How Trazo's Records and Calculations Work

1. The Records Model: One Entry, Two Outputs

Trazo's unit of work is the activity record: a dated, parcel-linked entry — an input application, a harvest, an irrigation set, a cleanout — whose structured fields follow what 7 CFR 205.103 recordkeeping asks an organic operation to retain: what was done, where, when, with what, by whom. Input applications carry product, rate, area treated, and applicator; harvests carry quantity, how the amount was determined, destination, and buffer-row disposition; equipment shared with conventional ground links to its cleanout log.

The same records then serve two audiences without re-entry. For the certifier: records assemble into Organic System Plan modules that follow the Common OSP structure (§ 205.201(a)), per-parcel applications (§ 205.202), an audit PDF with inline Part 205 citations, and a renewal packet. Where Trazo can derive a suggested answer from the season's records, it presents the suggestion with its evidence named — the operator confirms or edits every answer, because the operator authors the plan. For the buyer: the same records produce a scannable product page showing documented practices, photos, and third-party certifications — facts, never a self-assigned grade.

A design consequence worth stating: because Trazo measures documentation completeness rather than scoring environmental impact, a farm is never penalized for logging more. More records mean a more complete file — for the inspector, and for the buyer.

2. Documentation Completeness (Internal Metric)

Trazo computes a 0–100 completeness metric internally — historically called the transparency score. Since August 2026 it is not displayed as a number anywhere: not to consumers, not on the buyer page, and no longer as a widget on the producer dashboard either, where the farmer instead sees the per-record checklist of what is still missing. It is never presented as a sustainability rating of any kind. It measures documentation completeness: what percentage of expected farming practices have been recorded for a given production cycle, adjusted for documentation quality (photo evidence).

Seven Weighted Categories

Category weights were determined by the Trazo team based on consumer concern analysis and documentation impact assessment. Production activities receive the highest weight (30%) as they represent core farming practices. Weights have not been externally validated and may be adjusted in future versions based on peer feedback.

Formula

Expected event counts vary by crop type and farming approach (conventional, organic, biodynamic). For example, tree fruit crops expect 12 events for conventional farming, 15 for organic, and 18 for biodynamic.

Why there is no grade

Trazo does not assign a letter grade, tier, medal or badge to a farm. An earlier version of this document published an A+ to F scale; that construct was removed from the product in August 2026 and is no longer published here. A self-created rating shown to consumers is the label class restricted by EU Directive 2024/825 and carries general-environmental-benefit risk under FTC 16 CFR 260.6 — and a grade Trazo invented says more about our thresholds than about the farm. The score below is a completeness measure on the producer's own dashboard; consumers see the documented practices, photos and third-party certifications themselves.

3. Anti-Gaming Protections

A scoring system is only as credible as its resistance to manipulation. The following rules were implemented after a comprehensive gaming vector audit (March 2026):

• Zero-event protection: A production with zero logged events always receives 0%, regardless of any other factors.
• Minimum events threshold: Productions with fewer than 3 events are capped at a maximum score of 30%.
• USDA Organic verification: Organic certification claims are cross-referenced against the USDA Organic Integrity Database (36,000+ operations) using fuzzy string matching at a 90% confidence threshold. Results display as "Verified via USDA Organic Integrity Database" (green badge) or "Self-Reported" (gray badge).
• Regional irrigation defaults: Irrigation depth defaults are adjusted by state (e.g., 2.5" for Arizona vs. 1.0" for humid eastern states) to prevent underestimation in arid regions.

4. Photo Verification System

Photos provide accountability beyond text-only logging. A three-tier penalty structure incentivizes visual documentation for high-stakes events:

Photos undergo 5-layer anti-gaming validation: file size checks (10KB–50MB), format verification (magic byte analysis), blank image detection (pixel variance), and timestamp correlation with event dates (within 7 days).

5. Carbon Calculation Methodology

Trazo calculates Direct Operational Emissions, meaning emissions from activities that farmers directly control during the production cycle. This is not a full Life Cycle Assessment (LCA). It covers only emissions under direct farmer control and excludes supply chain, transportation, retail, and end-of-life emissions.

Calculation Examples

Method-Aware Accuracy for Manual Labor

USDA per-hectare fuel factors (Section 6) assume mechanical equipment. When a farmer documents a manual method (for example, "Hand thinning with pruning shears"), applying a mechanical fuel average would overstate emissions by 10–100x.

