Physical Condition Measurement

Framework position: TG-4.8 sits in the data methods layer. It provides the measurement protocols and variable selection guidance for physical state variables that feed upstream into the ecosystem condition indicators described in TG-2.1 and TG-2.8, and ultimately into asset accounts compiled under TG-3.1. Users compiling condition accounts should read this Circular alongside TG-4.9 Biological Condition Measurement, which covers the biotic counterpart to the abiotic variables described here.

1. Outcome

This Circular provides guidance on measuring the physical condition of ocean and coastal ecosystems for use in SEEA EA ecosystem condition accounts. On completing this Circular, users will be able to select appropriate physical state variables, establish reference conditions, rescale variables to dimensionless condition indicators on a 0--1 scale, and integrate physical condition data into the asset account structure defined in TG-3.1 Asset Accounts.

Physical condition measurement addresses the abiotic subset of ecosystem condition—temperature, salinity, dissolved oxygen, pH, turbidity, nutrients, and light attenuation—corresponding to ECT Classes A and B as defined in SEEA EA Table 5.1. This Circular does not cover biological condition variables (covered by TG-4.9) or the compilation of condition account tables (covered by TG-3.1).

2. Requirements

Users should be familiar with the following circulars before working through this guidance:

• TG-0.1 General Introduction to Ocean Accounts—for the overall structure and purpose of ocean accounts
• TG-0.6 Glossary of Key Terms—for canonical definitions of "condition variable", "condition indicator", and "reference condition"
• TG-0.7 Quality Assurance Principles—for data quality, metadata, and uncertainty documentation requirements
• TG-4.1 Remote Sensing and Geospatial Data—for satellite sensor selection, pre-processing, and spatial data handling
• TG-4.2 Survey Methods for Ocean Economic Activity—for in-situ sampling design, field protocols, and quality control
• TG-4.4 Citizen Science and Community-Based Monitoring—for citizen science and community-based monitoring data in data-poor contexts

3. Guidance Material

3.1 Conceptual Framework

Physical condition measurement is anchored in SEEA EA Chapter 5, which establishes the Ecosystem Condition Typology (ECT), the concept of reference condition, and the rescaling of state variables to dimensionless condition indicators on a 0--1 scale. Physical condition captures the abiotic state of the ocean environment, corresponding to two ECT classes defined in SEEA EA Table 5.1:

• ECT Class A—Physical and chemical characteristics: variables describing the thermal, chemical, and optical properties of the water column (temperature, salinity, pH, dissolved oxygen, turbidity).
• ECT Class B—Hydrological and hydrochemical characteristics: variables describing water circulation, stratification, nutrient dynamics, and light penetration relevant to ecosystem functioning.

Physical condition is distinct from biological condition (TG-4.9 Biological Condition Measurement): physical variables measure the abiotic habitat envelope within which organisms live, while biological variables capture the biotic response of living organisms to that environment. Both are components of ecosystem condition as defined by SEEA EA Chapter 5 and both feed into asset accounts (TG-3.1 Asset Accounts).

Reference condition is the baseline against which current state is compared. SEEA EA §§5.36--5.43 establishes three acceptable approaches, summarised in Table 3.1.1 below.

The choice of reference condition must be documented and consistently applied across all accounting periods. Changes to the reference baseline require retroactive recalculation of all condition indicators to maintain time-series comparability.

Ecosystem-specific physical priorities vary by habitat type:

For detailed habitat-specific guidance, see TG-6.1 Coral Reef Ecosystem Accounting, TG-6.2 Mangrove and Coastal Wetland Accounting, TG-6.3 Seagrass Ecosystem Accounting, and TG-6.5 Pelagic and Open Ocean Accounting.

3.2 Indicator Selection

Indicators are selected according to four criteria set out in SEEA EA §§5.22--5.27. Table 3.2.1 below summarises these criteria.

