Ocean health variables · 170 variables

What is known about the state of the ocean

Each variable carries either a measured value or a stated reason why no value exists. For 20 variables nothing is measured anywhere in the world — that is a finding about the state of ocean science, not a missing entry.

Measured: 15.7 % of index weight
Global value exists: 39.3 % of index weight
Local scale only: 33.6 % of index weight
Not measured anywhere: 11.4 % of index weight

Why the gaps exist

Weighted by variable significance, not counted by number of variables.

18.6%Data exist but are not comparable · 30 variables
16.9%No observation network · 27 variables
8.7%Technical barrier to measurement · 14 variables
8.3%Data exist but are not public · 17 variables
7.5%Data exist but the sample is biased · 9 variables
6.0%A global average would be meaningless · 7 variables

Ecosystem Assets

1.1 · Coastal-marine vegetated systems

1.1.1.1Cover areahaGlobal value exists0.46%
14,725,600 ha (147,256 km²) of mangrove; the rate of loss fell by 44% between 2000 and 2020 · 2020 (publ. 2024)
Global Mangrove Watch v4.0; Global Mangrove Alliance, State of the World's Mangroves 2024; FAO, The World's Mangroves 2000–2020
1.1.1.2Canopy closure%Local scale only0.46%
No global figure. Field surveys of healthy stands: 60-90%
No observation network
Collected locally, never globally. Closure has to be measured from beneath the canopy (fisheye photography) or by drone; satellites reliably give extent and canopy height but not closure. No shared method and no observation network exist, only scattered research.
Simard et al. 2019 (Global Mangrove Canopy Height, NASA ORNL DAAC); local surveys
1.1.1.3Productivityt/haGlobal value exists0.46%
No direct global productivity figure. The closest global proxy is carbon stock: 394 tC/ha (biomass plus the top metre of soil), exceeding 650 tC/ha in some regions · 2024
No observation network
Collected locally in individual studies (litterfall traps, stem increment). No global programme measures mangrove productivity: it requires multi-year plots and cannot be scaled by satellite.
Global Mangrove Alliance, State of the World's Mangroves 2024
1.1.1.4Degradation share%Measured0.46%
About 50% of the world's mangrove provinces are classified as threatened (IUCN Red List of Ecosystems); 43% of losses in 2000-2020 were directly human-driven (aquaculture, oil palm, rice) · 2024
Data exist but are not comparable
Observations exist, but degradation is not operationally defined: cleared forest is visible from space, thinned or stressed forest is not. No agreed international threshold for mangrove degradation exists, so no global percentage is computed.
IUCN Red List of Mangrove Ecosystems 2024; FAO 2023
1.1.2.1Meadow areahaGlobal value exists0.75%
About 16,000,000 ha (160,000 km²) of seagrass mapped; full extent estimated at 300,000-600,000 km². Roughly 30% of known extent has been lost since the late 19th century, and losses continue · 2020
UNEP, Out of the Blue: The Value of Seagrasses to the Environment and to People (2020); Waycott et al. 2009 (PNAS)
1.1.2.2Shoot densityshoots/m2Local scale only0.75%
No global figure. Range from local surveys: roughly 50-1,500 shoots/m² depending on species (Zostera, Posidonia, Thalassia)
A global average would be meaningless
Collected locally by diver transects. A global aggregate is impossible in principle: shoot density differs by an order of magnitude between Zostera and Posidonia, so an average across species carries no biological meaning. No global seagrass monitoring network exists at all.
local surveys; UNEP, Out of the Blue: The Value of Seagrasses (2020) — review, no global mean
1.1.2.3Growth depthmLocal scale only0.75%
No global figure. Lower limit typically 5-40 m, governed by water clarity
A global average would be meaningless
Collected locally. Depth limit is a function of clarity at the specific site. Monitoring exists only in some EU states under the Marine Strategy Framework Directive.
local surveys; Duarte 1991 (depth limit from light penetration)
1.1.3.1Forest areahaGlobal value exists0.68%
About 146,900,000 ha (1.469 million km²) of kelp forest by model estimate; global decline around 1.8% per year (median across regions) · 2020 / 2016
Jayathilake & Costello 2020 (Biological Conservation); Krumhansl et al. 2016 (PNAS)
1.1.3.2BiomasstReference only, no weight
No global biomass estimate. Locally 1-20 kg wet mass per m² depending on species and region
No observation network
Collected locally by harvest quadrats. No global network exists, and seasonality (biomass varies severalfold within a year) makes even local data hard to compare across studies. Satellites see kelp only at the surface and only at low tide.
Krumhansl & Scheibling 2012; local surveys
1.1.3.3Distribution depthmLocal scale only0.68%
No global figure. Typically 2-30 m, reaching 60-80 m in very clear water
No observation network
Collected locally by diver or ROV survey. No global aggregate exists or is planned.
local surveys

1.2 · Hard substrates and reefs

1.2.1.1Live cover%Measured0.75%
Mean live hard coral cover about 29-30% (a 14% decline over 2009-2018); cover has roughly halved since the 1950s. Caribbean: down 48% between 1980 and 2024 · 2019 globally / 2024 Caribbean
GCRMN/ICRI, Status of Coral Reefs of the World: 2020 (publ. 2021); GCRMN Caribbean 1970–2024 (2025); ICRI 2025
1.2.1.2Bleaching level%Measured0.75%
84% of the world's reef area was affected by the fourth global bleaching event, the most extensive on record · 2023–2025
NOAA Coral Reef Watch; ICRI, 4th Global Bleaching Event (2025)
1.2.1.3Species diversity (Shannon index)indexLocal scale only0.75%
No global aggregate. Indirectly: in the Caribbean the community has shifted from branching to massive forms, with macroalgae up 85% since 1980
Data exist but are not comparable
Data are collected globally within GCRMN surveys, but no aggregate is published: the index depends on survey area and taxonomic resolution, and sites are not randomly chosen.
GCRMN Caribbean 1970–2024 (2025) — regional trend; local surveys
1.2.2.1Colony area (absence of necrosis)m2Local scale only0.55%
No global aggregate. The key observation from the best-studied province, the Blake Plateau in the Atlantic: 83,908 individual coral mounds mapped, with one near-continuous stretch of about 400 km built mainly of DEAD coral — the mounds themselves are rubble of Desmophyllum pertusum, with living cover forming only a thin veneer on top · 2024
Technical barrier to measurement
Requires ROV or submersible survey at 200-1,000 m: the Blake Plateau received 23 dives over more than a decade across an area the size of Florida. Such a sample cannot be extrapolated to a global figure.
Sowers et al. 2024 (Geomatics, 'Mapping and Geomorphic Characterization of the Vast Cold-Water Coral Mounds of the Blake Plateau'); NOAA Ocean Exploration
1.2.2.2Skeleton condition (structural integrity)conv.unitsLocal scale only0.55%
No global index. The threat mechanism for cold-water coral is precisely established and tied to a planetary boundary already crossed: below an aragonite saturation state of 1 the dead carbonate framework dissolves, and that framework is the load-bearing structure of the mound. The saturation horizon is shoaling; models project that most known cold-water coral habitat will sit in undersaturated water by 2100 · 2006 (projection) / 2025 (Ω)
Technical barrier to measurement
Assessing framework integrity requires either sampling by submersible or repeat high-resolution bathymetry years apart. No country runs a regular programme; existing data are a by-product of exploratory cruises.
Guinotte et al. 2006 (Frontiers in Ecology and the Environment); Planetary Health Check 2025 (PIK); Sowers et al. 2024 (Geomatics)
1.2.2.3Depth of occurrencemLocal scale only0.55%
Reef-building cold-water coral typically occurs at 200-1,000 m in water of about 4 °C (full known range from a few tens of metres in Norwegian fjords to over 3,000 m). The largest mapped province is the Blake Plateau: a continuous mound belt about 500 km long and up to 110 km wide, core area about 6,215 km², maximum outline about 26,000 km². No global time series of change in depth exists · 2024
Technical barrier to measurement
Technical barrier: multibeam survey at 200-1,000 m at resolution sufficient to resolve mounds requires a research vessel, and nobody funds repeating it years later. Note also that the province area depends on the chosen mound-density threshold — 6,215 km² for the core against 26,000 km² for the maximum outline, a fourfold difference. The same problem as with garbage patch extent: the number is set by the threshold, not by the measurement.
Sowers et al. 2024 (Geomatics); NOAA Ocean Exploration, Blake Plateau mapping (January 2024); Hydro International (2025)
1.2.3.1Settlement densityind/m2Not measured anywhere0.82%
No global data. Locally 0-50 recruits/m²
No observation network
Collected locally and rarely. It requires deploying settlement plates and recovering them months later — laborious, and done only in individual research projects. No national programmes, no global network.
—
1.2.3.2Reef heightmNot measured anywhere0.82%
No global data
Technical barrier to measurement
Barely collected. Reef height requires high-resolution bathymetric survey repeated over years; such data exist for a handful of reefs worldwide, mostly the Great Barrier Reef.
—
1.2.4.1Number of modules installedunitsReference only, no weight
No global register of artificial reefs exists
No observation network
Not collected globally or in most countries. Artificial reefs are installed by municipalities, dive centres, fishing cooperatives and private foundations; no register exists even nationally, apart from Japan and South Korea.
—
1.2.4.2Submersion ageyearsReference only, no weight
No global data
No observation network
Not collected. The deployment date is known only to whoever installed the structure; there is no central record. The data sit in private archives and are never aggregated.
—
1.2.4.3Fouling degree%Not measured anywhere0.34%
No global data
No observation network
Not collected. Requires repeat diver survey of the specific module; done only within the projects themselves and usually unpublished.
—

