September 2025 | Oceanography
77
OBSERVATIONS
MODELS
DATA MANAGEMENT
COLLABORATION
Ocean Acidification (OA)
The absorption of ~25% of total anthropogenic CO2 emissions (Friedlingstein et al., 2023) has shifted ocean chemistry toward reduced pH and carbonate ion concentration,
with adverse consequences for marine life at all trophic levels, particularly calcifiers. Constraining OA trends on local/regional to global scales and developing predictive
capacity to inform decision-making requires cutting-edge technology for climate-quality measurements and advanced models.
• Observing gaps: Historical DIC and TA
data to validate hindcasts, subsurface
observations where big OA impacts occur,
organic alkalinity, calcification rates
• Develop high-quality, cost-effective CO2
system sensors, especially DIC and TA
• More careful QA/QC (w/climate- and
weather-quality guidance) of OA-relevant
BGC sensor data
• Targeted sampling of critical and historically
under-sampled ecosystems (e.g., polar
regions)
• Leverage existing observational
data to validate four-dimensional OA
model simulations
• Make model outputs and model-
based visualization systems more
accessible
• Assess data overlap and connectivity
(interoperability) across existing
specialized ocean carbon
repositories from regional to global
scale
• Develop centralized catalog of
OA-relevant data resources to
facilitate discovery and uptake by
stakeholders
• Promote cross-disciplinary
communications to create universal
OA terminology and identify
underutilized OA resources
Ocean Carbon Budget
The ocean is the largest reservoir of mobile carbon on our planet, containing 45 times more carbon than the atmosphere and 15 times more than land plants and soils
(Friedlingstein et al., 2023). Despite its importance to global climate dynamics and environmental policies, the ocean carbon budget is not thoroughly constrained. Equipping
the marine BGC observing system to produce useful data for carbon cycle models is critical for quantifying the role of the ocean in the global carbon cycle.
• More subsurface and coastal observations,
increased temporal resolution (hourly when
possible), use of autonomous assets to
reduce seasonal bias
• Integrate in situ and satellite-based ocean
carbon measurements
• Increase biological sampling to better
understand contributions of zooplankton,
viruses, microbes, etc., to ocean carbon
cycling
• Continued development of carbon system
sensors and platforms designed with easy
integration of those sensors
• Use models to inform temporal
and spatial sampling resolution
requirements in different regions
(i.e., OSSEs)
• Conduct quantitative model-data
comparison studies and uncertainty
analyses of ocean carbon uptake,
transport, and storage
• Need tools to streamline access and
search for synthesis and modeling
in data from platforms like Argo
(e.g., searchable dashboards that
filter data by parameter, standardized
entrance points for accessing all
available observations)
• Standardized metadata reporting,
especially data uncertainty source
and quantification
• Collaboration between data
managers and scientists to design
FAIR data systems
• Centralized library of relevant tools,
data products, and content creators
to promote collaboration
• Workshops to support uptake
and processing of observations,
e.g., BGC-Argo
• Ongoing investment in training,
knowledge exchange, and funding
opportunities for data-model
integration
Polar Systems
Polar regions are crucial to Earth’s BGC cycles, profoundly affecting climate and marine ecosystems. Recent advancements in observing platforms and technologies have
mitigated some of the challenges associated with data collection in these remote and often extreme environments. Despite this progress, there is an urgent need for more
model-data integration to explore complex interactions and feedbacks in these ecosystems.
• Geographic: more coastal/shelf and under-
ice observations, expanded coverage in the
Southern Ocean (beyond the Weddell Sea
and Drake Passage) and the Arctic (Eurasian
data)
• Temporal: reduce seasonal bias, high-
resolution/continuous coverage across
transitions like seasonal ice melt, more time
series programs
• Integration of in situ, autonomous, and
remote observations
• Increased resolution of polar
observations in space and time
through logistical and technical
innovation (platforms, sensors,
vehicles, etc.) that overcome
environmental challenges
• Develop/improve AI methods to
address contrasting data vs. model
resolution
• Develop unified databases with
advanced search capabilities to
enhance data discoverability and
usability
• Standardize data collection and data/
metadata reporting guidelines to
improve interoperability
• Standardize inclusion of funding in
projects for implementing FAIR data
practices that ensure accessibility of
both new and historical data sets
• Capacity building activities
(e.g., hackathons) on data uptake,
assimilation, synthesis, data-model
integration
• Opportunities for sustained dialog
among data managers, observers,
and modelers (e.g., community
workshops, project meetings,
webinars)
Trophic Interactions
Studying trophic interactions requires a holistic understanding of marine food webs, from physics to plankton to marine mammals. Marine trophic interactions span several
temporal and spatial scales, making them a complex system to observe and model.
• Biological rate measurements
• Improved biological baselines through
sustained time series programs to detect
patterns of variability and change
• Imaging data to measure community
composition and size distribution to better
constrain trophic transfer
• Resource management and stakeholder
needs should guide sampling
• Support goal-oriented pre-
observation communication among
all involved stakeholders to improve
model pipeline for complex and
under-observed trophic interactions
• Overall need for discoverable data
in common formats for uptake and
aggregation, including:
• Standardized reporting of variable
names and formats
• Clear descriptions of data type,
included variables, and access
point(s)
• Adoption of standard protocols for
data manipulation, re-uploading, and
metadata
• Cross-training activities to build
mutual awareness of data collection
and modeling challenges
TABLE 1. Continued…