Trazo detects manual operations from the farmer's free-text method field and routes to a manual labor estimate (0.1 kg CO₂e per acre), representing hand-tool wear and worker logistics. Mechanical keywords (tractor, harvester, sprayer) take precedence on ambiguous descriptions.

This ensures that farmers who document more receive more accurate carbon values, not blanket averages. When a farmer explicitly reports zero fuel consumption, the calculator respects that value rather than falling back to USDA estimates.

6. Emission Factor Sources

All emission factors are sourced from government publications and peer-reviewed literature. Each factor includes a confidence level (0.0–1.0) representing source evidence strength.

Fuel Combustion (EPA 40 CFR Part 98)

Electricity (EPA eGRID 2023)

US Average: 0.39 kg CO₂/kWh (delivered electricity, including 5.12% transmission & distribution losses). Country-specific factors available for Uruguay (0.04), Brazil (0.07), Colombia (0.13), Peru (0.19), Chile (0.31), Argentina (0.35), and Mexico (0.42).

Correction note (2025-12-03): This factor was corrected from 0.821 to 0.39 kg CO₂/kWh after discovering the previous value was 2.1x too high due to a unit conversion error.

Fertilizer Manufacturing

Chemical Manufacturing

Additional chemical-specific factors are available for copper fungicide (5.5), sulfur fungicide (1.8), biological fungicide (2.0), organic herbicide (1.2), and glyphosate (31.3 kg CO₂/kg AI). The full emission factor table with confidence levels and detailed source citations is available in the complete methodology document.

Equipment Operations (USDA ARS)

Per-hectare fuel factors for mechanical operations. When a farmer documents manual methods, the manual labor factor (0.1 kg CO₂e/acre) is used instead (see Section 5).

7. NOP Audit PDF Methodology

For organic operations, Trazo generates a 24-page Audit PDF designed to support certifier review under the USDA National Organic Program (7 CFR Part 205). The PDF is reference material farmers attach to their certifier's annual inspection package. Trazo is a documentation platform and is not a USDA-accredited certifier; the PDF does not constitute certification or any determination of compliance.

7.1 Document Structure

Each export is filtered by certification type (organic, GLOBALG.A.P., Regenerative Organic, or all). The organic-typed PDF contains the following sections, in order:

• Page-1 Compliance Snapshot: 13+ row table citing 7 CFR § 205.201 (OSP), § 205.201(a)(2) (OSP narrative vs. registry sync), § 205.103 (records), § 205.103(b)(2) (SOE traceability), § 205.202 (land requirements), § 205.203(c) (compost / raw manure), § 205.206 (pest management), § 205.272 (contamination prevention), § 205.307 (lot traceability), § 205.400(f)(2) (unapproved-input drift findings), and § 205.201(a)(3) (OFPP). Each row is automatically classified PASS / REVIEW / FAIL based on the data the farmer has entered.
• Producer Attestation (§ 205.401(d)): electronic signature block per 21 CFR Part 11 trustworthy-records principles.
• Organic System Plan reference: active OSP version, status, and § 205.201 narrative.
• Approved Inputs Registry: OMRI-Listed and National List substances on file for the operation.
• Drift Findings: per § 205.400(f)(1), inputs used that are not on the approved registry, with explanation.
• Field Events: chemical applications, soil management, pest management, equipment cleaning, weather, and general documentation, each with timestamps, GPS, and CFR-paired column headers.
• Compost & Manure Handling: compost temperature curves vs. § 205.203(c)(2) (windrow or static pile); raw manure 90-day or 120-day pre-harvest interval per § 205.203(c)(1).
• Pest IPM Threshold Monitoring: scouting observations, action thresholds, and rationale per § 205.206(a) and (e).
• Photo Gallery: embedded thumbnails in NOP-priority order (pest, soil/compost, chemical, equipment cleaning, contamination, then routine production), with EXIF date/GPS captions. Capped at 4 photos per event and 60 total.
• SOE Compliance Status: attestations per the 2024 Strengthening Organic Enforcement final rule: Fraud Prevention Plan (§ 205.201(a)(3)), 5-Year Retention (§ 205.103(b)(4)), Unannounced Inspection Readiness (§ 205.661), Non-Retail Container Labeling (§ 205.307), and NOP Import Certificate. Each attestation reports actual platform state, never a false claim.
• Record Change History: every edit logged with user attribution and timestamp; flagged when changes occur after production finalization.

