Show the source. Preserve the reasoning.

Traceability connects a claim to its sources and method. Relevance requires the correct field, activity and period. Completeness makes missing required information visible. Consistency surfaces conflicting observations. Reproducibility retains rules, calculations and method versions. Accountability records consequential decisions and changes.

• A record can be complete but unreliable.
• Independent sources may disagree even when both are relevant.
• A reviewer needs the limitations as well as the supporting records.

Each source type needs metadata appropriate to its collection method. A proposed source registry preserves both useful context and interpretation boundaries.

• Satellite imagery: acquisition date, coverage, dataset and processing; may not establish a practice directly.
• Field observation: observer, time, location context and protocol; quality depends on the protocol.
• Photograph: capture time, source and field association; may be ambiguous or incomplete.
• Sensor: device, units, calibration and timestamp; faults and measurement error remain possible.
• Farm record: responsible party and reporting period; a declaration may need corroboration.
• Weather dataset: source, period, units and coverage; regional values may not match exact field conditions.

Different claims require different methods. A methodology record should state its claim or indicator, eligible crops and contexts, input requirements, baseline and comparison rules, calculations, treatment of missing data, uncertainty, acceptance criteria and review responsibilities.

• Retain methodology version and effective period.
• Record why the selected method applies to the claim.
• Do not silently apply revised criteria to an earlier assessment.

This is a proposed registry structure, not a published accredited methodology.

Missing references and duplicate observations can be flagged through structural checks; a reviewer must determine whether sources are redundant or contradictory. Dates outside the reporting window and spatial mismatches require correct association or replacement evidence. Inconsistent units need normalization with source retention.

• Missing required evidence → request the specified source; do not infer completion.
• Contradictory observations → inspect dates, methods and source quality.
• Inappropriate method version → reassess under the applicable version.
• Out-of-range measurement → inspect device, units and collection method; do not automatically label fraud.

Evidence completeness, source quality, agreement, measurement uncertainty and methodology compliance should remain separate from final status. Authorized reviewers inspect records, request information and record rationale. A proposed audit trail captures actions, actors, timestamps and changes to material decisions.

A future numeric score would need a published definition and validation basis. Confidence is not certification.

Illustrative claim: a specified rice practice was implemented in the demonstration season. Required evidence: registered boundary, dated activity log and corroborating observation within the relevant window. Received: boundary and self-reported log. Missing: suitable corroborating observation.

• Rule applied: all required sources must relate to field A-024 and the same reporting period.
• Finding: the log alone does not meet the demonstration evidence requirement.
• Review outcome: more evidence requested; assessment remains preliminary.

This is not a production methodology or real field result. No environmental benefit has been calculated.

Collection, retention, sharing and publication require purpose-specific permissions. An authorized report should expose only the sources necessary for the recipient’s responsibility. Public case studies need separate approval and protection for identifiable participants and precise field locations.