1.3 · Soft substrates and pelagic zone

1.3.1.1LengthkmGlobal value exists0.48%
About 31% of the world's ice-free shoreline is sandy beach (roughly 250,000-360,000 km against a total coastline of about 1.1 million km) · 2018
Luijendijk et al. 2018 (Scientific Reports, 'The State of the World's Beaches')
1.3.1.2Beach widthmGlobal value exists0.48%
No global mean. Trend: 24% of the world's sandy beaches are eroding faster than 0.5 m/year, 28% are accreting · 2018
Data exist but are not comparable
The trend is global, the absolute width is not. Width requires locating the dune or vegetation line, which satellites resolve poorly and inconsistently across methods, so shoreline change rate is published instead of width.
Luijendijk et al. 2018 (Scientific Reports)
1.3.1.3Grain sizemmNot measured anywhere0.48%
No global data
No observation network
Not collected globally. Grain size requires physical sampling and sieving; satellites cannot see particle size. Only national and local datasets exist, often held in closed engineering survey archives.
—
1.3.2.1Exposed area at low tidehaGlobal value exists0.61%
12,792,100 ha (127,921 km²) of tidal flats; 16% of the area was lost between 1984 and 2016 · 2016 (publ. 2019)
Murray et al. 2019 (Nature, 'The global distribution and trajectory of tidal flats')
1.3.2.2Mud thicknessmNot measured anywhere0.61%
No global data
Data exist but are not public
Not collected systematically. Mud thickness is measured by drilling or sub-bottom profiling during engineering surveys; results stay in client reports and never reach open databases. No scientific monitoring exists.
—
1.3.2.3Organic matter content%Local scale only0.61%
No global mean for tidal flats. Marine sediments generally run 0.5-5% total organic carbon
No observation network
Collected locally by sampling. Flats are too narrow a strip to resolve in global grids, so they are not separated out.
Seiter et al. 2004 (global TOC distribution); local samples
1.3.3.1Photic zone boundarymGlobal value exists1.33%
Mean euphotic depth in the open ocean about 60-80 m (up to 120-150 m in oligotrophic subtropical gyres, 10-30 m nearshore) · continuous (satellite)
NASA OceanColor (MODIS/VIIRS Zeu, Lee et al. algorithm)
1.3.3.2Upwelling intensityconv.unitsLocal scale only1.33%
No single global value. Upwelling indices (Bakun, CUTI, BEUTI) exist for the four major upwelling systems · continuous (regional)
A global average would be meaningless
Collected regionally and continuously, but only for four systems. A global aggregate is meaningless in substance: upwelling is a local phenomenon, the index is undefined outside these zones, so a world average carries no information.
NOAA/PFEL Bakun Upwelling Index; CUTI/BEUTI (Jacox et al. 2018)
1.3.4.1Seafloor relief (absence of disturbance)conv.unitsGlobal value exists0.75%
No direct index. The best global proxy for disturbance: about 14% of the world's continental shelf area is touched by bottom trawling · 2018
Data exist but the sample is biased
The proxy is computed globally but incompletely: coverage depends on VMS and AIS data, which countries release differently. Asian and African data are fragmentary, so 14% is almost certainly an underestimate. No direct index of seafloor relief with repeat surveys exists.
Amoroso et al. 2018 (PNAS, 'Bottom trawl fishing footprints')
1.3.4.2Sediment type (naturalness)conv.unitsNot measured anywhere0.75%
Global maps of sediment type exist; an index of naturalness does not
Data exist but are not comparable
Maps exist, an index does not. Naturalness requires a pre-disturbance reference that does not exist for most of the shelf: systematic seabed mapping began after industrial trawling, so the baseline is gone.
dbSEABED (Jenkins); EMODnet Geology (Europe)
1.3.4.3DepthmGlobal value exists0.75%
Mean ocean depth 3,682 m; maximum 10,935 m (Challenger Deep) · 2023 (GEBCO 2023)
GEBCO 2023 Grid; Seabed 2030
1.3.5.1Canyon lengthkmGlobal value exists0.61%
9,477 large submarine canyons inventoried worldwide; mean length about 41 km · 2011
Harris & Whiteway 2011 (Marine Geology, 'Global distribution of large submarine canyons')
1.3.5.2Depth rangemGlobal value exists0.61%
Typical range: incision from 100-200 m at the shelf break down to 3,000-4,000 m; mean vertical relief about 2,000 m · 2011
Harris & Whiteway 2011 (Marine Geology)
1.3.5.3Turbidity current velocitym/sLocal scale only0.61%
No global figure. Direct measurements: 0.5-2 m/s typically, up to 5-8 m/s in the Congo Canyon · 2020
Technical barrier to measurement
Barely collected. Direct measurement requires instruments inside an active canyon, and the flow frequently destroys them; only a few successful measurement series exist worldwide. No global network exists, and none will for technical reasons.
Talling et al. 2022; Azpiroz-Zabala et al. 2017 (Congo Canyon)

1.4 · Ecosystem protection

1.4.1.1MPA areakm2Measured0.82%
About 36,100,000 km² (10.01% of the ocean) in protected and conserved marine areas; more than 16,600 MPAs · April 2026
UNEP-WCMC & IUCN, World Database on Protected and Conserved Areas / Protected Planet (Apr. 2026)
1.4.1.2Share of EEZ%Measured0.82%
23.05% of waters within national jurisdiction (EEZ) are under protection; in areas beyond national jurisdiction, only 1.66% · 2026
Protected Planet, Marine Protected and Conserved Areas (2026); UNEP-WCMC press release 01.04.2026
1.4.1.3No-take zoneskm2Measured0.82%
About 11,900,000 km² (3.3% of the ocean) is fully or highly protected; WWF assesses 2.8% as effectively protected · March 2026
Marine Conservation Institute, MPAtlas (March 2026); WWF Protecting Our Ocean report

Water Column State

2.1 · Physical indicators

2.1.1.1SST°CMeasured1.02%
Anomaly of +0.49 °C against the 1981-2010 baseline (2025 was the third warmest year on record); absolute mean about 18 °C. Ocean heat content in the top 2,000 m set a record, rising 23±8 ZJ in the year · 2025 (publ. Jan. 2026)
Pan et al. 2026, Advances in Atmospheric Sciences (doi:10.1007/s00376-026-5876-0); NOAA NCEI Global Climate Report 2025
2.1.1.2Bottom temperature°CGlobal value exists1.02%
Abyssal bottom temperature about 0.5-2 °C below 4,000 m; the deep ocean is also warming
Technical barrier to measurement
Data are collected but the network is thin. Argo profiles routinely to 2,000 m; Deep Argo, reaching 6,000 m, comprises roughly 200 floats for the whole ocean against about 4,000 standard ones. Bottom temperature is therefore known globally as a climatology, not as a current value.
World Ocean Atlas 2023 (NOAA NCEI); Deep Argo
2.1.1.3Thermocline positionmGlobal value exists1.02%
No absolute global value; the global trend is a 5.3% strengthening of upper-200 m stratification over 1960-2018, with the thermocline shoaling · 2020
Data exist but are not comparable
The trend exists, an absolute global value does not. Thermocline position is defined differently across studies (by gradient, by isotherm, by mixed layer depth) and varies strongly by season, so no single global number is published.
Li et al. 2020 (Nature Climate Change, 'Increasing ocean stratification over the past half-century')
2.1.2.1PSUPSUMeasured1.23%
Mean about 34.7 PSU; the contrast between salty and freshened regions is intensifying as the water cycle strengthens · continuous (Argo)
Argo / World Ocean Atlas 2023; Cheng et al. 2020 (salinity contrast)
2.1.2.2Halocline depthmLocal scale only1.23%
No global value. The measure matters primarily for the Arctic and the Baltic
A global average would be meaningless
Collected regionally. A halocline is pronounced only in the Arctic, the Baltic and freshened gulfs; across most of the ocean it does not exist, so the measure is undefined globally.
regional observations (Arctic Report Card, NOAA)
2.1.3.1Sigma-tkg/m3Global value exists1.43%
Surface sigma-t about 22-27 kg/m³ (density 1,022-1,027 kg/m³); at depth up to about 27.9 · continuous (Argo)
Argo / World Ocean Atlas 2023 (computed per TEOS-10)
2.1.3.2Buoyancy frequency (Brunt-Vaisala)conv.unitsReference only, no weight
No absolute global value; the trend is a 5.3% strengthening of upper-200 m stratification over 1960-2018 · 2020
Li et al. 2020 (Nature Climate Change)
2.1.4.1Secchi disk depthmLocal scale only0.82%
No global mean. Open ocean 20-45 m, nearshore 2-10 m. The global proxy is satellite Kd490, which shows rising turbidity in coastal zones · continuous (satellite proxy)
No observation network
Direct measurement has largely ceased. The Secchi disk is a 19th-century method displaced by optical sensors and satellites; historical series were gathered irregularly by ships with different observers, and global trends derived from them are disputed. No observation network exists today.
NASA OceanColor Kd490; historical Secchi series (Boyce et al. 2010, disputed)
2.1.4.2TurbidityNTULocal scale only0.82%
No global aggregate in NTU. Coastal turbidity is tracked by satellite as Kd490 or total suspended matter
Data exist but are not comparable
Collected locally and never aggregated. NTU is instrument-dependent, and no international conversion from satellite quantities exists without local calibration.
Copernicus Marine (Ocean Colour TSM); local measurements
2.1.5.1Current speedm/sGlobal value exists0.67%
Mean surface current speed about 0.1-0.25 m/s; western boundary currents (Gulf Stream, Kuroshio) 1-2 m/s · continuous
NOAA Global Drifter Program; OSCAR / Copernicus Marine GlobCurrent
2.1.5.2DirectiondegreesGlobal value exists0.67%
No global value exists — this is a vector field · continuous
A global average would be meaningless
Data exist, but the measure does not reduce to a scalar. Direction is a vector; an ocean-wide average is mathematically meaningless, since opposing currents cancel to zero.
OSCAR / Copernicus Marine
2.1.5.3Wave heightmGlobal value exists0.67%
Mean significant wave height about 2.0-2.5 m; extremes (99th percentile) rising by roughly 0.4% per year · 2019
Young & Ribal 2019 (Science, 'Multiplatform evaluation of global trends in wind speed and wave height')
2.1.5.4Wave energykW/mGlobal value exists0.67%
Mean wave power flux about 10-20 kW/m (up to 60-70 kW/m in the Southern Ocean); global wave resource around 2 TW, growing about 0.4% per year · 2019
Reguero et al. 2019 (Nature Communications, 'A recent increase in global wave power')

