystems with HIcyBot

By E. Cimoli, J.C. Montes-Herrera, V. Cummings, P. Marriott, R.S. Haynes, and V. Lucieer

14 FEATURE ARTICLE. Biophysical Dynamics of Mesoscale Eddies: Coincident Observations from SWOT and PACE

By L.A. Dove and M.A. Freilich

24 FEATURE ARTICLE. Turning Forecasts into Actions: Marine Heatwaves and Ecosystem-Wide Impacts in Australian

Waters During Summer 2024/25

By A.J. Hobday, C.M. Spillman, J.R. Hartog, G.A. Smith, J. Allnutt, F. Bailleul, L.K. Blamey, S. Brodie, C. Champion, A. Chandrapavan,

M.A. Coleman, M.J. Doubell, R. Duffy, G.L. Grammer, D. Maynard, E.E. Plaganyi, and D. Thomson

34 FEATURE ARTICLE. Establishing the Attractiveness of Seamounts to Marine Megafauna as a Useful Marine

Conservation Tool

By L.J. Broadus, G. Carroll, J. González-Solís, L.-A. Henry, A.B. Leitner, S.U. Stæhr, C. Göke, A.M. Holbach, V.C. Neves, V.H. Paiva,

P. Quillfeldt, F. Ventura, M. Frederiksen, and C. Mohn

44 WORKSHOP REPORT. Strengthening Connections in Observing the North Atlantic Meridional Overturning Circulation:

Outcomes from a Joint RAPID-OSNAP Workshop

By N. Foukal, I. Le Bras, Y. Fu, T. Petit, T.C. Biló, S. Elipot, and B. Moat

50 OCEAN EDUCATION. OceanHackWeek: An Inclusive, Collaborative Approach to Developing Oceanography Data

Science Skills

By C. Mitchell, W.-J. Lee, F. Fernandes, J. Gum, A. Kerney, E. Mayorga, T. Moore, N. Mortimer, N. Ribeiro, and V. Staneva

56 DIY OCEANOGRAPHY. acspype: An Open-Source Python Package for ac-s Data Acquisition and Processing

By I.T. Black, J.T. Black, and M.T. Kavanaugh

63 THE OCEANOGRAPHY CLASSROOM. Debates in the Oceanography Classroom: Challenging Misconceptions and

Engaging with New Knowledge

By L. Svendsen

65 CAREER PROFILES. Alec Bogdanoff, Chief Executive Officer & Co-Founder, Brizaga • Patricia Delgado, Superintendent,

Jug Bay Wetlands Sanctuary

Oceanography | Vol. 39, No. 1

ON THE COVER

A remote observatory tent on Antarctic

fast ice serves as the base for scientists

deploying a novel imaging system

beneath the sea ice, creating new

opportunities to observe the connected

icescape and seafloor ecosystems below.

See Cimoli et al. on p. 6 for the full story.

The Oceanography Society

1 Research Court, Suite 450-117

Rockville, MD 20850 USA

t: (1) 301-251-7708

info@tos.org

HAVE YOU MOVED?

Send changes of address to info@tos.org

or go to https://tosmc.memberclicks.net,

click on Login, and update your profile.

ADVERTISING INFO

Please send advertising inquiries to

info@tos.org or go to https://tos.org/

oceanography/advertise.

CORRECTIONS

Please send corrections to

magazine@tos.org.

SUBMIT A MANUSCRIPT TO Oceanography

PUBLISH WITH US

• EXPAND YOUR AUDIENCE beyond your subdiscipline.

• SHARE YOUR WORK with high-quality graphics that readers can use in classrooms.

• NO TEMPLATES: We design the pages for you.

• FLEXIBLE MANUSCRIPT STRUCTURE accommodates different article types.

• BRING YOUR SCIENCE TO LIFE with videos, animations, audio, and photo galleries.

• MOVE QUICKLY: Our goal is to publish papers within six months of submission.

WHAT GETS OUR ATTENTION

FEATURE ARTICLES (<7,000 words) provide an outlet for making significant advances in oceanography acces

sible to a broad readership. They can include review or synthesis papers, papers that discuss new findings and

how they significantly revise our thinking about a topic, or more traditional scientific research papers.

BREAKING WAVES (<3,500 words) articles describe novel approaches to multidisciplinary problems in ocean

ography. These provocative papers present findings that have the potential to move the field of oceanography

forward or in new directions.

OCEAN EDUCATION (<3,500 words) articles should inspire teachers in higher education to try new active,

student-centered instruction (ranging from short activities to curricula) and provide ideas/materials to do so.

DIY OCEANOGRAPHY (<3,500 words) articles share all of the relevant information on a homemade sensor,

instrument, or software tool(s) so that others can build, or build upon, it. These articles also showcase how this

technology was used successfully in the field.

MEETING/WORKSHOP/CONFERENCE REPORTS (<3,500 words) describe the goals, activities, and accom

plishments of meetings/workshops/conferences in all aspects of ocean science.

COMMENTARIES (<3,500 words) present analyses of issue of interest to Oceanography readers, written by

experts in the field.

RIP CURRENT – NEWS IN OCEANOGRAPHY (<1,500 words) articles describe newsworthy items in the field

of oceanography.

LEARN MORE

See the online Oceanography Author Guidelines for a full listing of manuscript categories and

descriptions, publication fees, and details of the manuscript submission process.

March 2026 | Oceanography

The Oceanography Society was founded in 1988 to advance

oceanographic research, technology, and education, and to dis

seminate knowledge of oceanography and its application through

research and education. TOS promotes the broad understanding

of oceanography, facilitates consensus building across all the dis

ciplines of the field, and informs the public about ocean research,

innovative technology, and educational opportunities throughout

the spectrum of oceanographic inquiry.

OFFICERS

PRESIDENT: Paula Bontempi

PRESIDENT-ELECT: Heidi Dierssen

PAST-PRESIDENT: Deborah Bronk

SECRETARY: Allison Miller

TREASURER: Susan Banahan

COUNCILORS

APPLIED TECHNOLOGY: Alana Sherman

AT-LARGE: Leonard Pace

BIOLOGICAL OCEANOGRAPHY: Grace Saba

CHEMICAL OCEANOGRAPHY: Phoebe Lam

EARLY CAREER: Hilary Palevsky

EDUCATION: Leilani Arthurs

GEOLOGICAL OCEANOGRAPHY: Jon Lewis

JUSTICE, EQUITY, DIVERSITY, AND INCLUSION:

Onjalé Scott Price

OCEAN DATA SCIENCE: Jeremy Werdell

OCEAN SOCIAL SCIENCE AND POLICY: Eric Wade

STUDENT REPRESENTATIVE: Moronke Harris

EXECUTIVE DIRECTOR

Jennifer Ramarui

CORPORATE AND INSTITUTIONAL MEMBERS

Ganey Science

Greenwater Marine Sciences Offshore

Integral Consulting Inc.

National Oceanography Centre

Sea-Bird Scientific

Sequoia

tos.org

EDITOR

Ellen S. Kappel, Geosciences Professional Services Inc.

ASSISTANT EDITOR

Vicky Cullen

DESIGN/PRODUCTION

Johanna Adams

ASSOCIATE EDITORS

Claudia Benitez-Nelson, University of South Carolina

Emmanuel Boss, The University of Maine

Philip Bresnahan, University of North Carolina Wilmington

Luca Centurioni, Scripps Institution of Oceanography

Grace Chang, Integral Consulting Inc.

Tim Conway, University of South Florida

Kjersti Daae, University of Bergen

Mirjam S. Glessmer, University of Bergen

Charles H. Greene, University of Washington

Alistair Hobday, CSIRO Environment

Camille Pagniello, University of Hawai‘i at Mānoa

Helen R. Pillar, University of Texas at Austin

Carol Robinson, University of East Anglia

Amelia Shevenell, University of South Florida

Robert E. Todd, Woods Hole Oceanographic Institution

Peter Wadhams, University of Cambridge

Oceanography is an open-access, peer-reviewed journal pub

lished quarterly by The Oceanography Society. Its articles

present significant research, noteworthy achievements, and

emerging tools and technologies, and connect ocean science

with public policy and education. By publishing articles that

are accessible to a broad audience, Oceanography promotes

cross-disciplinary communication in the ocean sciences.

