UC12: Instrument Lower Limit of Detection (LLOD) Flags and Values

[gaochen-larc]

Context:
When a measurand value falls below an instrument's sensitivity threshold, an actual physical value cannot be reported. The lower limits of detection (LLOD) are often handled using data flag codes, where the LLOD value itself takes one of two structural forms:

1. Static / Constant LLOD (Header-based): A fixed threshold recorded in site metadata or file headers.
   Example (EBAS steel canister data): Detection limit: 5.0 pmol/mol, Detection limit expl.: Estimated, values below detection limit are flagged with 780.

2. Time-Varying LLOD (Variable-based): A threshold that changes continuously with ambient conditions, instrument drift, or sampling time, stored as a dedicated time-series variable.
   Example (NASA AMS field campaign data): Flag code defined as LLOD_FLAG: -8888, with time-varying LLOD values recorded in a paired data variable (e.g., OA_DL_AMS).

Challenge:
The metadata schema must systematically record both static and dynamic LLOD values, along with their associated flag semantics, to enable automated data discovery and filtering without requiring users to download and parse individual data files.

[gaochen-larc]

Gemini assessment under guidance:
Both schemas can be easily adjusted to properly handle real-world Lower Limit of Detection (LLOD) requirements:

For wmdr2-bundled.json: Add standardized JSON properties for static detection limits on instruments, introduce a relatedVariable tag to link dynamic LLOD time series to primary measurements, and include a flagDefinitions block to expose flag semantics (like 780 or -8888).

For WMDR2.pdf: Explicitly standardize a static limitOfDetection attribute within Equipment or ObservingConfiguration capabilities, building on its already strong support for dynamic streams (ObservationSeries) and quality flag definitions (ReportingProcedure).

By making these targeted adjustments, both the lightweight GeoJSON format and the full WMDR 2 domain model can fully support real-world atmospheric chemistry workflows without losing machine actionability.