Before
Baseline variability, fronts, community, carbon system, gases, particles, ecology and potential rightsholder concerns.
Measurement and MRV explorer
Ocean iron fertilisation: Observation systems and accounting claims.
No instrument observes net atmospheric removal by itself. A defensible design combines physical, biological, chemical, particle and ecological measurements across treatment, reference and downstream domains.
Research observatory architecture
Patch, reference and downstream corridor evolve together.
The measurement architecture resolves the physical frame while tracking carbon and ecology from baseline through seasonal mixing.
During
Material mass balance, tracer, physical transport, biological response, carbonate change and early risk signals.
After
Export, remineralisation, mixing, food-web response, gases and evolving matched references.
Through return
Winter mixing, downstream nutrient effects, ventilation and model-data reconciliation across the accounting boundary.
Source: Buesseler et al. (2024). Scientific guidance for an OIF research observatory. The architecture has not been validated for crediting and does not establish permission
Instrument library
What each method sees—and misses.
Satellite ocean colour
Role: Baseline, response, corridor
Surface chlorophyll proxies, colour, spatial evolution and regional context.
Cannot establish alone: Cloud, ice, depth and algorithm limits; chlorophyll is not carbon export or removal.
Research vessel
Role: All stages
Water-column profiles, process studies, calibration, taxonomy, gases and adaptive sampling.
Cannot establish alone: Discrete coverage, weather and ship effects; expensive time does not guarantee representativeness.
BGC-Argo floats
Role: Baseline, downstream, winter
Repeated autonomous profiles of physical and biogeochemical variables over long periods.
Cannot establish alone: Sensor suite and calibration vary; floats do not remain inside a treatment patch by default.
Gliders
Role: Patch and fronts
High-resolution sections across physical gradients and repeat transects.
Cannot establish alone: Navigation, payload, depth and severe-weather constraints; not a complete carbon budget.
Autonomous surface vehicles
Role: Air–sea exchange
Meteorology, surface pCO₂ and repeated transects without continuous ship presence.
Cannot establish alone: Gas-transfer uncertainty and platform coverage remain; attribution needs a counterfactual.
Material tracer
Role: Input and attribution
Patch identity, dilution, movement and mass-balance constraints.
Cannot establish alone: Tracer and treatment can separate; recovery is incomplete; it does not measure all carbon fate.
Carbonate-system sampling
Role: Uptake and partition
DIC, alkalinity, pCO₂, pH and constraints on inorganic carbon change.
Cannot establish alone: Natural variability, mixing and gas exchange require matched physical observations.
234Th disequilibrium
Role: Particle export
Time-integrated proxy for particulate organic carbon export with site-specific assumptions.
Cannot establish alone: POC:234Th ratios and non-steady-state effects add uncertainty; one depth is not durability.
Sediment traps
Role: Particle export
Directly collected sinking particles, composition and flux at selected depths.
Cannot establish alone: Hydrodynamic bias, swimmers, spatial coverage and deployment duration affect inference.
Optical particle sensors
Role: Aggregation and sinking
Particle abundance, size spectra and high-frequency vertical patterns.
Cannot establish alone: Optics need calibration and composition assumptions; particles can fragment or remineralise.
Genomics and microscopy
Role: Ecology
Taxonomy, functional genes, community shifts and potential toxin-producing organisms.
Cannot establish alone: Presence is not activity, dose or impact; interpretation needs matched chemistry and food-web data.
Dissolved gases
Role: System effects
O₂, N₂O, CH₄ and DMS changes alongside carbon pathways.
Cannot establish alone: Sparse observations and air–sea exchange complicate budgets; effects may be non-local.
Space × time matrix
One cruise cannot cover every return pathway.
| Domain | Before | Response window | Export window | Seasonal / downstream |
|---|---|---|---|---|
| Treatment patch | Baseline and variability | Input, physics, biology, carbon | Particles and remineralisation | Water-mass fate |
| Matched reference | Comparability test | Untreated evolution | Parallel particle observations | Reference drift and divergence |
| Downstream corridor | Initial gradients | Advected treatment signal | Nutrient and food-web effects | Winter mixing, ventilation, non-local change |
| Regional context | Climatology plus observations | Weather and circulation | Background export | Interannual variability |
The table is a research-design concept, not a sampling plan. Spatial resolution, duration, statistical power and stopping thresholds remain proposal-specific.
Data integrity
MRV requires traceable measurements and falsifiable tests.
Study registration
Question, counterfactual, polygon, season, primary outcomes, exclusions, uncertainty method and study thresholds.
Calibrate & reconcile
Instrument metadata, standards, blanks, duplicates, cross-platform comparison, drift and missingness.
Data lineage
Raw observations, processed datasets, analytical code, methods, protocol deviations, quality-control flags and correction history preserve traceability from measurement to estimate.