Oceanography (Print ISSN 1042-8275; Online ISSN 2377-617X)

is published by The Oceanography Society, 1 Research Court,

Suite 450-117, Rockville, MD 20850 USA. Oceanography arti

cles are licensed under a Creative Commons Attribution 4.0

International License, which permits use, sharing, 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. Third-party material used in arti

cles are included in the Creative Commons license unless indi

cated otherwise in a credit line to the material. If the material is

not included in the article’s Creative Commons license, users

will need to obtain permission directly from the license holder

to reproduce the material. Please contact Jennifer Ramarui at

info@tos.org for further information.

Oceanography

tos.org/oceanography

Oceanography | Vol. 39, No. 1

Image courtesy of ESA

OCEANOPTICSCONFERENCE.ORG

OCEAN OPTICS XXVII

SEPTEMBER 13–18, 2026 | GHENT, BELGIUM

REGISTRATION, ABSTRACT SUBMISSION,

AND HOUSING ARE OPEN

It is our pleasure to invite you to attend Ocean Optics XXVII, the twenty-seventh

edition of the biennial international Ocean Optics Conference. This event brings

together a diverse ocean optics community, including oceanographers, limnolo-

gists, optical engineers, Earth observation scientists, resource managers, and pol

icy professionals from across the globe, all united by a shared passion for optics in

aquatic environments.

Continuing the rich tradition of the Ocean Optics Conference, this year’s program

will feature invited plenary speakers, selected oral presentations, and numerous

poster presentations showcasing the latest advancements and discoveries in the

field. For those looking to deepen their skills, pre-conference short courses and

workshops will offer hands-on training in data analysis and emerging technologies.

IMPORTANT DATES

EARLY REGISTRATION: Closes May 8, 2026

ABSTRACT SUBMISSION: Closes May 8, 2026

HOUSING: Book soon as rooms may sell out quickly

PRESENTATION/POSTER ASSIGNMENTS:

Notifications will be sent by late June 2026

March 2026 | Oceanography

QUARTERDECK

1. In-person participation in a team with shared goals and daily

interactions with more senior staff is critical for professional

development. Such engagement of course requires that the

older generation be present in the office as well.

2. Interactions with staff you otherwise might have no reason to

connect with may prove to be consequential. Aside from sci

entific colleagues, there is a lot to learn from technical and

administrative staff and others who facilitate your work—or

someone else’s. Things you learn from an accountant, a con

tracts officer, or an HR manager might turn out to be very

useful in your career.

3. In the office, you can get meaningful feedback on your work,

sooner rather than later. Nothing in Zoom, Teams, Google

Meet, Slack, or email is the same as walking to the next office

and asking a colleague for input. That quick exchange or a

longer discussion invariably improves your work.

4. Sometimes unplanned interactions can open doors. You

bump into a colleague while getting a cup of coffee and

exchange ideas or discuss issues. Who knows where those

kinds of relationships may lead someday? Any exchange can

be part of developing connections and trust so that you can

do your job better and advance your career.

There are, of course, other advantages to being in the office

every day. One web site I found listed 18. But, to underscore the

importance of simply “being there,” the early career profession

als I spoke with gained some valuable information by attending a

conference in person and by randomly being assigned to me as a

mentor. You just never know where such serendipitous encounters

may lead and what you might learn.

FOUR REASONS (AMONG MANY)

WHY EARLY-CAREER PROFESSIONALS

SHOULD GO TO THE OFFICE

Ellen S. Kappel, Editor

ARTICLE DOI. https://doi.org/10.5670/oceanog.2026.e106

At this year’s Ocean Sciences Meeting in Glasgow (February 22–27,

2026), I had the privilege of participating as a mentor in a couple

of “Career Conversations and Coffee” events where students and

early-career scientists could learn about career opportunities from

ocean science professionals representing government, academia,

industry, and nonprofit organizations. Mentors were scattered

among a dozen or more tables, and mentees were assigned tables

that aligned with their career interests.

After I provided some background about my own career tra

jectory, discussions wandered depending on the interests of the

mentees (generally about half dozen at each table at any one

time). Both mornings, participants had plenty of questions, and

I stayed well beyond the scheduled end time. At one meeting, I

emphasized—as I often do when engaging with students and early-

career professionals—the importance of going to the office every

day. Having worked in a vibrant office for a dozen years and then

from a home office for more than 25, I am very familiar with the

advantages and pitfalls of a remote work environment.

Much to my surprise, the early career folks I was speaking with

lit up when I raised this issue. They said that they like their remote

work situation but are being forced to go back to the office by

their employers, with potential consequences if they do not. They

commented that their employers never articulated the benefits of

returning to an office environment, a lack of communication that

created some bad feelings and misunderstandings.

Although the debate regarding remote versus in-office work

environments seems like an old story by now, my discussions with

these early-career scientists encouraged me to write (again) about

this subject. If rising professionals were lacking information about

the benefits of going to an office, perhaps others were, too. So, I

offer the following benefits of working in an office environment.

Oceanography | Vol. 39, No. 1

ADVANCING

MONITORING OF

NEARSHORE ANTARCTIC

SEA ICE AND BENTHIC

ECOSYSTEMS WITH

HIcyBot

By Emiliano Cimoli, Juan Carlos Montes-Herrera, Vonda Cummings,

Peter Marriott, Ryan S. Haynes, and Vanessa Lucieer

BREAKING WAVES

March 2026 | Oceanography

BACKGROUND

Sea ice is the dominant feature structuring Antarctic coastal marine

ecosystems. Fast ice (i.e., annual sea ice attached to the shore) can

support some of the most productive marine habitats on Earth,

contributing approximately 55%–68% of the total primary pro

duction in coastal areas (McMinn et al., 2010). Other contributors

include phytoplankton as well as benthic microalgae and macro-

algae. As coastal sea ice extent and stability shift, the balance of pri

mary production between phytoplankton, sea ice algae, and ben-

thic microalgae may change significantly, with cascading impacts

on tightly linked seafloor biodiversity and overall ecosystem func

tioning (Norkko et al., 2007; Wing et al., 2018).

Seasonal changes in fast ice thickness, extent, and snow cover

regulate light transmission, strongly influencing under-ice pri

mary production. The presence or absence of sea ice alters aquatic

photosynthesis by modifying light intensity and spectral qual

ity (Soja-Woźniak et al., 2025). The shallow depths typical of ice-

covered coastal zones further promote efficient export of ice algae

and other particulate matter to the seafloor via both primary and

secondary production, resulting in tight cryo-pelagic-benthic

coupling. The resulting quantity and quality of phytodetritus

through the seasons play a role in shaping the diversity and abun

dance of benthic microbial, algal, and invertebrate communities

(Rossi et al., 2019).

Historically, we have struggled to capture the high spatio-

temporal variability at which sea ice biogeochemical processes

occur (from 1 µm to 10–1,000 m, from seconds to decades).

Shifts in the sources and fluxes of organic matter from the ice to

benthic food webs may serve as early indicators of broader eco

logical responses to physical environmental changes driven by cli

mate change in Antarctica (Clark et al., 2017). Looking directly

below the surface, the complexity of the Antarctic seafloor speaks

for itself. Shallow benthic ecosystems display remarkable struc

tural, functional, and biological diversity, shaped by dense assem

blages of sessile epifaunal and infaunal invertebrates, fishes

and algae, patchy habitat mosaics, and tightly coupled trophic

interactions that persist despite extreme conditions (Gutt et al.,

2015). The scales and intensity of this coupling vary consider

ably with depth and connectivity with primary food sources (from

meters to kilometers).

