LIM-01
Regional & Spatial Coverage Bounds (Not Global Monitoring)
Coverage BoundsOceanSig focuses on specific regional investigation hubs (e.g., Florida Keys Reef Tract, Monterey Bay, Cascadia Margin). It does not provide continuous 100% global ocean sensor monitoring.
Telemetry is ingested from designated primary stations and regional satellite virtual stations. Unmonitored marine areas, offshore high seas gaps, and non-reporting regional buoys are not captured in live anomaly calculations.
LIM-02
Auxiliary Feeds (IOOS & ERDDAP) Inactive by Default
Ingestion ArchitecturePrimary live telemetry is driven by NOAA NDBC and NOAA Coral Reef Watch. Auxiliary connectors (IOOS, ERDDAP) are currently auxiliary and not active in primary continuous background ingestion.
Operational status indicators clearly reflect whether auxiliary ingestion feeds are active, stale, or offline. Live risk assessments rely strictly on verified primary pipelines.
LIM-03
Acoustic Hydrophone Grounding State (ACOUSTIC_EMPIRICAL_INTEGRATION_NOT_YET_GROUNDED)
Acoustic MethodologyMBARI MARS deep benthic hydrophone integration is governed under a strict invariant framework and is classified as not yet empirically grounded to raw retained audio files.
Calibrated sound pressure levels (SPL dB re 1 µPa) measure ambient sound energy in the 900m benthic canyon. OceanSig enforces strict epistemic invariants: SPL spikes cannot be used as direct proxies for cetacean population abundance, and acoustic silence does not prove animal absence.
LIM-04
Cascadia Hypoxia Investigation is a PARAMETRIC_SYNTHESIS_MODEL
Investigation ClassificationThe Cascadia Margin Hypoxia 2026 investigation is explicitly classified as a parametric synthesis model simulation, not an empirical investigation backed by retained raw CTD cast bytes.
The Cascadia dossier serves as an architectural evaluation benchmark for multimodal alignment and benthic dissolved oxygen stress tracking. Researchers must not cite Cascadia simulation numbers as physical observational telemetry.
LIM-05
Shared Institutional Lineage is Not Independent Replication
Comparison ContractsSynthesizing two NOAA products (e.g. NDBC Buoy 46042 and CO-OPS Tide Station 9413450) does not constitute independent institutional scientific replication.
Both instruments share calibration standards, operational telemetry infrastructure, and institutional quality control within NOAA. OceanSig Comparison Contracts explicitly flag shared institutional lineage (NOAA_FAMILY) and block independent replication claims.
LIM-06
Physical Non-Equivalence of Disparate Sensing Datums
Physical ComparabilitySatellite radiometric skin sea surface temperature (~10–20 µm depth) is physically non-equivalent to shallow coastal pier thermistors (0.8m depth) or moored buoy hull sensors (1.0m depth).
Direct arithmetic subtraction between disparate instruments reflects multi-station differences across disparate supports, NOT a direct physical or spatial thermal gradient. Comparison Contracts evaluate all 18 dimensions before permitting cross-source assertions.
LIM-07
Absence of Evidence is Not Evidence of Ecological Absence
Epistemic GuardrailsThe absence of a reported anomaly, alert, or detection in OceanSig does not prove that an ecosystem is unstressed or healthy.
Sensor outages, sampling latency, geographic data gaps, and unmonitored ecological stress factors (e.g. coral disease, ocean acidification, nutrient runoff) mean that 'no alert' simply indicates no threshold exceedance in reporting sensors.