Recommended physical condition indicators by ECT class

ECT Class A—Physical and chemical characteristics

• Sea Surface Temperature (SST): primary indicator of thermal stress and climate forcing. Unit: °C. Reference condition: long-term climatological mean (e.g., 1985--2014 NOAA OISST baseline). Where the full 1985--2014 period is unavailable nationally, use the longest consistent national record as the reference baseline, document the period explicitly in metadata, and note the departure from the global standard baseline. Rescaling: temperature anomaly relative to baseline; condition indicator decreases as anomaly increases above the reference variability range. For coral reef accounting units: threshold-based rescaling using the Maximum Monthly Mean (MMM) climatology is the preferred approach. When sea surface temperature exceeds the MMM, Degree Heating Weeks (DHW) begin to accumulate; DHW ≥ 8°C-weeks corresponds to significant bleaching risk. See NOAA Coral Reef Watch products for operational DHW data and MMM climatologies.^[1]
• Salinity: critical indicator for estuarine and coastal boundary zones. Unit: practical salinity units (PSU). Reference condition: long-term climatological mean from WOA or ARGO programme for the accounting unit.
• Ocean pH / aragonite saturation (Ω): indicator of ocean acidification state. Unit: pH units (dimensionless Ω). Reference condition: pre-industrial pH (~8.18) or a regional multi-decadal baseline from SOCAT/GLODAP. Condition indicator decreases as pH declines or Ω falls below the aragonite saturation threshold (Ω = 1), below which shell-forming organisms cannot maintain calcified structures.
• Dissolved oxygen (DO): indicator of habitat viability and eutrophication stress. Unit: mg/L or % saturation. Hypoxia threshold: <2 mg/L (below which most marine fauna cannot persist). Reference condition: site-specific or regional historical baseline.

ECT Class B—Hydrological and hydrochemical characteristics

• Water clarity / light attenuation (Kd490): relevant for seagrass and coral light requirements. Unit: m⁻¹. Reference condition: site-specific long-term mean from satellite ocean colour products. Condition indicator decreases as Kd490 increases beyond the reference level (higher attenuation = reduced light penetration = poorer condition).
• Dissolved inorganic nutrients (NO₃, PO₄): eutrophication indicators for coastal zones. Unit: μmol/L. Reference condition: oligotrophic background level for the relevant ocean region. Condition indicator decreases as nutrient concentrations exceed reference levels.
• Mixed layer depth (MLD): indicator of ocean stratification and vertical mixing, relevant for pelagic and open-ocean accounting units. Unit: metres. Reference condition: long-term climatological mean MLD from ARGO programme or CMEMS reanalysis. Shoaling of MLD beyond the climatological range indicates increased stratification, which constrains nutrient upwelling and affects productivity. Condition indicator decreases as MLD shoals beyond the climatological reference range. Source: ARGO programme density profiles; CMEMS physical reanalysis products. See TG-6.5 Pelagic and Open Ocean Accounting for guidance on MLD in pelagic condition accounts.

Tiered indicator selection approach

Indicator selection follows a three-tier framework based on data availability. Table 3.2.2 below summarises the three tiers.

Rescaling and aggregation pathway

The transition from raw measurement to reported condition indicator follows three steps, consistent with SEEA EA §§5.44--5.58. Table 3.2.3 below summarises these steps.

3.3 Measurement Protocols

In-situ measurement

CTD (conductivity-temperature-depth) casts provide vertical profiles of temperature, salinity, and (where equipped) dissolved oxygen and fluorescence. Moored buoys record continuous surface and sub-surface time series. ARGO profiling floats provide open-ocean profiles to 2,000 m depth with near-real-time transmission. Repeat hydrography programmes (WOCE/GO-SHIP sections) provide basin-scale reference data for calibration.

For coastal and estuarine zones, national tide gauge networks, fixed moored sensors, and periodic survey cruises provide the primary in-situ data. Water sampling for nutrient and pH analysis should follow IOC-UNESCO/IODE standard operating procedures, including certified reference materials for pH (NOAA-distributed CRMs) and oxygen (Winkler titration).

Remote sensing

Satellite-derived SST (MODIS Aqua, AVHRR, Sentinel-3 SLSTR) provides daily global coverage with approximately 1 km spatial resolution. Ocean colour products (MODIS Aqua, SeaWiFS, Sentinel-3 OLCI) deliver Kd490 and turbidity estimates at 300 m—4 km resolution. Gap-filling via optimal interpolation (OI) is standard for SST products (NOAA OISST, CMEMS, GHRSST Level 4). See TG-4.1 Remote Sensing and Geospatial Data for sensor selection and pre-processing protocols.

Tiered minimum time requirements

Table 3.3.1 below summarises the minimum time-coverage requirements for each tier.

Quality control

Apply IOC-UNESCO/IODE data quality procedures: flag anomalies, report measurement precision, document spatial representativeness, and record gap-filling method. All sampling stations must be georeferenced to WGS84. Temporal frequency: annual composites are the minimum for condition accounts; sub-annual composites improve trend analysis. Calibration of satellite retrievals against in-situ measurements should be performed where co-located data exist, with calibration coefficients and residual uncertainty documented in product metadata.