Because nearshore productivity and biodiversity are closely

linked to sea ice dynamics, there is a pressing need for monitor

ing methods that can efficiently and simultaneously track changes

in light levels, changes in cryo-benthic communities, and food

pulses. These techniques must also be practical for use in extremely

remote and challenging environments—such as under two meters

of thick Antarctic fast ice. Historically, this has been done using

diver-based surveys, which are necessarily restricted in spatial and

temporal extent and in maximum depth. Recently, the availabil

ity of small remotely operated vehicles (ROVs) has enabled the

expansion of coastal under-ice sampling areas at reduced logis

tical complexity. Incorporating hyperspectral imaging into these

platforms can expand their utility beyond mapping the under-ice

seascape and associated biodiversity, enabling the spatial and tem

poral quantification of ecosystem function through measurements

of biomass and accessory pigment estimates that can serve as indi

cators of organismal health.

UNDERWATER HYPERSPECTRAL IMAGING

The devil, of course, is in the detail. Process studies should strive

to take observations at the spatial resolution over which the pro

cess operates (Levin, 1992). Analogous to the rapid rise of close-

range terrestrial hyperspectral imaging using drones over the

past decade, in situ underwater hyperspectral imaging (UHI) has

emerged as a powerful tool for generating high-resolution biogeo-

chemical maps of seafloor properties. First introduced as a ded

icated subdiscipline of marine remote sensing by Johnsen et al.

(2013), proximity-based hyperspectral sensing in marine envi

ronments is an emerging methodology; to date, only a handful of

documented systems have been deployed on diver-operated plat

forms, ROVs, or under-ice sleds across diverse habitats from trop

ical reefs to the deep sea (Montes-Herrera et al., 2021; Summers

ABSTRACT. In Antarctica’s coastal waters, the seafloor hosts a surprisingly rich diversity of life, shaped by intricate interactions

between sea ice and flora and faunal communities. As sea ice forms and melts seasonally, it modulates the availability of light to both

the microscopic algae living within and beneath the ice itself (sympagic algae), and, in turn, influences the productivity and biodiver

sity of the benthic ecosystems below. The complex pathways that connect sea ice dynamics with the benthos are particularly vulnerable

to climate-driven changes in sea ice cover. Understanding responses to a rapidly changing icescape, shaped by warming air and ocean

conditions, is essential for predicting the future of benthic biodiversity and its ecosystem functions and services, including biogeochem-

ical cycling and carbon storage. Key to this understanding is coupling structural and biological observations across these mirrored eco

systems. We describe a proof-of-concept under-ice hyperspectral imaging platform, HIcyBot, that we deployed in the Ross Sea Marine

Protected Area in 2023. HIcyBot is a remotely operated vehicle that leverages emerging sea ice bio-optical models to quantify ice algae

biomass and classifies fine-scale seafloor features through integration of underwater hyperspectral imaging, stereophotogrammetry, and

acoustic positioning. Through near simultaneous high-resolution characterization of these under-ice realms, the platform introduces a

novel, spatially explicit approach to understanding how biodiversity and the ecosystem function beneath Antarctic sea ice and how they

are responding to a changing icescape.

Oceanography | Vol. 39, No. 1

et al., 2022; Lange et al., 2024; Anhaus et al., 2025). Its repeatable,

non-invasive capabilities make UHI especially suited for tracing

and quantifying fine-scale biogeochemical processes in remote

and sensitive environments.

UNDER-ICE MAPPING WITH HICYBOT

The HIcyBot system aims to produce co-georeferenced UHI and

three-dimensional (3D) structural data of the seafloor and overlying

fast ice based on several surveys conducted within hours of one

another (Figure 1). The optical components include three primary

subsystems: a central hyperspectral imager, a stereophotogrammet-

ric dual-camera system, and a USB live-stream reference camera

contained within a tethered underwater enclosure (Figure 2).

An eight-hydrophone-array ultra-short baseline (USBL) sen

sor was integrated to enable geolocation of pushbroom hyperspec-

tral frames with synchronized, timestamped Global Navigation

Satellite System (GNSS) and inertial navigation data. The

co-mounted dual-vision cameras leverage stereophotogramme-

try techniques to derive camera positions and orientations from

overlapping stereo images, providing independent vehicle pose

estimates that support local motion compensation for the UHI in

conditions where USBL signal quality is degraded beneath the ice

(Figure 2). The payload enclosure was incorporated into a custom

ized BlueROV2 heavy-configuration kit. A key design feature was

the addition of bottom ballast to create a pendulum-like configu

ration, enabling passive, stable vertical alignment for nadir-facing

(a) Sea-Ice Mode

(b) Seafloor Mode

FIGURE 1. HIcyBot is a proof-of-concept under-ice remotely operated vehicle (ROV) with hyperspectral imaging capability, designed for concomitant mapping

of (a) sympagic (ice-associated) habitats, and (b) benthic, seafloor-associated habitats that remain largely inaccessible to other marine sensing and survey

techniques. Pushbroom sensors record one spatial line per frame as the platform moves, forming a 3D hyperspectral cube (X, Y, λ) linking spatial and spec

tral domains. GNSS = Global Navigation Satellite System. USBL = ultra-short baseline. MPA = Marine Protected Area.

March 2026 | Oceanography

imaging without active control. Technical specifications, image

pre-processing, and georeferencing challenges for UHI are pro

vided in the online Supplementary Materials.

FIELD TEST IN THE ROSS SEA MPA

The system was first deployed from November 7 to 22, 2023,

under land-fast sea ice at Cape Evans, within the Ross Sea Marine

Protected Area (MPA), Antarctica, the world’s largest MPA. This

region currently represents one of the least human-impacted envi

ronments, yet one of the most challenging to study due to extreme

Antarctic conditions, remoteness, and extensive sea ice cover. It

is considered a benthic biodiversity hotspot fueled by overlying

under-ice algae of the highest concentrations on record. The pri

mary objective was to generate preliminary products showing

ice-associated (sympagic) algal biomass or chlorophyll a (Chl-a)

concomitant with seafloor biodiversity, including class-level maps

incorporating photosynthetic and accessory pigments of selected

organism groups, all resolved along the hyperspectral transects at

millimeter scales.

LOOKING ABOVE: AN ICE ALGAL COMMUNITY

PRODUCT EXAMPLE

Quantifying sympagic biomass was previously only achievable

with labor-intensive ice coring techniques and laboratory process

ing. UHI offers a significantly improved way to monitor the bio

physical process in sea ice at varying spatial and temporal scales.

Building on emerging work to develop bio-optical models for

under-ice hyperspectral imaging in Antarctic land-fast sea ice, a

“spectra to biomass” correlation was applied at a pixel level to the

“x, y, λ hypercube” to fundamentally enhance the estimation of

under-ice algal biomass and their microscale distribution patterns

remotely and quantitatively (e.g., LAUC650–700 from Cimoli et al.,

2025; see Figure 2 workflow).

Figure 3a provides an example of a quantitative high-

resolution estimation of sympagic Chl-a from an ROV. It shows

an ice algal biomass map derived from optimized bio-optical

algorithms with fine-scale variability associated with features

such as brinicles. Brinicles, often referred as “ice stalactites,” are

complex physical structures that form as dense, saline brine is

excluded from seawater as it freezes and drains downward into

the ocean, freezing the surrounding seawater and creating hol

low, downward-growing ice tubes (Testón-Martínez et al., 2024).

Brinicles appear to support up to twice the amount of algal bio

mass around their edges compared with surrounding flat areas

(Figure 3a), potentially providing important habitat for under-ice

grazers through both enhanced food availability and their shel

tering role. With HIcyBot, we can scale up efforts to quantify

and characterize ice algal biomass through larger spatial swaths,

achieving millimeter-scale resolution while extending cover

age to survey-grid scales (e.g., 100 × 100 m), and potentially

providing pigment proxies of algal photoacclimation state and

health (Cimoli et al., 2025).

LOOKING BELOW: A BENTHIC BIODIVERSITY

PRODUCT EXAMPLE

HIcyBot’s internal enclosure can quickly be physically inverted

within the customized BlueROV2 frame between deployments,

enabling a downward-facing configuration for seafloor surveys.

Operating roughly half a meter above the dark benthos at depths

of 10–40 m with active illumination, this setup preserves vehi

cle nadir stability and allows detailed imaging of benthic habitats

located beneath previously scanned ice surfaces.