2.2 · Chemical indicators

2.2.1.1pHlog scaleMeasured1.54%
Surface pH about 8.05 against roughly 8.17 pre-industrial — a decline of about 0.1 units, equivalent to a 30% rise in hydrogen ion concentration. Current rate about -0.017 to -0.019 units per decade · 2024–2025
NOAA PMEL Ocean Acidification Program; GLODAPv2 / SOCAT; Copernicus Marine Ocean State Report; IPCC AR6 WG1 Ch.5
2.2.1.2Aragonite saturation (Omega)OmegaMeasured1.54%
Omega = 2.84 against a planetary boundary of 2.86 and a pre-industrial value of 3.44 — the boundary has been formally crossed for the first time; declining at about -0.08 per decade · 2025
Planetary Health Check 2025 (PIK); Findlay et al. 2025 (Global Change Biology, doi:10.1111/gcb.70238)
2.2.2.1Concentrationmg/LMeasured0.82%
Surface 5-9 mg/L depending on temperature and salinity; the global ocean oxygen inventory has fallen by about 2% since 1960 · 2017 (trend)
Schmidtko et al. 2017 (Nature, 'Decline in global oceanic oxygen content'); World Ocean Atlas 2023
2.2.2.2Saturation%Measured0.82%
Surface waters about 95-105% saturation; minima in oxygen minimum zones below 5%; the volume of water below 70 µmol/kg is expanding · 2017
Schmidtko et al. 2017 (Nature); Breitburg et al. 2018 (Science)
2.2.2.3Dead zone areakm2Global value exists0.82%
About 245,000 km² affected by hypoxia; more than 700 documented eutrophic and hypoxic sites, up from roughly 400 in 2008 · 2008 / 2011
No observation network
Observations are made locally but the global synthesis is not maintained. Hypoxia is recorded by national programmes (USA, EU, China), but no international body consolidates them — the 2008 synthesis was a one-off researcher initiative. Satellites cannot see hypoxia: it is a near-bottom phenomenon.
Diaz & Rosenberg 2008 (Science); WRI Eutrophication & Hypoxia database (2011)
2.2.3.1NitratesµMGlobal value exists0.68%
Surface: below 0.1 µM in subtropical gyres, 5-25 µM in upwelling zones and high latitudes; at depth (below 1,000 m) 30-40 µM · 2023 (WOA23)
World Ocean Atlas 2023 (NOAA NCEI); GLODAPv2
2.2.3.2PhosphatesµMMeasured0.68%
Surface 0.01-0.5 µM; at depth 2.0-3.0 µM · 2023 (WOA23)
World Ocean Atlas 2023 (NOAA NCEI); GLODAPv2
2.2.3.3N/P ratioratioReference only, no weight
The canonical Redfield ratio is 16:1; the observed deep-water value is about 14.7:1, and surface values vary widely
Redfield 1958; World Ocean Atlas 2023 (computed)
2.2.4.1DOCmg/LGlobal value exists0.92%
Surface 60-80 µM C (about 0.7-1.0 mg/L); at depth 34-45 µM C (about 0.4-0.5 mg/L). The global dissolved organic carbon stock is about 662 Pg C, comparable to atmospheric CO₂
Technical barrier to measurement
The spatial picture exists, the trend does not. Measuring DOC requires shipboard sampling and laboratory analysis; repeat sections are too few and too infrequent (every 5-10 years under GO-SHIP) to resolve a global temporal trend.
Hansell et al. 2009, 2021 (Global DOC database, NCEI); CLIVAR/GO-SHIP
2.2.4.2POCmg/LGlobal value exists0.92%
Surface 1-10 µM C (about 0.01-0.12 mg/L); a continuous global satellite POC product is available · continuous (satellite)
NASA OceanColor (MODIS/VIIRS POC); Stramski et al. 2008
2.2.5.1Biomass stockt C/haGlobal value exists1.23%
Mangrove: 394 tC/ha on average (biomass plus the top metre of soil), exceeding 650 locally. Seagrass: about 140-330 tC/ha. Salt marsh: about 250 tC/ha · 2024
Global Mangrove Alliance 2024; Fourqurean et al. 2012 (Nature Geoscience); UNEP Blue Carbon assessments
2.2.5.2Sediment stockt C/haGlobal value exists1.23%
In mangrove systems the top metre of soil is included in the 394 tC/ha figure and typically accounts for 50-70% of the stock; for seagrass the soil stock dominates, exceeding 80% · 2024
Data exist but are not comparable
Data are collected but are not comparable. Carbon stock in sediment depends on sampling depth: some studies take 30 cm, others 1 m, others down to the base of the peat. No single international protocol exists, so consolidating into one global figure produces a severalfold spread.
Global Mangrove Alliance 2024; Fourqurean et al. 2012; Kauffman et al. 2020

Biological Resources

3.1 · Nekton (fish and large animals)

3.1.1.1SSB / SSBmsyratio (SSB/SSBmsy)Local scale only0.61%
No global sum of SSB is published. The operational global measure instead: 62.4% of assessed marine stocks are within biologically sustainable limits, down from 64.5% two years earlier. For individual stocks: Peruvian anchoveta about 11 million t (2025) · 2026 (2023 data)
Data exist but are not comparable
There is no global sum in substance: SSB is modelled separately for each stock, in different units and against different base years; adding them is invalid and FAO deliberately does not. For roughly 40% of the world's stocks there is not even enough data to assign a category.
FAO, The State of World Fisheries and Aquaculture (SOFIA) 2026; FAO State of Stocks Index (revised methodology 2025)
3.1.1.2Total biomass / Bmsyratio (B/Bmsy)Local scale only0.61%
No global sum of exploitable biomass exists. Model estimates of mesopelagic fish biomass differ by an order of magnitude (1-10 billion t). FAO projects that exploitable fish biomass will fall by more than 10% by 2050 in several regions under high emissions · 2026
No observation network
The global sum is not computed for the same reason as SSB. The mesopelagic zone — the largest component by biomass — is barely fished and barely monitored: there is neither a fishery generating statistics nor an observation programme.
FAO SOFIA 2026; Irigoien et al. 2014 (Nature Communications) — mesopelagic estimate
3.1.1.3Recruitmentmillion individualsLocal scale only0.61%
No global aggregate. Recruitment is estimated per stock in national assessments (ICES, NOAA, AFMA)
No observation network
Collected locally and only for stocks with a full assessment, roughly 10% of the world's stocks. It requires annual ichthyoplankton or juvenile trawl surveys, run by about 20 countries. Global aggregation is impossible: different species, units and year classes.
ICES Stock Assessment Database; NOAA Fisheries Stock SMART; FAO SOFIA 2026
3.1.1.4CPUE (catch per unit effort)kg/dayLocal scale only0.61%
No official global CPUE is published. Reconstructions show catch per unit effort in industrial fisheries falling roughly 80% since the 1950s; global fishing effort is tracked by AIS through Global Fishing Watch · 2018–2024
Data exist but are not comparable
Technically computable, but nobody publishes it as a global indicator: CPUE is not comparable across gears and species, and effort in vessel-days ignores rising technical efficiency (sonar, GPS). FAO deliberately does not consolidate CPUE into a global measure.
Global Fishing Watch; Sea Around Us (Pauly & Zeller); Rousseau et al. 2019 (PNAS) — global effort
3.1.2.1Demersalratio (B/Bmsy)Global value exists0.75%
No biomass figure; global demersal fish catch about 20-25 million t/year (out of 80 million t total marine capture in 2024) · 2024 (publ. 2026)
Data exist but the sample is biased
Catch is collected globally, biomass is not. Demersal biomass is estimated by trawl surveys run by about 25 countries on their own shelves; the open ocean and the shelves of Africa and South Asia are barely covered. Note that catch is not biomass — for a level basis this substitutes one quantity for another.
FAO FishStat / SOFIA 2026 (catches by group ISSCAAP)
3.1.2.2Pelagicratio (B/Bmsy)Global value exists0.75%
No biomass figure; small pelagic fish are the largest group in world catch at about 25-30 million t/year · 2024 (publ. 2026)
Data exist but the sample is biased
Same situation: global statistics exist for catch, not biomass. Pelagic stocks fluctuate naturally by multiples (El Niño), so even with estimates a global level would reflect a climate cycle rather than ecosystem condition.
FAO FishStat / SOFIA 2026; IMARPE (Peruvian anchoveta)
3.1.2.3Reef-associatedind/haReference only, no weight
No global figure. Reef Life Survey and similar programmes cover more than 3,000 sites worldwide but publish no single aggregate
Data exist but the sample is biased
Data are collected globally but no aggregate is published: fish density on a reef depends on depth, exposure and census method, and sites are chosen non-randomly (places accessible to divers), so an average is not representative of the world's reefs.
Reef Life Survey (Edgar & Stuart-Smith); GCRMN (for some nodes)
3.1.3.1Population sizeindividualsGlobal value exists0.82%
No single sum across marine mammals. By species: humpback whale about 135,000 (recovered), blue whale about 5,000-15,000, North Atlantic right whale about 370-380 (critically low). About 22% of marine mammal species sit in threatened IUCN categories · 2024–2025
A global average would be meaningless
No aggregate is published because summing individuals of different species is meaningless. Counts are species- and region-specific: the USA and EU run regular surveys, most of the tropics do not. A better measure for the platform would be the share of species with rising populations, not a sum of individuals.
IUCN Red List (2025 assessments); IWC Population Estimates; NOAA Fisheries Stock Assessment Reports
3.1.3.2Herd structure (calf share)%Not measured anywhere0.82%
No global data. Available for individual well-studied populations (for example the annual calf count for the North Atlantic right whale)
No observation network
Not collected systematically even locally: it requires either multi-year photo-identification of every individual in the population or aerial surveys able to distinguish calves. Worldwide this is done for a handful of populations — those close to extinction and given dedicated funding.
NOAA North Atlantic Right Whale Report Card; individual long-term projects
3.1.3.3Rangekm2Global value exists0.82%
Global areas-of-habitat maps exist for every assessed species in the IUCN database; ranges are shifting poleward by tens of kilometres per decade on average · 2025
No observation network
Maps exist, but they are not measurements. IUCN range maps are drawn by experts from accumulated encounters, updated every 5-10 years, and carry no explicit uncertainty. No systematic repeat surveys of marine mammal distribution exist.
IUCN Red List spatial data; AquaMaps; Poloczanska et al. 2013 (Nature Climate Change) — range shifts
3.1.4.1Nesting turtle numbers (females)individualsGlobal value exists0.68%
No global sum of nesting females. Global trend: most regional management units show rising nest counts; six of seven species remain in threatened IUCN categories · 2017 (trend) / 2025 (status)
Data exist but are not comparable
There is no absolute sum because what is counted is nests, not females, and the nests-per-female conversion is species- and region-specific (from 1.5 to 7). Many of the world's beaches are not patrolled at all — counting in West Africa and Southeast Asia is fragmentary.
Mazaris et al. 2017 (Science Advances, 'Global sea turtle conservation successes'); SWOT — State of the World's Sea Turtles database; IUCN Red List
3.1.4.2Clutch survival rate%Local scale only0.68%
No global mean. Hatching success typically 50-90% on protected beaches, noticeably lower on unprotected ones because of predators, trampling and inundation
No observation network
Collected locally by hundreds of projects but never aggregated: SWOT collects nest counts, not hatching success. The measure depends heavily on whether clutches were moved to hatcheries, so it reflects conservation intensity rather than environmental condition.
local project reports; SWOT (nest counts, no hatching success)
3.1.4.3Sea snake abundanceind/km2Not measured anywhere0.68%
No global data. IUCN has assessed about 70 sea snake species, a substantial share of them data deficient · 2025
Data exist but the sample is biased
Not collected globally or in most countries. Sea snakes have no commercial value (beyond local use in Asia), attract no conservation funding and are hard to census. The main source of data about them is trawl bycatch, so the sample is entirely biased.
IUCN Red List (Hydrophiinae assessments); Elfes et al. 2013