Data-poor contexts

Where national in-situ programmes are limited, global reanalysis products provide viable Tier 3 alternatives: CMEMS Physical Analysis and Forecast System, ERA5-ocean, and WOA climatologies. These products carry known systematic biases in shallow coastal regions and enclosed seas—document the source, version, and known limitations explicitly in the metadata record per TG-0.7 Quality Assurance Principles. Consider supplementing with citizen science and community-based monitoring data (TG-4.4 Citizen Science and Community-Based Monitoring) where national survey coverage is sparse.

3.4 Data Requirements and Sources

The following matrix specifies the minimum data requirements for each recommended physical condition indicator:

Key institutional data sources: Copernicus Marine Service (CMEMS), NOAA CoastWatch, NOAA Coral Reef Watch, IMOS (Australia), ICES Data Centre, IOC-UNESCO IODE, national oceanographic institutes, national meteorological services, Argo Data Assembly Centres (GDAC).

Known data limitations:

• Sparse in-situ coverage in developing-country EEZs creates reliance on Tier 2--3 products with higher uncertainty.
• Satellite retrievals are degraded by cloud cover, sun glint, river plumes, and coastal adjacency effects within approximately 1 km of the shoreline.
• Reanalysis products may not resolve sub-mesoscale processes in enclosed seas, shallow bays, or areas of strong tidal forcing.
• BGC-ARGO pH and oxygen profiles are sparse in coastal and shelf regions; WOD/GLODAP cover is more complete in the open ocean.
• MLD estimates from density profiles are sensitive to the threshold criterion applied (density or temperature criterion); the criterion used must be documented for reproducibility.

Compiling agencies should document data source, product version, spatial resolution, and temporal coverage for every variable in the condition account metadata. See TG-4.6 Data Harmonisation and Interoperability for guidance on integrating data from multiple sources.

3.5 Reporting and Integration

Reporting format

Physical condition indicators feed directly into the ecosystem condition account (TG-3.1 Asset Accounts) as rows in the SEEA EA Table 5.2 format, with the following fields for each variable: ECT class, variable name, unit of measurement, reference condition value, current condition value, and rescaled condition indicator (0--1 scale).

Rescaling formula

For variables where higher values indicate better condition (e.g., dissolved oxygen, aragonite saturation):

CI = (V—V_min) / (V_ref—V_min)

For variables where lower values indicate better condition (e.g., temperature anomaly, nutrient loading, turbidity):

CI = 1—[(V—V_min) / (V_max—V_min)]

Where:

• CI = condition indicator (0--1 scale; 1 = at or above reference condition; 0 = fully degraded or absent)
• V = observed variable value for the accounting period
• V_ref = reference condition value (the value of the variable under the chosen reference condition)
• V_min = minimum observed value or ecologically meaningful lower bound (e.g., 0 mg/L for dissolved oxygen; Ωaragonite = 0; pH = 7.75). Examples by variable: dissolved oxygen V_min = 0 mg/L; Ωaragonite V_min = 0; pH V_min = 7.75; temperature anomaly V_min = 0. Compiling agencies must document the V_min and V_max values applied for each variable in condition account metadata.
• V_max = maximum observed value or ecologically meaningful upper bound (used in negative-condition formula only)

For binary threshold indicators (e.g., DO hypoxia): CI = 1 if V ≥ threshold (2 mg/L); CI = 0 if V < threshold (step rescaling).

Where current condition exceeds the reference baseline (CI > 1 in the positive-condition formula), CI is reported as 1.0 in the condition account table and the raw calculated value is recorded in the accompanying metadata. This situation can arise where recent conservation gains have improved condition above the historical reference.

Integration with biological condition

For integration with biological condition data (TG-4.9 Biological Condition Measurement): physical variables establish the abiotic habitat envelope. Report SST, dissolved oxygen, turbidity, and pH alongside biological indicators to enable attribution analysis—for example, whether a decline in biological condition indicators is driven by thermal stress, oxygen depletion, or reduced light availability rather than direct anthropogenic impact on biota.

Uncertainty reporting

Uncertainty reporting is required alongside each indicator: specify measurement error, temporal coverage, spatial resolution, gap-filling method, and data source reliability tier (1, 2, or 3). Uncertainty estimates should distinguish between measurement uncertainty (instrument precision), representativeness uncertainty (spatial or temporal extrapolation), and model uncertainty (where reanalysis products are used). Dataset versioning and metadata documentation must follow TG-0.7 Quality Assurance Principles—include dataset provenance and version in all metadata records.

4. Acknowledgements

This Circular has been approved for public circulation and comment by the GOAP Technical Experts Group.

Authors: [To be confirmed]

Reviewers: [To be confirmed]