FIGURE 2. The data processing workflow for both sea-ice and seafloor

HIcyBot operating modes (upward- and downward-looking) includes opti

cal and acoustic data streams and subsequent processing steps that lead

to example data products, with both modes sharing the same general pre

processing pipeline. RGB = red, green, blue. UHI = underwater hyperspec-

tral imaging. USBL = ultra-short baseline. DEM = digital elevation model.

CCA = crustose coralline algae. MPB = microphytobenthos.

Oceanography | Vol. 39, No. 1

(a) Sea Ice

(b) Seafloor Seep

FIGURE 3. Stereophotogrammetry enabled 3D reconstructions of the (a) sea-ice underside and (b,c) seafloor. Georeferenced models provided camera

pose data (pitch, roll, yaw) that, combined with USBL positioning, aided hyperspectral image geolocation. (a) Under-ice algal biomass map derived from

bio-optical algorithms showing fine-scale variability associated with brinicles (“ice stalactites”). (b) and (c) Spectral Angle Mapper-based seafloor classi

fication outputs identify major benthic classes, including a methane seep, millimeter-scale spectral proxies maps of Chl-a (MPBI), and phycoerythrin in

crustose coralline algae (ANMB).

March 2026 | Oceanography

Following reflectance conversion, we used the Spectral Angle

Mapper (SAM) algorithm as an example of UHI classifica

tion. SAM compares the angle between the spectral signature of

each pixel and known target reference spectra, essentially mea

suring how similar two spectral “color patterns” are, regard

less of brightness. Because of the close distance to the target and

use of artificial lighting, even small changes in distance from

the sensor to the seafloor can cause large intensity variations

(Dumke et al., 2018).

Automated hyperspectral classification of the seafloor at

Cape Evans enabled generation of high-resolution maps show

ing benthic biodiversity, habitat composition, and the area cov

ered by each class instantly and independent of the shape or illu

mination of the targets (Figure 3b). Substrate was distinguished

between pebble/gravel and fine sand, aligning well with the known

geomorphology of the site. A diffuse surface layer, likely composed

of microalgae and/or diatoms, was observed across these sedi

ment types. This “fluff layer” may be epipsammic, living on sand

grains, or epipelic, occurring on finer sediments like mud or silt

(Sutherland, 2008). Its origin remains uncertain and could either

reflect in situ microphytobenthic growth or settled material from

the water column or the ice-algae above.

The red macroalgae, Phyllophora antarctica, dominated the

benthic assemblages, often attached to the sea urchin Sterechinus

neumayeri via their tube feet or overgrown by brown microalgal

assemblages. The urchins camouflage themselves with algal fronds

and debris, forming a mutualism in which the algae avoid deep

displacement while the urchins gain protection, a known inter

action reinforced by Phyllophora’s chemical defenses against

herbivory. Crustose coralline algae, considered important eco

system engineers, were found encrusting many of the exposed

rocks. Automated image classification further revealed a diversity

of invertebrates, including Abatus spp. (burrowing echinoids with

five radial grooves), and shells of the bivalve Laternula elliptica. The

cod icefish (Trematomus bernacchii) can be seen swimming above

a “starry” seafloor populated by the sea star Odontaster spp., whose

colors range from pink to orange-yellow. Nemerteans (Parborlasia

corrugatus), conspicuous worm-like scavengers and predators,

were found traversing the seafloor or curled. We also captured and

spectrally classified imagery of a newly emerged methane seep at

Cape Evans (Figure 3b) adding to recent observations along the

Ross Sea coast (Seabrook et al., 2025).

Following classification, the true potential of UHI lies in

targeted millimeter-scale mapping of benthic organisms’ bio

chemical traits in situ. For example, here we applied algorithms

developed under controlled laboratory conditions to estimate

spectral proxies of R-phycoerythrin, an indicator of environ

mental state and health in coralline algae (using ANMB565 from

Montes-Herrera et al., 2024), and Chl-a in epipelic communities,

using models derived from comparable biological systems such

as microphytobenthos (using MPBI from Chennu et al., 2013;

Figure 3c). While these are realistic proxies, consistent quantita

tive field calibrations remain limited due to the need for species-

and site-specific validation, robust calibration datasets, and stan

dardized protocols.

OPTICAL MONITORING OF CRYO-BENTHIC

LINKAGES IN A CHANGING ICESCAPE

Projected shifts in seasonal sea ice dynamics underscore the grow

ing need to assess how well existing tools can capture the immediate

responses of shallow, ice-covered marine habitats across different

spatio-temporal scales. While traditional ice coring and seafloor

surveys provide valuable ecological insights, new high-resolution

optical systems offer non-invasive methods for capturing the spec

tral fingerprints of different sympagic and benthic organisms and

substrates that allow their classification and quantification based

on their spectral features (Figure 4a–c).

Spatially explicit UHI data open a new frontier for exploring

and quantifying the dynamic exchange between sea ice and the

seafloor, offering a “spectral breadcrumb trail” that links the ice

canopy to the life below. For example, can spectral fingerprints help

quantitatively track biogeochemical linkages between sea ice and

the seafloor? Do sea ice algae retain pigment-specific signatures,

and/or remain detectable after seafloor deposition, and can they

be distinguished from native benthic phototrophs? Understanding

how light transmission, algal productivity, and “slough-off ” affect

benthic biodiversity (both phylogenetic and functional) demands

an integrated approach combining hyperspectral imaging with

process-based measurements (e.g., ice-hanging sediment traps).

Figure 4 highlights spectral characteristics of the coupled ice-

benthic system, enabling quantification of light-driven impacts

on benthic communities and associated biogeochemical pro

cesses. The system’s ability to capture both structural and compo

sitional complexity, including organism volume, habitat architec

ture, and phototrophic biomass per unit area, provides a solid basis

for developing indicators of the emerging Antarctic blue carbon

potential. It may also help in differentiation of the relative con

tributions of sympagic, pelagic, and authigenic carbon sources,

which is important for guiding conservation priorities and climate

mitigation strategies in polar regions (Sands et al., 2023).

FUTURE DEVELOPMENTS

The HIcyBot system is transitioning from a proof-of-concept to a

functional monitoring toolkit, with the potential to enable remote

observations of under-ice ecological and biogeochemical pro

cesses that are otherwise too difficult to sustain. Such efforts aim

to be integrated within broader, standardized long-term observa

tion programs at representative sites like those coordinated by the

Antarctic Nearshore and Terrestrial Observing System (ANTOS)

Expert Group, and may ultimately evolve into fixed, station-like

robotic observatories with edge artificial intelligence capabilities,

forming part of integrated coastal monitoring networks.

Oceanography | Vol. 39, No. 1

The UHI community now faces a pressing need for develop

ing standardized protocols in georectification, radiometric cor

rection, and reflectance calibration to fully unlock the technique’s

potential. Encouragingly, efforts are already underway to estab

lish greater consistency and rigor to this emerging research field

(Løvås et al., 2022; Liu et al., 2024).

Broader implementation also depends on both technology

and accessibility. Given the high costs and logistical complexity

of UHI and Antarctic work, maximizing data cost performance

remains critical. Identifying the most informative narrow spec

tral bands will support the creation of simpler, lower-cost multi-

spectral systems and help to democratize underwater spectral

imaging applications.

SUPPLEMENTARY MATERIALS

The supplementary materials are available online at https://doi.org/10.5670/

oceanog.2026.e107.

REFERENCES

Anhaus, P., C. Katlein, S. Arndt, D. Krampe, B.A. Lange, I. Matero, E. Salganik, and

M. Nicolaus. 2025. Under-ice environment observations from a remotely operated

vehicle during the MOSAiC expedition. Scientific Data 12(1):944, https://doi.org/

10.1038/s41597-025-05223-1.

Chennu, A., P. Färber, N. Volkenborn, M.A.A. Al-Najjar, F. Janssen, D. de Beer, and

L. Polerecky. 2013. Hyperspectral imaging of the microscale distribution and dynam

ics of microphytobenthos in intertidal sediments: Hyperspectral imaging of MPB

biofilms. Limnology and Oceanography: Methods 11(10):511–528, https://doi.org/

10.4319/lom.2013.11.511.