3.2 · Benthos (bottom-dwelling organisms)

3.2.1.1Bivalve stocktReference only, no weight
No global wild bivalve stock figure. Aquaculture production is known: world aquaculture reached 142 million t in 2024, of which molluscs account for roughly 17-20 million t · 2024 (publ. 2026)
No observation network
The wild stock is not assessed globally: bivalves live in shallow water where neither trawl surveys nor satellites reach, so assessment requires diver or dredge surveys site by site. National estimates exist in a handful of countries (USA, Netherlands, Japan).
FAO SOFIA 2026 (aquaculture and mollusc catches)
3.2.1.2Gastropod densityind/m2Reference only, no weight
No global data
No observation network
Not collected globally. Gastropod density is measured by quadrats on the intertidal — laborious and local. The closest large-scale initiative is citizen science (MarClim in the UK, rarely elsewhere); no global programme exists.
MarClim (UK); local surveys
3.2.1.3Cephalopod biomasstReference only, no weight
No global biomass figure. World cephalopod catch is about 3.5-4 million t/year; the long-term trend is increasing cephalopod abundance worldwide since the 1950s · 2024 / 2016 (trend)
Data exist but are not comparable
Biomass is not assessed globally: cephalopods live about a year and their abundance changes severalfold between seasons, so classical stock assessment models apply poorly. Even national estimates exist for only a few stocks (Argentine shortfin squid, Japanese flying squid).
Doubleday et al. 2016 (Current Biology, 'Global proliferation of cephalopods'); FAO FishStat
3.2.2.1Crab stocktReference only, no weight
No global stock figure. World crab catch is about 1.5-2 million t/year · 2024
Data exist but are not comparable
No global sum is computed: crab stocks are managed nationally, surveys are run by the USA, Canada, Norway, Russia and Japan on their own grounds, and methods are not comparable. For tropical crab fisheries no assessments exist at all.
FAO FishStat; NOAA Eastern Bering Sea Crab Survey
3.2.2.2Lobster abundanceindividualsReference only, no weight
No global abundance figure. World lobster catch is about 300,000-350,000 t/year · 2024
Data exist but are not comparable
Not collected in individuals as a matter of principle: lobsters are tracked through trap CPUE and tagging, which yield an index rather than an absolute count. Absolute estimates exist for two or three stocks worldwide (Gulf of Maine, Western Australia).
FAO FishStat; Maine DMR; Western Rock Lobster (Australia)
3.2.2.3Krill biomassg/m2Reference only, no weight
Antarctic krill: current best estimate about 379 million t circumpolar (range 60-420 million t); the 2019 survey of Area 48 gave 62.6 million t. The 2024 catch of about 500,000 t is 0.8% of biomass against a precautionary limit of 5.61 million t · 2019–2025
Data exist but are not comparable
The unit in the set, t/km³, does not match how the data are published (g/m² areal density, or million t per area), so conversion will require an assumption about the thickness of the inhabited layer. Krill is a rare case where a global estimate exists because a single international body (CCAMLR) has a survey mandate.
CCAMLR, Species Description: Euphausia superba (2024/2025); 2019 International Synoptic Krill Survey Area 48
3.2.3.1Sea cucumber densityind/haLocal scale only0.82%
No global density figure. It is established that most of the world's sea cucumber fisheries are overexploited or depleted; Holothuria was listed in CITES Appendix II in 2020 · 2020–2024
Data exist but the sample is biased
Density is collected locally by diver transects, almost always within one-off projects. Sea cucumber fishing is largely artisanal and substantially illegal — statistics are systematically understated, and density data are absent exactly where pressure is highest (the Indo-Pacific).
FAO, Sea Cucumbers: A Global Review of Fisheries and Trade; CITES Appendix II listings
3.2.3.2Sea urchin abundanceind/m2Not measured anywhere0.82%
No global data. Locally from 0 to more than 30 individuals/m² (sharply elevated in an urchin barren)
No observation network
No global network exists. Collected locally in two contexts: fisheries (Japan, Chile, Maine) and kelp forest degradation. Reef Life Survey records urchins at its sites but publishes no global aggregate, for the same reasons as for reef fish.
Reef Life Survey; local surveys; Filbee-Dexter & Scheibling 2014 (urchin barrens)
3.2.4.1Biomassg/m2Global value exists0.55%
A modelled global map of macrobenthic biomass exists: typically 5-50 g/m² on the shelf falling below 0.1 g/m² in the abyss; total benthic macrofaunal biomass is estimated on the order of 10⁸-10⁹ t wet mass · 2010
Technical barrier to measurement
Direct measurement requires a grab sampler and laboratory sorting — the entire history of oceanography has produced only thousands of points for the whole seafloor. A global network is economically impossible; everything available is extrapolation.
Wei et al. 2010 (PLOS ONE, 'Global patterns and predictions of seafloor biomass'); Rex et al. 2006
3.2.4.2Diversity index (Shannon)indexNot measured anywhere0.55%
No global aggregate. Locally, macrobenthic H' usually runs 1.5-4.0
Data exist but are not comparable
The Shannon index depends on sample area and on the taxonomic resolution of identification (to species or to family), so consolidating studies into one number produces an artefact rather than a signal. No international protocol exists; the EU uses it under the MSFD, but even there it is not comparable between countries.
local studies; EMODnet Biology (primary data, no aggregate)
3.2.4.3Sensitive taxa share (AMBI)indexLocal scale only0.55%
No global value. AMBI is applied regionally, primarily in the EU under the MSFD and the Water Framework Directive
Data exist but are not comparable
Barely applied outside Europe: AMBI requires a sensitivity list for every taxon, and such lists exist for the Northeast Atlantic and the Mediterranean. For the Indo-Pacific and the Arctic no validated lists exist — the barrier is the absence of a methodological base rather than of observations.
AZTI Marine Biotic Index (Borja et al.); EU MSFD reporting

3.3 · Plankton

3.3.1.1Chlorophyll-amg/m3Global value exists0.75%
Ocean mean about 0.2-0.3 mg/m³ (0.03-0.1 in oligotrophic gyres, above 1 nearshore and in upwelling zones). Trend: decline at low and middle latitudes, greening of high latitudes with a poleward shift · continuous, trend 1998–2024
NASA OceanColor (MODIS/VIIRS); ESA Ocean Colour CCI; Science Advances 2025 (chlorophyll decline at low-mid latitudes); Zhao et al. 2025 (Science, poleward greening)
3.3.1.2Primary productionmg C/m3/dayGlobal value exists0.75%
Global ocean primary production about 45-50 Pg C/year, roughly half the biosphere total. A statistically significant decline in net primary production now covers almost half the ocean area, strongest in the stratified tropics and subtropics · 2025 (1998-2023 record)
Nature Communications 2025, 'Global declines in net primary production in the ocean color era' (doi:10.1038/s41467-025-60906-y); Communications Earth & Environment 2025 (doi:10.1038/s43247-025-02051-4)
3.3.1.3Bloom frequencyevents/yearGlobal value exists0.75%
No global count of events per year. Established: there is no evidence of a global increase in harmful algal blooms once the growth in observational effort is accounted for; meanwhile phytoplankton blooms at low and middle latitudes are becoming less frequent · 2021 / 2025
Data exist but the sample is biased
The absolute number of events is not collected globally: recording a bloom depends on whether a monitoring programme exists and on whether aquaculture or beaches are affected. HAEDAT, the global event database, records only what is reported — it measures attention rather than the phenomenon.
Hallegraeff et al. 2021 (Nature Communications, 'Perceived global increase in algal blooms'); IOC-UNESCO HAEDAT; Science Advances 2025
3.3.2.1Copepod biomassmg/m3Local scale only1.23%
No global value. Long Continuous Plankton Recorder series exist for the North Atlantic, North Pacific and Southern Ocean; the trend is a poleward shift of communities and a shift towards smaller species
Data exist but the sample is biased
Global coverage is physically absent: the CPR is towed by merchant ships along fixed routes, so data exist only along those routes. The tropics, the Indian Ocean and most of the South Pacific are uncovered — there are neither participating ships nor national programmes.
Global Alliance of CPR Surveys (GACS); Marine Biological Association CPR Survey
3.3.2.2Larval abundance (ichthyoplankton)ind/m3Local scale only1.23%
No global aggregate. Regular ichthyoplankton surveys are run by about 20 countries (CalCOFI, since 1949, is the longest series)
Data exist but the sample is biased
Not consolidated globally: surveys follow national fishery interests and cover the shelves of the USA, EU, Japan and Australia. The open ocean and the shelves of developing countries are uncovered. No single protocol exists (net type, tow depth, time of day), so even the existing series are not comparable between countries.
CalCOFI (NOAA/Scripps); ICES Eggs and Larvae Database

Geology and Minerals

4.1 · Bottom substrates

4.1.1.1Sediment map (natural distribution)conv.unitsGlobal value exists0.56%
Global maps of seabed sediment type exist: a seafloor lithology map built from about 14,400 samples, supplemented by the dbSEABED compilation. Separately: 28.7% of the ocean floor (about 104 million km²) has been surveyed to modern standards, an increase of nearly 5 million km² in a year · 2015 / April 2026
Technical barrier to measurement
Sampling requires ship time: a single grab station takes hours of research vessel operation. The entire history of oceanography has produced some tens of thousands of samples. The barrier is economic and physical, not organisational.
Dutkiewicz et al. 2015 (Geology, 'Census of seafloor sediments'); dbSEABED; Seabed 2030 / GEBCO 2026 Grid (IHO press release, April 2026)
4.1.1.2Grain sizemmLocal scale only0.56%
No global mean. Grain-size compilations exist but coverage is extremely sparse. The general pattern: sand and gravel are concentrated on shelves (about 7-8% of ocean area), fine fractions dominate the deep sea
Technical barrier to measurement
The same technical barrier: physical sampling is required. Satellites cannot see grain size at all. Acoustic backscatter gives a hardness proxy but not grain size without calibration against samples.
dbSEABED (Jenkins); EMODnet Geology (Europe); Dutkiewicz et al. 2015
4.1.1.3Sorting (Trask coefficient)coeff.Not measured anywhere0.56%
No global data. The coefficient is computed from the grain-size curve of a specific sample
Data exist but are not comparable
Beyond sparse sampling, there are competing methods: the Trask coefficient, Folk and Ward statistics, and the Inman method give different numbers for the same sample. Consolidating published values without recomputing from the original curves produces an artefact, and the original curves are usually not published.
Trask 1932; Folk & Ward 1957; local studies
4.1.1.4Bottom hardnesskPaNot measured anywhere0.56%
No global data. Measured in situ by penetrometer or computed from acoustic backscatter
Technical barrier to measurement
A double barrier: measurement requires lowering an instrument to the seabed (technical), and survey results for pipelines, cables and platforms belong to the client (closed). Open data are practically nonexistent for any region.
engineering surveys (reports unpublished); methods ISO 19901-8
4.1.2.1Rock (share)%Local scale only0.45%
No global share. Hard rock substrates concentrate on seamounts, mid-ocean ridges and shelf areas with strong hydrodynamics; the share is mapped for Europe in EMODnet
Data exist but are not comparable
A classification barrier: substrate description schemes are mutually incompatible (Folk, Wentworth, EUNIS, national systems), so shares from different sources cannot be summed. Add low bathymetric coverage: without high-resolution survey, rock outcrop cannot be distinguished from boulder field.
EMODnet Geology (Seabed Substrate); Harris et al. 2014 (Geomorphology of the oceans)
4.1.2.2Boulders (share)%Local scale only0.45%
No global share. Boulder fields are characteristic of formerly glaciated areas (the Baltic, the North Sea, the Canadian shelf)
Data exist but are not comparable
The same classification problem plus scale: boulders are resolvable only at sub-metre survey resolution. Such data were mostly gathered for engineering projects and belong to the developers.
EMODnet Geology; national geological surveys (BGS, GEUS)
4.1.2.3Gravel (share)%Local scale only0.45%
No global share. Gravel fractions are localised on shelves, including as a target for construction aggregate extraction
Data exist but are not comparable
Beyond classification incomparability there is a conflict of interest: gravel deposit maps have commercial value as resource maps, so detailed surveys are commissioned by extraction companies and not published.
EMODnet Geology; The Crown Estate (UK), extraction licensing
4.1.2.4Sand (share)%Local scale only0.45%
No global share. Sand dominates most shelves; the deep seafloor is essentially free of sand except in turbidite fans
Data exist but are not comparable
Classification incomparability and sparse sampling. Note the link to 6.1.4.1 (sand extraction, about 6 billion t/year): extraction intensity is better known than the resource itself — we know more precisely how much we remove than how much there is.
dbSEABED; EMODnet Geology; UNEP Marine Sand Watch (for comparison with extraction)
4.1.2.5Silt and clay (share)%Global value exists0.45%
No precise global share, but the order of magnitude is established: fine-grained sediments (pelagic clays, calcareous and siliceous oozes) cover most of the deep seafloor — by various estimates on the order of 70-80% of ocean area · 2015
Data exist but are not comparable
The only one of the five sub-measures in this group for which a global order of magnitude can be established: fine fractions dominate so strongly that the conclusion is robust to classification differences.
Dutkiewicz et al. 2015 (Geology, 'Census of seafloor sediments in the world's ocean')

4.2 · Mineral resources

4.2.1.1Oil reserves (leaks)tLocal scale only0.55%
Offshore fields hold roughly 25-30% of the world's proven oil reserves. In absolute terms, against world proven reserves of about 1.7 trillion barrels, that is on the order of 400-500 billion barrels (about 55-70 billion t) — but this figure is unreliable · 2024–2025
Data exist but are not public
The data are closed and politically motivated. Reserves are declared by state companies with no audit. Leaks — what the indicator actually measures — are detected by the operator from pressure loss and reported to the national regulator; no consolidated international statistics on offshore leaks exist.
BP/Energy Institute Statistical Review; EIA International Energy Statistics; industry-literature critique of OPEC estimates
4.2.1.2Gas reserves (leaks)m3Local scale only0.55%
Offshore fields supply roughly a third of world gas production; world proven gas reserves are about 188 trillion m³, of which the offshore share is estimated at 50-70 trillion m³ · 2024–2025
Data exist but are not public
The same closure and absence of audit. For leaks the situation is changing: satellite methane monitoring (TROPOMI, MethaneSAT) has begun detecting large offshore releases independently of operators — a rare case of technology routing around an institutional gap.
Energy Institute Statistical Review; IEA Methane Tracker; TROPOMI / MethaneSAT
4.2.1.3Gas hydrates (deposit integrity)m3Local scale only0.55%
Methane in gas hydrates is estimated at roughly 3,000-20,000 trillion m³ (1,000-5,000 Gt of carbon) — nearly an order of magnitude of uncertainty. Deposit integrity is not monitored globally at all
Technical barrier to measurement
Deposit integrity requires repeat seismic surveys and monitoring of seabed gas seeps — only a handful of such observations exist worldwide (the Svalbard margin, the Black Sea, the Cascadia slope). The barrier is technical and financial: this is the most expensive class of marine observation.
Boswell & Collett 2011 (Energy & Environmental Science); USGS Gas Hydrates Project; Ruppel & Kessler 2017
4.2.2.1Polymetallic nodules (surface disturbance)t/km2Global value exists0.55%
Nodule abundance in the Clarion-Clipperton Zone is typically 10-20 kg/m², that is 10,000-20,000 t/km²; the total resource of the zone is estimated on the order of 21 billion t. On disturbance: tracks from test mining in the 1970s and 1980s remain visible and biologically unhealed 30-40 years later · 2015–2026
The absence of commercial mining means disturbance is currently near zero — this is a baseline rather than a gap. Data on it will appear only alongside mining, and the question is whether they will be open: the data submission requirement currently binds ISA contractors but not permits issued under US national law.
ISA Deep Seabed Minerals database; DISCOL repeat surveys (Jones et al. 2017); ISA resource assessments CCZ
4.2.2.2Cobalt crusts (thickness)cmLocal scale only0.55%
Cobalt-rich crusts are typically 1-25 cm thick (characteristic values a few centimetres), growing at millimetres per million years. International Seabed Authority exploration contracts for crusts cover seamounts of the western Pacific
Data exist but are not public
The main body of data sits in contractor reports to the ISA, part of it treated as confidential commercial information and unpublished. Technically, measurement requires sampling on seamounts at 800-2,500 m.
ISA exploration contracts (cobalt-rich ferromanganese crusts); Hein et al. 2013 (Ore Geology Reviews)
4.2.2.3Placer deposits (concentration)g/tNot measured anywhere0.55%
No global data. Coastal placers (titanium, zircon, tin, diamonds, gold) are worked in Namibia, Indonesia, India, Australia and South Africa
Data exist but are not public
Classic commercial confidentiality: grade in a placer is the key commercial parameter determining licence value. The data belong to mining companies and are disclosed only to the extent required by exchange rules (JORC, NI 43-101), that is for individual projects of listed companies.
public company reporting under JORC/NI 43-101; national geological surveys
4.2.3.1Salinity for desalinationPSUReference only, no weight
Data are fully available globally: mean ocean salinity about 34.7 PSU, while in the most evaporative regions the Persian Gulf runs 40-42 PSU and the Red Sea about 40 PSU. World desalination capacity is about 95-100 million m³/day, discharging over 140 million m³/day of brine · 2019–2026
Duplication: this indicator measures the same quantity as 2.1.2.1 (PSU), only framed as a resource. In a cumulative index it would double-count salinity.
Argo / World Ocean Atlas 2023; Jones et al. 2019 (Science of the Total Environment) — state of desalination and brine production; IDRA / EU Blue Economy Observatory (2026)
4.2.3.2Salt content (NaCl)kg/m3Reference only, no weight
Total dissolved salts in seawater about 35 kg/m³, of which sodium chloride is about 27-30 kg/m³ (the ratio of major ions is constant throughout the ocean — the principle of constant proportions) · continuous
Duplication squared: the quantity is an arithmetic derivative of 4.2.3.1, which itself duplicates 2.1.2.1. Three indicators in the set describe one measurement.
Dittmar's principle (constant proportions); Argo / World Ocean Atlas 2023
4.2.3.3Temperature for OTEC°CReference only, no weight
Data are available globally. OTEC requires a difference of at least 20 °C between the surface and about 1,000 m depth, which holds in a band roughly 20° either side of the equator, where the surface is 25-30 °C and 1,000 m is 4-5 °C · continuous
Duplication: a combination of 2.1.1.1 (SST) and 2.1.1.2 (deep temperature) framed as a resource. As a standalone indicator of ocean condition it carries no meaning — it describes technical suitability, not the health of the environment.
Argo / World Ocean Atlas 2023; NREL / IRENA reviews OTEC