Cimoli, E., F. Kennedy, K.M. Meiners, L.C. Lund-Hansen, Z. Malenovský, B.K. Sorrell,

A. Martin, and V. Lucieer. 2025. Impact of experimental light regimes on Antarctic

fast-ice algal communities observed by combining ice coring and imaging spectros

copy. Elementa: Science of the Anthropocene 13(1):00094, https://doi.org/10.1525/

elementa.2024.00094.

Clark, G.F., J.S. Stark, A.S. Palmer, M.J. Riddle, and E.L. Johnston. 2017. The roles of

sea-ice, light and sedimentation in structuring shallow Antarctic benthic communi

ties. PLoS ONE 12(1):e0168391, https://doi.org/10.1371/journal.pone.0168391.

Dumke, I., S.M. Nornes, A. Purser, Y. Marcon, M. Ludvigsen, S.L. Ellefmo, G. Johnsen,

and F. Søreide. 2018. First hyperspectral imaging survey of the deep sea

floor: High-resolution mapping of manganese nodules. Remote Sensing of

Environment 209:19–30, https://doi.org/10.1016/j.rse.2018.02.024.

Gutt, J., V. Cummings, P. Dayton, E. Isla, A. Jentsch, and S. Schiaparelli. 2015. Antarctic

marine animal forests: Three-dimensional communities in Southern Ocean eco

systems. Pp. 1–30 in Marine Animal Forests: The Ecology of Benthic Biodiversity

Hotspots. S. Rossi, L. Bramanti, A. Gori, and C. Orejas Saco del Valle, eds, Springer

International Publishing, Cham, https://doi.org/10.1007/978-3-319-17001-5_8-1.

FIGURE 4. Comparison of pixel-derived spectral signatures from sea-ice and seafloor features. (a) Sea-ice spectra represent transmitted light through ice-

algal layers, while seafloor spectra (b) and (c) indicate reflectance from benthic surfaces. These distinct spectral “fingerprints” reveal substrate contrasts and

highlight biological phototrophic pigment absorbance (e.g., Chl-a, carotenoids, phycoerythrin) that characterize photophysiology. Observed organism struc

tural complexity, illumination, and imaging conditions can, however, influence these signatures, underscoring the need for further empirical calibration to

develop targeted bio-optical models.

March 2026 | Oceanography

Johnsen, G., Z. Volent, H. Dierssen, R. Pettersen, M.V. Ardelan, F. Søreide, P. Fearns,

M. Ludvigsen, and M. Moline. 2013. Underwater hyperspectral imagery to create

biogeochemical maps of seafloor properties. Pp. 508–540 in Subsea Optics and

Imaging, Elsevier.

Lange, B.A., I. Matero, E. Salganik, K.L. Campbell, C. Katlein, P. Anhaus, J. Osanen, and

M.A. Granskog. 2024. Biophysical characterization of summer Arctic sea-ice habi

tats using a remotely operated vehicle-mounted underwater hyperspectral imager.

Remote Sensing Applications: Society and Environment 35:101224, https://doi.org/

10.1016/j.rsase.2024.101224.

Levin, S.A. 1992. The problem of pattern and scale in ecology: The Robert H.

MacArthur Award Lecture. Ecology 73(6):1,943–1,967, https://doi.org/10.2307/

1941447.

Liu, H., N. Summers, Y.-C. Chen, H.S. Løvås, G. Johnsen, D. Koestner, C. Sætre, and

B. Hamre. 2024. Pixelwise immersion factor calibration for underwater hyper-

spectral imaging instruments. Optics Express 32(11):19854, https://doi.org/10.1364/

OE.523641.

Løvås, H.S., A.A. Mogstad, A.J. Sørensen, and G. Johnsen. 2022. A methodology for

consistent georegistration in underwater hyperspectral imaging. IEEE Journal of

Oceanic Engineering 47(2):331–349, https://doi.org/10.1109/JOE.2021.3108229.

McMinn, A., A. Pankowskii, C. Ashworth, R. Bhagooli, P. Ralph, and K. Ryan. 2010.

In situ net primary productivity and photosynthesis of Antarctic sea ice algal,

phytoplankton and benthic algal communities. Marine Biology 157(6):1,345–1,356,

https://doi.org/10.1007/s00227-010-1414-8.

Montes-Herrera, J.C., E. Cimoli, V. Cummings, N. Hill, A. Lucieer, and V. Lucieer. 2021.

Underwater hyperspectral imaging (UHI): A review of systems and applications for

proximal seafloor ecosystem studies. Remote Sensing 13(17):3451, https://doi.org/

10.3390/rs13173451.

Montes-Herrera, J.C., E. Cimoli, V.J. Cummings, R. D’Archino, W.A. Nelson, A. Lucieer,

and V. Lucieer. 2024. Quantifying pigment content in crustose coralline algae using

hyperspectral imaging: A case study with Tethysphytum antarcticum (Ross Sea,

Antarctica). Journal of Phycology 60(3):695–709, https://doi.org/10.1111/jpy.13449.

Norkko, A., S.F. Thrush, V.J. Cummings, M.M. Gibbs, N.L. Andrew, J. Norkko, and

A.-M. Schwarz. 2007. Trophic structure of coastal Antarctic food webs asso

ciated with changes in sea ice and food supply. Ecology 88(11):2,810–2,820,

https://doi.org/10.1890/06-1396.1.

Rossi, L., S. Sporta Caputi, E. Calizza, G. Careddu, M. Oliverio, S. Schiaparelli,

and M.L. Costantini. 2019. Antarctic food web architecture under varying

dynamics of sea ice cover. Scientific Reports 9(1):12454, https://doi.org/10.1038/

s41598-019-48245-7.

Sands, C.J., N. Zwerschke, N. Bax, D.K.A. Barnes, C. Moreau, R. Downey, B. Moreno,

C. Held, and M. Paulsen. 2023. Perspective: The growing potential of Antarctic

blue carbon. Pp. 16–17 in Frontiers in Ocean Observing: Emerging Technologies

for Understanding and Managing a Changing Ocean. E.S. Kappel, 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,

Oceanography 36(Supplement 1), https://doi.org/10.5670/oceanog.2023.s1.5.

Seabrook, S., C.S. Law, A.R. Thurber, Y. Ladroit, V. Cummings, L. Tait, A. Maurice,

and I. Hawes. 2025. Antarctic seep emergence and discovery in the shallow

coastal environment. Nature Communications 16(1):8740, https://doi.org/10.1038/

s41467-025-63404-3.

Soja-Woźniak, M., T. Holtrop, S. Woutersen, H.J. Van Der Woerd, L.C. Lund-Hansen,

and J. Huisman. 2025. Loss of sea ice alters light spectra for aquatic photosynthe

sis. Nature Communications 16(1):4059, https://doi.org/10.1038/s41467-025-59386-x.

Summers, N., G. Johnsen, A. Mogstad, H. Løvås, G. Fragoso, and J. Berge. 2022.

Underwater hyperspectral imaging of Arctic macroalgal habitats during the polar

night using a novel mini-ROV-UHI portable system. Remote Sensing 14(6):1325,

https://doi.org/10.3390/rs14061325.

Sutherland, D.L. 2008. Surface-associated diatoms from marine habitats at Cape

Evans, Antarctica, including the first record of living Eunotogramma marginopunc-

tatum. Polar Biology 31(7):879–888, https://doi.org/10.1007/s00300-008-0426-z.

Testón-Martínez, S., L.M. Barge, J. Eichler, C.I. Sainz-Díaz, and J.H.E. Cartwright. 2024.

Experimental modelling of the growth of tubular ice brinicles from brine flows under

sea ice. The Cryosphere 18(5):2,195–2,205, https://doi.org/10.5194/tc-18-2195-2024.

Wing, S.R., J.J. Leichter, L.C. Wing, D. Stokes, S.J. Genovese, R.M. McMullin, and

O.A. Shatova. 2018. Contribution of sea ice microbial production to Antarctic ben-

thic communities is driven by sea ice dynamics and composition of functional

guilds. Global Change Biology 24(8):3,642–3,653, https://doi.org/10.1111/gcb.14291.