Pollution

5.1 · Chemical pollution

5.1.1.1Hg in waterng/LMeasured0.41%
Open ocean: about 0.0002-0.0006 µg/L (1-3 pmol/kg). It is established that mercury in the upper ocean has roughly tripled relative to pre-industrial levels · 2014–2024
Data exist but are not comparable
The unit in the set does not match the real range: concentrations are 1,000-5,000 times below 1 µg/L, so entries will read as zero. Switching to ng/L or pmol/kg is required, otherwise the indicator is technically unusable.
GEOTRACES Intermediate Data Product; Lamborg et al. 2014 (Nature); Minamata Convention effectiveness evaluation
5.1.1.2Cd in waterng/LMeasured0.41%
Surface: about 0.0001-0.01 µg/L; at depth about 0.07-0.11 µg/L (a nutrient-type distribution) · 2021–2024
Data exist but are not comparable
The same unit-scale problem. In addition, the increase with depth is natural: without recording sampling depth the value cannot be interpreted.
GEOTRACES Intermediate Data Product 2021; World Ocean Atlas (for comparison with nutrients)
5.1.1.3Pb in waterng/LMeasured0.41%
North Atlantic surface: about 0.002-0.004 µg/L — a decline of roughly 70-80% since the 1980s following the phase-out of leaded petrol · 2020-
GEOTRACES; Boyle et al. 2014 (Geochimica et Cosmochimica Acta); Bridgestock et al. 2016
5.1.1.4Cu in waterng/LMeasured0.41%
Surface: about 0.03-0.1 µg/L; at depth about 0.13-0.25 µg/L · 2021
GEOTRACES Intermediate Data Product 2021
5.1.1.5Sediment accumulationmg/kgReference only, no weight
No global aggregate. National and regional sediment quality guidelines (ERL/ERM, OSPAR, NOAA SQuiRT) set thresholds but do not describe the actual state of the world
Data exist but are not comparable
Methodological incomparability plus absence of aggregation: concentration depends on sediment grain size (more metal in clays than in sands), so without normalisation to aluminium or organic content values from different sites cannot be compared. No international normalisation protocol and no global marine sediment database exist.
OSPAR CEMP; NOAA SQuiRT tables; national programmes
5.1.1.6Tissue accumulation in aquatic organismsµg/g wet weightGlobal value exists0.41%
Median total mercury in marine fish about 0.1-0.3 µg/g wet mass (markedly higher in large predators — tuna, swordfish, sharks). A global biotic mercury database exists and is used for Minamata Convention effectiveness evaluation · 2021–2025
Data exist but are not comparable
The unit in the set is dry mass, while food safety monitoring works in wet mass (conversion factor 4-5, varying by species). This is methodological incomparability in its plainest form: without recording the conversion basis, data from different sources will differ severalfold.
Global Biotic Mercury Synthesis (Evers et al.); Minamata Convention Effectiveness Evaluation; FAO/WHO GEMS/Food
5.1.2.1POPs (PCB, DDT) in waterng/LMeasured0.82%
Open ocean: PCBs about 0.001-0.1 ng/L, DDT lower. Trend: declining in the northern hemisphere after the Stockholm Convention (2004), though redistribution towards the Arctic continues
No observation network
The Stockholm Convention global monitoring plan is built on air and human milk rather than seawater — the network exists but measures other media. Seawater is excluded from the monitoring plan precisely because of analytical cost at picogram-per-litre concentrations.
Stockholm Convention Global Monitoring Plan; AMAP (Arctic); individual expedition studies
5.1.2.2Petroleum hydrocarbonsmg/LGlobal value exists0.82%
No global concentration figure. Indirectly: the number of large tanker spills fell from about 79 per year in the 1970s to fewer than 10 per year in the 2010s-2020s, while chronic input from shipping, runoff and natural seeps is not tracked · 2024–2025
No observation network
Water concentration is not measured globally: there is no network, and satellites see only surface slicks from large spills. Accounting is by event (number and volume of spills) because that is what interests insurers and the IMO, not by the state of the environment.
ITOPF Oil Tanker Spill Statistics 2025; NASEM, Oil in the Sea IV (2022)
5.1.2.3Pesticidesµg/LLocal scale only0.82%
No global data for seawater. Coastal measurements in individual countries record exceedances for neonicotinoids and herbicides near agricultural runoff
No observation network
No observation network exists in the marine environment: pesticides are monitored in rivers and drinking water where regulation requires it. Nobody is obliged to monitor seawater, so data appear only within research projects.
GBRMPA Marine Monitoring Program (Australia); HELCOM; local studies
5.1.3.1Antibioticsng/LLocal scale only0.55%
No global data for seawater. For rivers a global study exists (1,052 sites in 104 countries) in which a quarter of sites exceeded safe thresholds for pharmaceuticals · 2022
No observation network
The marine part is simply unmeasured: dilution in seawater requires detection limits of a few ng/L, which is costlier than river analytics. There is also no regulatory driver — no convention requires monitoring antibiotics in seawater.
Wilkinson et al. 2022 (PNAS, 'Pharmaceutical pollution of the world's rivers'); local marine studies
5.1.3.2Hormonesng/LNot measured anywhere0.55%
No global data. Coastal waters and estuaries show oestrogens at a few ng/L, enough to feminise fish
No observation network
Not collected globally or systematically locally. Hormones require either expensive mass spectrometry or bioassays (vitellogenin in fish), which are not standardised. No marine convention requires it.
individual estuary studies (Thames, Chesapeake Bay); UNEP reviews
5.1.3.3Caffeine (discharge marker)ng/LLocal scale only0.55%
No global data. Caffeine is widely used as a tracer of untreated domestic sewage: tens to hundreds of ng/L in polluted coastal water, single digits in clean water
No observation network
Collected locally and mostly in research. No global network exists, but the barrier here is low: this is one of the few measures in the category that could realistically be filled by project measurements under a common protocol.
local studies (reviews Marine Pollution Bulletin); method after Buerge et al. 2003
5.1.4.1BOD5mg O2/LLocal scale only0.75%
No global data for seawater. BOD5 is a wastewater control measure and is barely used in marine monitoring
Data exist but are not comparable
The measure is borrowed from wastewater control practice: in open water BOD5 sits near the detection limit and is uninformative. Data are missing not because nobody collects them but because the measure is not applied to seawater — a question about the composition of the indicator set rather than an observation gap.
national wastewater control programmes; HELCOM (estuaries)
5.1.4.2CODmg O2/LLocal scale only0.75%
No global data for seawater. Used in China as the principal coastal control standard, not in most other countries
Data exist but are not comparable
The same situation: an industrial control measure rather than a marine one. Complicated further by standard COD methods performing poorly in saline water (chloride interferes), so national methods differ and are not comparable.
China Marine Environment Bulletin; national standards
5.1.4.3Trophic index (TRIX)indexLocal scale only0.75%
No global value. TRIX is applied in the Mediterranean, the Adriatic and the Black Sea under regional conventions
Data exist but are not comparable
The barrier is methodological rather than observational: TRIX was calibrated for the Adriatic (Vollenweider 1998) and its scale is not validated for oceanic conditions. The components exist globally, so the indicator could technically be computed worldwide, but the result would not be interpretable on the original scale.
Vollenweider et al. 1998; UNEP/MAP MED POL; Black Sea Commission

5.2 · Solid waste

5.2.1.1Floatingg/km2Reference only, no weight
Floating plastic: about 171 trillion particles weighing about 2.3 million t on the ocean surface (2023 estimate). Annual input is estimated at wildly divergent figures: 1-2, 8, 11 and 12.7 million t/year in different studies · 2023 / 2014–2026
Data exist but are not comparable
The unit g/m³ does not match practice: floating debris is sampled by Manta trawl and expressed in g/km² of surface or particles/km², not as a volumetric concentration. Conversion requires an assumption about tow depth — in the current unit the indicator cannot be filled correctly.
Eriksen et al. 2023 (PLOS ONE); Jambeck et al. 2015 (Science); Meijer et al. 2021 (Science Advances); OECD Global Plastics Outlook 2022; UNEP
5.2.1.2Seaflooritems/m2Local scale only0.56%
No global aggregate. ROV and trawl surveys give from 0 to several items/m²; concentrations peak in submarine canyons and deep basins where currents carry debris
Technical barrier to measurement
Technical barrier: an ROV or submersible is required, at tens of thousands of dollars per ship-day. Existing data were gathered incidentally on geological and biological cruises, so the sample is determined entirely by where researchers happened to work.
Deep-sea Debris Database (JAMSTEC); Pham et al. 2014 (PLOS ONE); Kane et al. 2020 (Science, canyons)
5.2.1.3On beachesitems/mLocal scale only0.56%
No comparable global value. Regional programmes with a single protocol (OSPAR) give on the order of hundreds of items per 100 m of beach. Global clean-up campaigns collect data in many countries, but the sample is set by where volunteers are
Data exist but the sample is biased
A classic case of biased sampling: citizen science data are gathered where organisations and volunteers exist — in wealthy countries and on tourist beaches. This creates the illusion that those are the polluted places. The gap is structural rather than observational.
OSPAR Beach Litter Monitoring; UNEP/IOC Guidelines on Survey and Monitoring of Marine Litter; Ocean Conservancy ICC (with a caveat on sampling bias)
5.2.1.4Large accumulationshaGlobal value exists0.56%
The Great Pacific Garbage Patch: about 160 million ha (1.6 million km²) with a mass of about 79,000 t. Five subtropical accumulation zones are recognised in total · 2018
Lebreton et al. 2018 (Scientific Reports, 'Evidence that the Great Pacific Garbage Patch is rapidly accumulating plastic')
5.2.2.1In water columnparticles/m3Global value exists0.72%
Open ocean: usually 0.1-10 particles/m³; in the upper 200 m of the Atlantic values up to about 2,200 particles/m³ have been recorded (a synthesis of 1,257 samples from 1-60 m) · 2024–2026
Data exist but are not comparable
Methodological incomparability in pure form: no single international sampling protocol for microplastics exists, and ISO/ASTM standards are still in development. Data gathered before their adoption will remain unconsolidatable.
Nature (Atlantic sample synthesis, 2024); Lindeque et al. 2020 (effect of mesh size); GESAMP Reports and Studies No. 99
5.2.2.2In sedimentsparticles/kg dry weightLocal scale only0.72%
No global aggregate. Range from single digits to several thousand particles per kg of dry sediment; in deep-sea hotspots on canyon floors up to 1.9 million fibres/m² · 2020–2025
Data exist but are not comparable
Methodological incomparability and absence of aggregation. The measure is in principle collectable under a single protocol and is one of the most promising for filling: sediment, unlike water, integrates pollution over years and does not require repeat visits.
Kane et al. 2020 (Science); Van Cauwenberghe et al. 2013; GESAMP No. 99
5.2.2.3In tissues of aquatic organismsparticles/individualLocal scale only0.72%
No global aggregate. Microplastic has been found in every taxonomic group examined, from zooplankton to whales; typical values range from under 1 to tens of particles per individual
Data exist but are not comparable
Collected widely but never aggregated: thousands of publications with different species, organs and methods. The presence of microplastic is convincingly established; the quantity is not. For the platform it is more honest to show presence than a number.
reviews in Environmental Pollution; GESAMP No. 99; individual taxon studies
5.2.2.4Polymer types (detection)qualitativeGlobal value exists0.72%
Consistent worldwide: polyethylene and polypropylene dominate because they float; then polystyrene, PET and PVC, which sink and are therefore undercounted in surface samples
GESAMP Reports and Studies No. 99; Erni-Cassola et al. 2019 (polymer distribution review)
5.2.3.1Number of lost fishing gear itemsunitsGlobal value exists0.82%
About 2% of all fishing gear worldwide is lost each year, including more than 25 million pots and traps and about 739,600 km of longline mainline · 2022
Note: the widely cited figure of 640,000 t/year is now considered unsubstantiated by researchers — it circulated through NGO reports without a primary source. This is a textbook citation cascade: the measure looked global and dependable purely through repetition. For the platform it is worth recording not only the source but whether the source is primary.
Richardson et al. 2022 (Science Advances, doi:10.1126/sciadv.abq0135); critique of the 640,000 t figure — The Tyee/academic reviews 2024
5.2.3.2Area covered by netsm2Global value exists0.82%
Annual losses of net gear: about 75,049 km² of purse seine, 2,963 km² of gillnet, 218 km² of trawl — roughly 78,000 km² (7.8 billion m²) per year · 2022
Richardson et al. 2022 (Science Advances)
5.2.4.1Number of sunken vesselsunitsGlobal value exists0.34%
The total number of shipwrecks worldwide is estimated at around 3 million across all eras. Of these about 8,500 are classified as potentially polluting (mostly Second World War vessels), holding an estimated 6-25 billion gallons of oil · 2024–2025
No observation network
No systematic accounting exists. Data come from wartime loss archives, insurance registers and hydrographic finds; the coordinates of many wrecks are unknown. Surveying each object requires an ROV — the barrier is both organisational and technical.
UNESCO Underwater Cultural Heritage; Project Tangaroa / Potentially Polluting Wrecks database; Michel et al. (NOAA RULET)
5.2.4.2Cargo type (hazardous cargo)qualitativeLocal scale only0.34%
No global register of shipwreck cargoes exists. For some objects the crew lists and cargo manifests are known from archives; for most they are not
Data exist but are not public
Data exist but are scattered across national military and insurance archives, part of it classified (munitions, chemical weapons). This is the one measure in the set where the gap is partly explained by access restrictions rather than absence of observation.
Project Tangaroa; national archives; OSPAR (dumped munitions)
5.2.4.3Hull corrosion%Not measured anywhere0.34%
No global data
Technical barrier to measurement
Not collected: it requires direct hull survey by submersible with residual steel thickness measurement. Worldwide only a handful of wrecks have been surveyed this way — those of historical value or presenting an obvious threat. For the remaining thousands there will be no data.
NOAA/National Park Service (USS Arizona corrosion studies); Major Projects Foundation (Pacific)

5.3 · Physical impact

5.3.1.1Ship noisedBGlobal value exists0.55%
Source level of a container ship about 170-190 dB re 1 µPa. Globally, shipping noise is rising: historically about +3 dB per decade, and by model estimates the acoustic energy doubles on average every 11.5 years, fastest in the Arctic and the Norwegian Sea · 2022–2026
Data exist but are not public
Hydrophone networks with open data barely exist: most long-term recordings belong to navies and are declassified selectively. That is why the global picture of ocean noise is a reconstruction from traffic rather than an observation.
Jalkanen et al. 2022 (Environmental Pollution, 'Underwater noise emissions from ships during 2014–2020'); McDonald et al. 2006 (JASA); Frontiers in Marine Science 2024
5.3.1.2Seismic survey pulsesdBLocal scale only0.55%
Airgun source levels reach about 250-260 dB re 1 µPa, among the loudest anthropogenic sounds in the ocean. No global accounting of the number or geography of surveys exists
Data exist but are not public
Seismic survey data belong to oil and gas companies and contractors and are commercially confidential: survey locations and timing are disclosed only to national regulators. No global register exists and none will without a regulatory requirement.
IOGP guidelines; Nowacek et al. 2015 (Frontiers in Ecology and the Environment)
5.3.1.3Pile driving noisedBLocal scale only0.55%
Pile driving produces about 200-250 dB re 1 µPa at source. Data on actual operations exist regionally, primarily for North Sea offshore wind where noise limits apply
No observation network
Data appear only where a regulation requires measuring them. The EU has such requirements, most countries do not — so the gap is defined by the geography of regulation rather than by technical feasibility.
BSH (Germany), pile-driving noise limits; OSPAR Impulsive Noise Register
5.3.2.1Coastal zone illuminationluxGlobal value exists0.61%
First global assessment: about 1.9 million km² of the world's coastal waters are exposed to biologically meaningful artificial light at 1 m depth · 2021
Smyth et al. 2021 (Elementa, 'A global atlas of artificial light at night under the sea')
5.3.2.2Light penetration into waterlux at depthReference only, no weight
The same atlas gives estimates at 1, 10 and 20 m depth: about 1.6 million km² affected at 10 m, about 840,000 km² at 20 m · 2021
Smyth et al. 2021 (Elementa)
5.3.3.1Temperature gradient from discharge°C per 100mNot measured anywhere0.41%
No global data. Discharge standards exist nationally (typically limiting the rise to 1-3 °C at the edge of the mixing zone)
Data exist but are not public
Neither collected nor aggregated: measurements are made by the operator for the regulator and stay in company reporting. No international organisation maintains a register of thermal discharges to the sea, even though the facilities — some 200 coastal nuclear plants, the large desalination complexes of the Persian Gulf — are well known and in principle enumerable.
national discharge permits; desalination reviews (Persian Gulf, Middle East Desalination Research Center)
5.3.3.2Plume areakm2Not measured anywhere0.41%
No global data. Individual satellite studies of thermal plumes from large facilities give areas from a few to tens of km²
No observation network
No network and no client: the satellite data for this are available and free, so the barrier is purely organisational. One of the most realistic candidates for filling — the method is reproducible and the facilities are known.
Individual Landsat TIRS/ASTER studies; local monitoring programmes
5.3.4.1Suspended matter from dredgingmg/LLocal scale only0.61%
No global data. During dredging, plume suspended solids typically run 10-100 mg/L, reaching several hundred near the draghead
Data exist but are not public
Data are collected in large volumes but belong to contractors and port authorities. They are rarely published and only in aggregate. No international register of dredging operations exists, despite the enormous volumes worldwide.
CEDA/IADC guidance; project-specific reporting (Rotterdam Maasvlakte, Panama, Persian Gulf)
5.3.4.2Settling durationdaysNot measured anywhere0.61%
No global data. Settling after operations cease takes from hours (sands) to weeks (silts and clays)
No observation network
Practically never measured: it requires continuing monitoring after operations finish, when the contractor no longer has obligations. This gap is created by the structure of regulation — control is tied to the period of works rather than the period of impact.
dredging project environmental impact assessments; CEDA position papers