ACKNOWLEDGMENTS

This research was supported by the Australian Research Council Special Research

Initiative, Australian Centre for Excellence in Antarctic Science (project number

SR200100008), and the New Zealand Antarctic Science Platform (K882-2324-A

RSRED – Benthic Sentinel Sites; Ministry of Business, Innovation, and Employment

Grant/Award Number MBIE ANTA1801). We are grateful to staff at Antarctica

New Zealand for their field-based operational and logistical support. Mechatronics

support was provided by Adept Turnkey Ltd. and the UTAS Central Science

Laboratory, with special thanks to Sean Sarikas and Philip Hortin. This work contrib

utes to BEPSII Task Group 2: New Technologies.

AUTHORS

Emiliano Cimoli (emiliano.cimoli@utas.edu.au), Institute for Marine and Antarctic

Studies, College of Sciences and Engineering, University of Tasmania, and Australian

Centre for Excellence in Antarctic Science, University of Tasmania, Hobart, Tasmania,

Australia. Juan Carlos Montes-Herrera, Institute for Marine and Antarctic Studies,

College of Sciences and Engineering, University of Tasmania, and Discipline of

Geography and Spatial Sciences, School of Technology, Environments and Design,

College of Sciences and Engineering, University of Tasmania, Hobart, Tasmania,

Australia. Vonda Cummings and Peter Marriott, Earth Sciences New Zealand,

Wellington, New Zealand. Ryan S. Haynes, Discipline of Geography and Spatial

Sciences, School of Technology, Environments and Design, College of Sciences and

Engineering, University of Tasmania, Hobart, Tasmania, Australia. Vanessa Lucieer,

Institute for Marine and Antarctic Studies, College of Sciences and Engineering,

University of Tasmania, and Australian Centre for Excellence in Antarctic Science,

University of Tasmania, Hobart, Tasmania, Australia.

ARTICLE CITATION

Cimoli, E., J.C. Montes-Herrera, V. Cummings, P. Marriott, R.S. Haynes, and V. Lucieer.

2026. Advancing monitoring of nearshore Antarctic sea ice and benthic ecosystems

with HIcyBot. Oceanography 39(1):6–13, https://doi.org/10.5670/oceanog.2026.e107.

This is an open access article made available under the terms of the Creative

Commons Attribution 4.0 International License, which permits use, sharing, adapta

tion, 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.

Oceanography | Vol. 39, No. 1

FEATURE ARTICLE

BIOPHYSICAL DYNAMICS

OF MESOSCALE EDDIES

COINCIDENT OBSERVATIONS FROM SWOT AND PACE

By Lilian A. Dove and Mara A. Freilich

BACKGROUND IMAGE. Gulf Stream eddies seen as a green swirls of high chloro

phyll off the US East Coast. This image was taken by the Ocean Color Instrument

(OCI) sensor aboard the PACE satellite on May 2, 2024. Source: NASA

Oceanography | Vol. 39, No. 1

ABSTRACT. The overlapping missions of the NASA Surface Water and Ocean Topography (SWOT) satellite and NASA Plankton,

Aerosol, Cloud, ocean Ecosystem (PACE) satellite provide the opportunity to observe oceanic biophysical interactions from space at

unprecedented spatiotemporal scales. We use provisional datasets from these two cutting-edge missions to investigate subseasonal to sea

sonal variability in the microbial community compositions of subtropical mesoscale eddies. The results highlight the capacity of mesoscale

eddies to act as transient ecological niches that restructure the surface marine microbial community. For the first time, the combination of

SWOT and PACE enables space-based observations of plankton community composition alongside the physical processes that structure it.

March 2026 | Oceanography

INTRODUCTION

Satellite remote sensing revolutionized our understanding of the

ocean surface, revealing tight coupling between biological and phys

ical processes at scales of 100 km and larger and on timescales of

weeks to months (McGillicuddy et al., 1998; Siegel et al., 1999; Gaube

and McGillicuddy, 2017). Interactions between physical circula

tion and ocean biology at horizontal scales ranging from the sub-

mesoscale (O(0.1–10 km)) to the mesoscale (O(100 km)) have also

been implicated as vital for structuring marine ecosystems (White

et al., 2010; McGillicuddy, 2016; Lévy et al., 2018) as they move

nutrients and organic matter vertically in the water column (Klein

and Lapeyre, 2009; Mahadevan, 2016), stir communities laterally

(d’Ovidio et al., 2010; Clayton et al., 2013; Lehahn et al., 2017), and

connect the ocean’s surface mixed layer and upper mesopelagic zone

(Freilich et al., 2024, 2026). However, observational validation of

these tight biophysical couplings with implications for community

composition only exist through case studies. The overlapping mis

sions of the NASA Surface Water and Ocean Topography (SWOT;

Morrow et al., 2019; Fu et al., 2024) and NASA Plankton, Aerosol,

Cloud, ocean Ecosystem (PACE; Gorman et al., 2019; Werdell et al.,

2019) satellites—providing high-resolution sea surface altimetry and

ocean color data, respectively—offer an unprecedented opportunity

for expanding biophysical research at the ocean surface.

At global scales, satellite altimetry is used to estimate surface

velocity using the geostrophic approximation, which requires com

puting sea surface height gradients. Conventional altimetry has

been used to map sea surface height at scales of ¼° by averaging

over 10 or more days of observations (Pujol et al., 2016). This lim

its observations of currents to those that are in geostrophic bal

ance and are relatively large in scale. Through two-dimensional

sea surface height observations, SWOT observations have already

revealed energetic coherent submesoscale eddies (wavelengths of

15 km and larger; Morrow et al., 2019; Fu et al., 2024; Archer et al.,

2025; Wang et al., 2025) as well as fine-scale eddies and filamentary

structures (Coadou-Chaventon et al., 2025; Qiu and Chen, 2025;

Verger-Miralles et al., 2025) and ocean waves (Ardhuin et al., 2025).

Increased resolution in horizontal velocities can also be used to infer

vertical velocities using dynamical frameworks such as the omega

equation or surface quasi-geostrophic approaches (Carli et al., 2025),

although these methods rely on simplifying assumptions. Mesoscale

eddies have relatively weak vertical velocity (O(1−10 m/day)), but

the larger velocity gradients and faster temporal scales associated

with submesoscale dynamics lead to larger vertical velocities, even

when dynamics are largely geostrophic or cyclostrophic and result

in tighter biophysical coupling in models and observations of shifts

in plankton community structure (Lévy et al., 2015, 2018; Freilich

et al., 2022; Gray et al., 2024). Modeling studies show that vertical

velocity inferred from surface observations can reduce uncertainty

in heat and carbon budgets (Su et al., 2018; Siegelman et al., 2020),

suggesting that reductions in biogeochemical uncertainty, includ

ing the magnitude of the biological carbon pump and of nutrient

fluxes, are possible as well.

Observations suggest that oceanographic small- to fine-scale

dynamics do not just alter the abundance of phytoplankton but

also result in shifts in community composition due to biogeo-

chemical forcing. For example, diatoms are thought to respond

to rapid vertical nutrient injection (Benitez-Nelson et al., 2007;

McGillicuddy, 2016), including upwelling that can occur within

the interiors of cyclonic eddies. Shifts in community composi

tion have implications for food web dynamics as well for biogeo-

chemical function; such shifts have been observed in focused field

campaigns (e.g., Benitez-Nelson et al., 2007; Baltar et al., 2010;

Nelson et al., 2014; Tzortzis et al., 2021). Modeling studies are less

conclusive regarding the impacts of fronts and eddies on com

munity composition. The signal of lateral redistribution (passive

stirring) appears to overwhelm that of a reactive response to verti

cal nutrient supply in some global studies (Martin and Pondaven,

2003; Clayton et al., 2013; Lévy et al., 2014), while other mod

eling studies argue for the predominance of reactive dynamics

(Marinov et al., 2010; Levy and Martin, 2013; Clayton, 2017a).

While conventional ocean color sensors are multispectral, PACE

contains a hyperspectral instrument, opening the possibility to

infer phytoplankton pigment composition and plankton com

munity structure (Chase et al., 2022; Kramer et al., 2022; Cetinić

et al., 2024) on global scales. This global-scale perspective is nec

essary to constrain the role of fronts and eddies in shaping com

munity composition.

The combination of SWOT and PACE products enables new

insights into physical processes operating on timescales relevant to

biological growth that previously were limited from space and that

cannot be resolved by either platform alone. Here, we present some

of the first coincident analyses of SWOT and PACE data to inves

tigate the response of community composition to physical forc

ing. We demonstrate the potential of combining SWOT and PACE

observations through a case study, conditional averaging of bio

logical conditions based on physical characteristics, and seasonal

scale temporal analysis.

DATA AND METHODS

SWOT

A primary advancement of the SWOT mission (launched in

December 2022) compared to previous sea surface altimeters is

the increased horizontal resolution made possible by the Ka-band

radar interferometer (KaRIn; Rodriguez et al., 2018). KaRIn makes

measurements across two parallel 50 km-wide swaths, provid

ing two-dimensional observations of ocean topography of sub-

100 km oceanic processes (Wang et al. 2025; Figure 1a) and rou

tinely capturing ocean eddies at scales of a few kilometers (Archer

et al., 2025). SWOT’s two-swath design leaves a 20 km “nadir gap”

directly below the satellite, which is filled by a traditional altimeter

that provides pointwise measurements (dots in Figure 1a).

In contrast to the high-resolution swaths of SWOT, DUACS

(Data Unification and Altimeter Combination System), an oper

ational multi-mission altimeter data system developed by the

Oceanography | Vol. 39, No. 1

French national space agency (CNES), combines along-track

measurements from multiple nadir satellite altimeters into grid

ded, daily sea level fields. The current operational DUACS prod

ucts are gridded at ¼° (~25 km) resolution through mapping and

interpolation, which smooths variability at smaller scales. As a

result, DUACS products are optimized for high temporal cover

age, whereas SWOT enables the observation of finer-scale variabil

ity at the expense of less frequent sampling.

Sea level anomaly (SLA) and absolute dynamic topography

(ADT) from two different SWOT products are presented here.

Figure 1 uses a Level 3 (L3; version 3.0) product of individual

swaths on a 2 km × 2 km grid. The data presented in the rest of

the manuscript use the provisional Level 4 (L4; version 2.0) prod

uct Multiscale Interpolation Ocean Science Topography (MIOST),

which provides gridded fields by incorporating nadir altime-

try satellite data from DUACS with SWOT KaRIn measurements

(Ballarotta et al., 2025; Ubelmann et al., 2021). The MIOST prod

uct consists of daily gridded products at 1/10° (~10 km) spatial res

olution at global scale and is under provisional status, meaning it

is publicly available but still undergoing validation and refinement.

The MIOST product is generally too coarse to resolve inertia-

gravity waves, meaning that the sea surface height gradients pre

dominantly reflect geostrophic currents. MIOST data are accessed

from the AVISO+ server, supported by CNES. SWOT L3 data are

also available and documented on NASA EarthData.

The Rossby number (Ro) is a non-dimensional number that

represents the ratio of inertial forces to Coriolis forces, indicating

whether a fluid flow is dominated by Earth’s rotation (low magni

tude Ro) or by its own inertia (high magnitude Ro). In this study,

Ro is used to quantify the strength and polarity (sign of rotation) of

mesoscale circulation and to classify ocean regions into cyclonic,

anticyclonic, and background flow regimes. Ro is calculated as

Ro = ζ/f = (vx − u y)/f,

where u and v are the zonal and meridional geostrophic velocities,

respectively, and f is the Coriolis parameter. Subscripts represent

partial derivatives. The assumption of geostrophy breaks down

when the Rossby number approaches O(1), so the Rossby num

bers calculated here are only estimates.

PACE

PACE (launched in February 2024) introduced a number of

advances for research across the hydrosphere, biosphere, and

atmosphere (Gorman et al., 2019; Werdell et al., 2024). In the con

text of ocean biology, the hyperspectral nature of the Ocean Color

Instrument (OCI) allows us to derive different phytoplankton pig

mentations from spectral remote sensing reflectances, not only for

the calculation of chlorophyll a concentrations but also for first-

order derivation of community composition (Cetinić et al., 2024).

PACE L3 mapped products are provided with a horizontal resolu

tion of 4 km and 1/10°.

For this study, PACE data is accessed from the Ocean Biology

DAAC L3 and L4 browser. These data, including derived products,

are also available and documented on NASA EarthData. The PACE

OCI product for chlorophyll a is currently in provisional status.

The Multiple Ordination ANAlysis (MOANA) algorithm

takes advantage of the hyperspectral ocean color radiometry of

PACE-OCI and returns near-surface concentrations of three dif

ferent phytoplankton groups in cells mL–1: Prochlorococcus,

Synechococcus, and autotrophic picoeukaryotes (Figure 1c). The

approach, described in Lange et al. (2020), uses a principal com

ponent analysis and multiple regression, requiring inputs of

FIGURE 1. A mesoscale eddy is shown

as measured by the NASA Surface

Water and Ocean Topography (SWOT)

satellite and the NASA Plankton, Aero

sol, Cloud, ocean Ecosystem (PACE)

satellite between March 28 and April 1,

2024: (a) sea level anomaly (SLA in m)

as measured by the Data Unification

and Altimeter Combination System

(DUACS; background grid) and SWOT

(swaths), (b) chlorophyll a as measured

by PACE with SLA SWOT contours, and

estimated by the PACE Multiple Ordina

tion ANAlysis (MOANA) algorithm with

SLA SWOT contours.

37°N

36°N

35°N

34°N

33°N

32°N

1.00

0.75

0.50

0.25

0.00

–0.25

–0.50

–0.75

–1.00

SLA (m)

75°W

74°W

73°W

34.5°N

34.0°N

33.5°N

34.5°N

34.0°N

33.5°N

0.40

0.35

0.30

0.25

0.20

0.15

0.10

12,000

10,000

8,000

6,000

4,000

2,000

Chlorophyll-a (mg m–3 )

Picoeukaryotes (cells mL–1 )

74.0°W

73.5°W

73.0°W

March 2026 | Oceanography

remote- sensing reflectance collected by PACE-OCI. Sea surface

temperature is also used for the calculation of Prochlorococcus con

centrations, but not for the other phytoplankton groups. MOANA

is in provisional status, so currently is only produced for and val

idated in the Atlantic sector of the global ocean. Additional algo

rithms for delineating community composition from PACE-OCI’s

hyperspectral surface reflectance are also in development at NASA.

DATA MERGING

PACE and SWOT have fundamentally different global cover

age times due to their observing geometries: PACE’s wide-swath,

sun-synchronous orbit enables near-daily global coverage, whereas

SWOT’s narrower swath and interferometric sampling has a longer

effective revisit time (a global average of 11 days). Because our

focus is on mesoscale features rather than submesoscale structure

captured in swaths, we use the MIOST product, which sacrifices

some spatial resolution to provide more temporally consistent,

gap-free fields that are better suited for tracking the evolution of

mesoscale features. As a result, we are able to get near-daily over

laps between SWOT and PACE data.

For each day analyzed, we incorporate available PACE obser

vations within a ±2-day window, forming a five-day composite to

enhance spatial coverage while maintaining near-coincident sam

pling. Because ocean color observations are limited by cloud cover,

substantial spatial and temporal gaps in daily satellite measurements

occur. To mitigate these effects, we exclude data from the five-day

averages during which more than 50% of the region of interest is

cloud covered. This temporal averaging reduces noise associated

with intermittent cloud contamination and increases the number

of valid observations contributing to each estimate, while still pre

serving variability at the mesoscale. The five-day averaged observa

tions are then collocated onto the relevant sea surface altimetry grid

by averaging all measurements within each altimetry pixel.