Pressures and Use

6.1 · Exploitation

6.1.1.1Fishing mortality / Fmsyratio (F/Fmsy)Global value exists0.61%
World marine capture about 80 million t/year (92 million t total capture of aquatic animals in 2024 including inland waters). Independent reconstructions give a figure 30-50% above official reporting · 2024 (publ. 2026)
FAO SOFIA 2026; FAO FishStat; Pauly & Zeller 2016 (Nature Communications) — catch reconstruction
6.1.1.2Fishing intensity (days at sea/km2)kWh/km2Reference only, no weight
No global value in this unit. It is established that industrial fishing is detectable across more than 55% of the ocean surface — four times the area occupied by agriculture on land · 2018
Data exist but the sample is biased
The unit days/km² is not computed globally: effort is aggregated in kW-hours or vessel-hours per cell. The key limitation is that AIS is mandatory only for vessels above 300 GT on international voyages, so artisanal fishing is structurally invisible. This is sampling bias rather than an observation gap.
Kroodsma et al. 2018 (Science, 'Tracking the global footprint of fisheries'); Global Fishing Watch; critique — Amoroso et al. 2018
6.1.1.3Bycatch%Measured0.61%
No single value — the result depends entirely on the definition. Under a broad definition (all unmanaged and unused catch) the figure is around 40% of world catch; under a narrow one (only discarded catch) about 11%. Separately, at least 20 million individuals of protected and endangered species are estimated to be discarded annually · 2009 / 2019
Data exist but are not comparable
A methodological rather than observational problem: the term bycatch has no agreed international definition. Onboard observer programmes cover fractions of a percent of trips in most of the world's fleets — there is no complete accounting anywhere except a few EU, US and New Zealand fisheries.
Davies et al. 2009 (Marine Policy) — broad definition; Gilman et al. 2020 (Scientific Reports); FAO Technical Paper 633
6.1.1.4Discards%Measured0.61%
About 10.8% of world catch (9.1 million t/year against an average catch of 84.6 million t), confidence interval 7-16 million t. This is roughly half the rate of the late 1980s · 2010–2014 (publ. 2019)
Gilman et al. 2020 (Scientific Reports, 'Benchmarking global fisheries discards'); FAO Fisheries and Aquaculture Technical Paper 633 (Perez Roda et al.)
6.1.2.1Traffic densityvessels/dayGlobal value exists0.46%
No scalar global value. The world merchant fleet exceeds 100,000 vessels above 100 GT; shipping intensity has roughly quadrupled since 1992. Global AIS traffic density maps are published openly · 2020–2026
A global average would be meaningless
The measure does not reduce to a scalar: traffic density is extremely uneven (the Strait of Malacca against the South Pacific), so an ocean-wide average is meaningless. For an index, only density within the project cell is usable.
EMODnet Human Activities (Vessel Density); World Bank/IMF AIS-based trade tracking; Tournadre 2014 (traffic growth)
6.1.2.2Collision riskprobabilityGlobal value exists0.46%
No single global probability. Global maps of ship strike risk have been published for large whales (blue, fin, humpback, sperm); it is established that risk reduction measures apply to only a small share of the highest-risk areas · 2024
No observation network
There are almost no observations of the strikes themselves — only a model overlaying ranges on traffic. A vessel often does not notice the impact, and mandatory reporting does not exist in most countries. This gap arises from the nature of the event rather than from a lack of technology.
Nisi et al. 2024 (Science) — global risk maps; IWC Ship Strike Database
6.1.2.3Routes (presence in MPAs)qualitativeGlobal value exists0.46%
The data needed to compute this exist globally: overlaying AIS traffic on the world database of protected areas (WDPA) is technically routine. It is established that fishing, including bottom trawling, takes place in most of Europe's marine protected areas · 2018–2024
Data exist but are not comparable
The calculation is possible but no global published measure exists: it requires matching against each MPA's regime, and regimes are described inconsistently in the WDPA and often not machine-readable. The barrier is methodological.
Dureuil et al. 2018 (Science) — trawling in European MPAs; Global Fishing Watch × Protected Planet
6.1.2.4Vessel speedknotsGlobal value exists0.46%
Every vessel's speed is known from AIS. The practically significant relationship: reducing speed by 1 knot lowers a vessel's noise by roughly 1 dB, and slowing the world fleet by 10% would cut total acoustic energy by about 40% · 2020–2026
AIS (EMODnet, Global Fishing Watch); Leaper 2019 (speed-reduction analysis)
6.1.3.1Diving (person-days/year)person-days/yearLocal scale only0.34%
No global figure in person-days. The world's coral reefs are estimated to receive on the order of 70 million tourist visits a year worth about 36 billion dollars; dives are part of that flow · 2017
Data exist but are not public
Nobody keeps the record: dive logs belong to dive centres, and certifying bodies (PADI, SSI) publish certifications issued rather than activity. The measure is in principle collectable through partnership with dive industry associations — a realistic candidate for filling.
Spalding et al. 2017 (Marine Policy, 'Mapping the global value and distribution of coral reef tourism')
6.1.3.2Yachting (moorings/year)moorings/yearNot measured anywhere0.34%
No global data
Data exist but are not public
Not collected: mooring data are held by marinas and port authorities as commercial information. Wild anchoring — the main cause of damage to seagrass and coral — is not recorded anywhere at all, even though it is what does the harm.
marina records; individual studies anchor damage (Mediterranean, Balearics)
6.1.3.3Coastal recreation (visitors/year)visitors/day/kmLocal scale only0.34%
No global figure in this unit. Coastal and marine tourism is the largest sector of the ocean economy by value; country-level arrival statistics exist, but their distribution along the coastline does not
Data exist but are not public
The unit visitors per day per km requires a spatial distribution that tourism statistics do not contain. It can be reconstructed indirectly from mobile operator and social media geotags — the method exists, but the data are commercial.
UN Tourism (UNWTO) country-level statistics; OECD Ocean Economy; studies geotag-based
6.1.4.1Sand extraction volumetReference only, no weight
About 6 billion t of sand and gravel are extracted from the marine environment each year — comparable to the natural riverine supply to the ocean · 2023
UNEP/GRID-Geneva, Marine Sand Watch (2023)
6.1.4.2Oil extractiont/dayGlobal value exists0.55%
Offshore production supplies roughly 27-30% of world oil, that is on the order of 3.5-4 million t/day (about 27-30 million barrels/day) · 2024–2025
IEA Oil Market Report; EIA International Energy Statistics; industry reviews Rystad/Wood Mackenzie
6.1.4.3Deep-sea mining (presence)qualitativeGlobal value exists0.55%
No commercial mining exists anywhere yet. The International Seabed Authority has issued 31 exploration contracts to 22 contractors (19 for polymetallic nodules, 18 of those in the Clarion-Clipperton Zone); exploitation regulations remain unfinished as of June 2026. Separately, the US subsidiary of a Canadian company has applied to NOAA for commercial mining over about 65,000 km² with a resource of about 619 million t of wet nodules, with a permit expected in the first quarter of 2027 · 2026
ISA Exploration Contracts registry; NOAA Deep Seabed Mining final rule (21.01.2026); TMC USA consolidated application (22.01.2026); The Conversation/Phys.org, June 2026
6.1.5.1Farm areahaReference only, no weight
No global figure for the area of marine farms. Production is known: aquaculture of aquatic animals passed 100 million t for the first time in 2024 (103 million t), or 141 million t including algae · 2024 (publ. 2026)
No observation network
Area is in principle visible from space — cage and algae farms are identifiable in Sentinel imagery, and mapping work exists for China and Korea. No global product exists, but this is one of the most realistic gaps to close: the method exists and the data are free.
FAO SOFIA 2026 (production); satellite mapping studies of mariculture (China, South Korea)
6.1.5.2Feed volumetReference only, no weight
No global feed volume figure. Indirectly: 11% of world aquatic animal production goes to non-food uses, overwhelmingly fishmeal and fish oil, and 90% of fishmeal is consumed by aquaculture · 2024 (publ. 2026)
Data exist but are not public
Feed data are commercial information held by producers (Skretting, BioMar, Cargill) and disclosed only in aggregate in subscription industry reports. This is not an absence of observation but closure.
FAO SOFIA 2026; IFFO (The Marine Ingredients Organisation) — aggregated statistics
6.1.5.3Antibiotic usekgLocal scale only0.36%
No global data. The contrast between countries is enormous: Norwegian salmon farming has reduced antibiotic use to almost nil through vaccination, while use remains high in parts of Asia and Latin America
Data exist but are not public
The data are closed and politically sensitive: disclosure affects export access and the industry's reputation. Norway publishes because its figures are good. Monitoring exists (veterinary prescriptions are recorded) but results are not published — the barrier is entirely institutional.
Norwegian Veterinary Institute (Fish Health Report); WOAH (OIE) antimicrobial use reporting; reviews Schar et al.
6.1.5.4Escapes% of production volumeNot measured anywhere0.36%
No global data. Norwegian salmon farming maintains mandatory reporting of escapes (on the order of hundreds of thousands of fish a year); in most countries no reporting exists
Data exist but are not public
Barely collected anywhere: reporting an escape brings inspections and fines, so whoever reports has a direct incentive to undercount. The indirect but objective method — the share of hybrids in the wild population from genetics — is applied only in Norway and Scotland.
Norwegian Directorate of Fisheries (escape statistics); Glover et al. (genetic introgression)

6.2 · Infrastructure

6.2.1.1Wind farms (impact on marine mammals)MW/turbinesGlobal value exists0.55%
Installed offshore wind capacity worldwide about 90-92 GW (2026), with roughly 70% concentrated in China and the United Kingdom. Projected to reach about 220-230 GW by 2030 · 2026
No observation network
Impact as such has not been assessed globally: studies exist for individual North Sea farms (harbour porpoise displacement during pile driving, with subsequent return), but no single method and no global synthesis exist.
RenewableUK EnergyPulse, Global offshore wind pipeline (June 2026); Westwood Offshore Wind in Numbers (2026); GWEC
6.2.1.2Underwater cableskmLocal scale only0.55%
Export and array cables of operating offshore wind farms exceed 28,000 km; interconnectors between countries exist separately · 2026
Data exist but are not public
Power cable routes are disclosed to hydrographic offices for navigational safety, but no consolidated open dataset exists — partly for critical infrastructure protection reasons. The gap sits close to the closed-data category.
Industry reviews submarine cable market (2026); national hydrographic offices; EMODnet Human Activities (partial)
6.2.1.3Drilling platformsunitsLocal scale only0.55%
No precise global register. Estimates of the number of offshore oil and gas installations range from about 7,000 to 12,000 depending on what counts as an installation (fixed platforms, subsea production systems, floating systems)
Data exist but are not comparable
National regulators maintain complete registers (BSEE in the USA, NSTA in the UK, NPD in Norway), but no global consolidation exists — nobody holds a mandate to maintain it. The objects are meanwhile clearly visible in satellite imagery and through gas flaring (the VIIRS Nightfire product). A realistic gap to close.
BSEE / NSTA / Norwegian Offshore Directorate (national registers); VIIRS Nightfire (flare detection)
6.2.2.1Underwater communication cableskmGlobal value exists0.72%
More than 1.5 million km of communication cable in service; 694 systems (603 active, 91 planned or under construction) and 1,893 landing stations. The longest system, 45,000 km around Africa, entered service in November 2025 · early 2026
TeleGeography, Submarine Cable Map 2026 and Submarine Cable FAQs; ICPC
6.2.2.2Pipelines (leaks)km, diameterLocal scale only0.72%
No consolidated global register of marine pipelines exists; total length is estimated in the tens of thousands of kilometres. Leak accounting is national and inconsistent
Data exist but are not public
Routes are known to operators and hydrographic offices, but no consolidated open dataset exists on infrastructure protection grounds. Leaks are detected either by pressure loss (that is, by the operator) or by chance — there is no independent monitoring. Satellite detection of methane plumes (TROPOMI, MethaneSAT) is beginning to close part of this gap.
EMODnet Human Activities (Europe, partial); national regulators; satellite methane monitoring