REGIONAL STUDY SETTING

This study focuses on analyzing data from 63°–75°W, 30°–40°N in

the North Atlantic Ocean (between April and November 2024. This

region was chosen for its energetic mesoscale variability associated

with the Gulf Stream and its frequent shedding of cyclonic cold-core

eddies into the subtropical gyre. These eddies, typically O(100 km)

in diameter, form from Gulf Stream meanders and propagate

southwestward, transporting relatively nutrient-rich and biolog

ically distinct Slope Sea waters into the oligotrophic subtropical

Atlantic (The Ring Group, 1981; McGillicuddy et al., 1998).

The region encompasses a strong meridional gradient in physi

cal and biogeochemical properties across the Gulf Stream that sep

arates cooler, more productive waters to the north from warmer,

nutrient-poor waters to the south. Cyclonic eddies in this region are

associated with depressed sea surface height and positive relative

vorticity (Gaube and McGillicuddy, 2017). The April–November

time frame captures the transition from the spring bloom through

stratified summer conditions and into early fall. This seasonal win

dow allows for investigation of both episodic (subseasonal) vari

ability associated with individual eddies and broader seasonal

shifts in biological response to physical forcing.

To focus on subtropical eddies and minimize contamination

from the Gulf Stream core, analyses are restricted to waters south

of the Gulf Stream, identified using an absolute dynamic topog

raphy (ADT) threshold of 0.3 m. Closed contours exceeding this

threshold are retained to preserve coherent eddy structures with

relatively low ADT that extend south of the Gulf Stream.

RESULTS

SWOT’s two-dimensional swath data are higher resolution than

the ¼° data from DUACS, so they capture submesoscale features

and provide finer resolution of sea level anomalies associated with

mesoscale eddies (Figure 1a). Examining swath data with collo

cated PACE observations demonstrates tight correspondence

between enhanced chlorophyll a concentration and the core of a

cyclonic eddy (Figure 1b).

Compared to the conventional DUACS product, the SWOT-

informed MIOST product also resolves features with larger mag

nitude Rossby numbers, indicating a greater prevalence of high-

magnitude vorticity (long tails; Figure 2a). This expanded range

reflects SWOT’s ability to resolve smaller-scale, higher-intensity

motions. In the context of the Gulf Stream region, SWOT captures a

FIGURE 2. These panels illustrate the

advancement of SWOT over DUACS for

capturing small-scale features. (a) Prob

ability density function (PDF) of the

Rossby number measured across the

study region from March 2024 through

October 2024. (b) Five-day average

(April 11–15, 2024) of derived concen

trations of chlorophyll a (mg m−3) from

PACE overlain by contours of Ro > 0.4

for both SWOT (red) and DUACS (yel

low). (c) Concentration of autotrophic

picoeukaryotes (cells mL−1) binned per

Ro for both SWOT (red) and DUACS

(yellow) south of the Gulf Stream.

Picoeukaryotes (cells mL–1 )

PDF (normalized)

Rossby Number

7,000

6,000

5,000

4,000

3,000

2,000

1,000

–0.9 –0.7 –0.5 –0.3 –0.1 0.1 0.3 0.5 0.7 0.9

SWOT

DUACS

75°W

70°W

65°W

40°N

35°N

30°N

Chlorophyll-a (mg m–3 )

100

10–1

SWOT; Ro > 0.4

DUACS; Ro > 0.4

Oceanography | Vol. 39, No. 1

wider range of physical features than DUACS (Figure 2b). Though

DUACS captures the larger coherent cyclonic mesoscale eddies, it

misses smaller submesoscale eddies. Even in eddies that are present

in both the SWOT and DUACS products, the contours of the eddy

edges (as defined by an SLA or the Rossby number threshold) are

smoother with SWOT than DUACS, and SWOT identifies larger

regions with high Rossby numbers, higher eddy core Rossby num

bers, and more elongated (anisotropic) features (Middleton et al.,

2025). In addition, within the Gulf Stream itself, DUACS cap

tures some cyclonic features; however, SWOT captures both more

cyclonic features and the elliptical natures of these features.

Binning observations of PACE-OCI derived community compo

sition products by Rossby number demonstrates that the extended

tails of the SLA distribution captured by SWOT and increased fidel

ity contribute biologically relevant information (Figure 2c). In the

areas of the Ro distribution with 90% of the data (–0.3 < Ro < 0.3),

the binned averages and ranges of chlorophyll a concentrations

and cell counts (e.g., picoeukaryote cells in Figure 2c) look simi

lar between both DUACS and SWOT. However, the greatest devi

ations in phytoplankton concentration are observed at the lowest

and highest Ro. The extreme Ro values observed only in the SWOT

dataset, –0.7 and 0.9, reveal the tightest biophysical coupling with

the lowest and highest phytoplankton concentrations, respectively.

At high Ro (Ro ≥ 0.5; cyclonic), we also observe a distinct signal

between DUACS and SWOT. Both products reveal that higher Ro

is associated with higher phytoplankton concentrations. However,

we observe generally lower concentrations of autotrophic pico-

eukaryotes for a given Ro in SWOT than in DUACS because

DUACS products underestimate Ro for a given feature compared

with SWOT. This result provides concrete evidence that submeso-

scale dynamics (Ro of order 1) are most significant for primary

producers and for the growth of autotrophic picoeukaryotes.

While prior literature has mostly focused on responses by fast

growing taxa such as diatoms to physical forcing, the combination

of PACE and SWOT observations suggests systematic variation in

photosynthetic cyanobacteria (Prochlorococcus and Synechococcus)

across physical features, in addition to the systematic enhancement

of autotrophic picoeukaryotes (Figure 3). In contrast to larger dia

toms, which can achieve maximum growth rates of ~1–2 day–1

under favorable conditions (Marañón, 2015), smaller cyanobacte-

ria such as Prochlorococcus and Synechococcus typically grow more

slowly, with rates on the order of ~0.2–0.7 day–1, depending on

light and nutrient availability (H. Liu et al., 1998). The highest con

centrations of Synechococcus (~20,000 cells mL–1; Figure 3c,d) and

the autotrophic picoeukaryotes (~3,000 cells mL–1; Figure 3e,f) are

aligned with features that exhibit strongly positive Ro, the cyclonic

eddies shed from the Gulf Stream or that have local upwelling in

their cores. The concentrations of these taxa in what could be con

sidered background waters without eddies (i.e., |Ro| < 0.1) and

in anticyclones, which are downwelling, are significantly lower

(~10,000 cells mL–1 for Synechococcus and ~1,000 cells mL–1 for

picoeukaryotes). The inferred concentration of Prochlorococcus

(Figure 3a,b) also appears to respond to eddies, with evidence of

elevated concentrations in both cyclonic and anticyclonic features.

FIGURE 3. Derived plankton concen

trations from the MOANA algorithm

applied to PACE-OCI (June 1–5, 2024).

Concentrations (cells per mL) are plot

ted per binned Rossby number for

(a) Prochlorococcus, (c) Synechococ-

cus, and (e) picoeukaryotes. Points in

(a), (c), and (e) show the five-day aver

aged points and are not binned. Snap

shots (June 1–5, 2024) are shown of

derived plankton concentrations in the

region of consideration for (b) Prochlo-

rococcus, (d) Synechococcus, and

(f) picoeukaryotes. Red contours indi

cate locations of Ro > 0.4 as mea

sured by SWOT, and areas north of

the Gulf Stream (as determined by an

ADT > 0.3 m unclosed contour thresh

old) are masked.

Prochlorococcus (cells mL–1 )

75°W

70°W

65°W

40°N

35°N

30°N

40°N

35°N

30°N

40°N

35°N

30°N

Picoeukaryotes (cells mL–1 )

Synechococcus (cells mL–1 )

400,000

300,000

200,000

5,000

4,000

3,000

2,000

1,000

30,000

20,000

10,000

Rossby Number

–0.8 –0.6 –0.4 –0.2 0.0 0.2 0.4

0.6 0.8

Picoeukaryotes

Synechococcus

Prochlorococcus

6,000

4,000

2,000

30,000

20,000

10,000

400,000

300,000

200,000