7.2 Drift Detector Logic

When a farmer logs a chemical or soil amendment event, a post-save signal compares the input against the operation's Approved Inputs Registry using a three-layer cascade:

1. Foreign-key membership check: if the event references a specific OMRI Product or National List Substance by ID, the detector confirms that ID is in the approved set.
2. Fuzzy name match: if no FK is set, the free-text commercial name is matched against the OMRI catalog and National List using token-based comparison with brand- anchor constraints. Confidence below the threshold is treated as unresolved.
3. Unresolved → flagged: events that match no approved input are surfaced as drift findings with a § 205.400(f)(1) citation. The detector never blocks a save: farmers may use unapproved inputs provided they notify their certifier per § 205.400(f)(1).

Fuzzy matching can produce false positives or false negatives. Drift findings should be treated as advisory and reviewed before sharing the PDF with a certifier.

7.3 What the PDF Does Not Do

• It does not certify. Only USDA-accredited certifying agents (CCOF, Oregon Tilth, MOFGA, OEFFA, PCO, and others) issue organic certificates. The PDF is reference material, not a certification.
• It does not verify. Trazo does not visit farms, sample soil, audit supplier invoices, or independently validate that documented practices actually occurred.
• It does not guarantee outcomes. A farmer whose PDF passes Trazo's automated snapshot may still receive a non-compliance finding from their certifier.
• It is compliance-assisting, not compliance-granting. The intended use is to make inspector pre-read faster, not to replace inspector judgement.

7.4 Tamper-Evidence Footer

Each photo in the gallery includes a footer listing the original S3 object key, upload timestamp, file size, and uploader account, metadata retained in the producer dashboard for cross- reference. The PDF does not currently include a SHA-256 hash on the file itself; that capability may be added in a future release.

8. Known Limitations

Transparency about limitations is fundamental to this methodology's credibility.

• Self-reported data: The transparency score is entirely based on farmer-reported data. There is no independent verification of event accuracy. The consumer-facing disclaimer reads: "Based on farmer-reported documentation. Not independently audited."
• Documentation completeness, not quality: The score measures how many events were documented, not whether the documented practices are good or sustainable.
• Photo content verification: Photo anti-gaming validates file integrity and temporal consistency but cannot verify that the photo depicts the actual event on the actual farm.
• Carbon scope: Carbon calculations cover direct operational emissions only (Scope 1 equivalent). This is deliberately narrower than a full Life Cycle Assessment per ISO 14067.
• Emission factor age: Some factors derive from studies published in 2004–2009. While these remain the most widely-cited sources in agricultural carbon literature, annual review against newer publications is planned.
• EU regulatory status: Under the EU Directive 2024/825 (effective September 2026), a self-created sustainability score shown to consumers could be classified as a self-created sustainability label. For this reason the transparency score is not displayed on the consumer QR page. It is an internal completeness metric only, and consumers see the documented practices and third-party certifications directly. Full EU substantiation may still require external methodology audit.

9. Regulatory Context

This methodology was developed with reference to the following standards and regulations. Trazo has not undergone formal conformity assessment against these standards.

• FTC Green Guides (16 CFR Part 260): No implied third-party endorsement. All emission factors cite specific government publications. Blockchain described as "blockchain-recorded" (immutability), never "blockchain-verified."
• ISO 14021:2016 (Self-Declared Environmental Claims): Claims are specific, substantiated, and verifiable. Methodology published openly.
• ISO 22005:2007 (Food Traceability): Production event tracking follows traceability design principles: each event records who, what, when, where, and how.
• ISO 14067:2018 (Carbon Footprint of Products): Carbon calculations reference ISO 14067 principles for emission factor selection, though Trazo's direct operational scope is narrower than a full product carbon footprint.

10. References

Government Data Sources

1. U.S. Environmental Protection Agency. Greenhouse Gas Emission Factors Hub (40 CFR Part 98). epa.gov
2. U.S. Environmental Protection Agency. eGRID2023 Summary Data. epa.gov/egrid
3. U.S. Energy Information Administration. (2023). How much carbon dioxide is produced per kilowatthour of U.S. electricity generation? eia.gov
4. IPCC. (2006/2019). Guidelines for National Greenhouse Gas Inventories. N₂O emission factors (EF1=1%, GWP=265). ipcc-nggip.iges.or.jp
5. USDA Natural Resources Conservation Service. National Engineering Handbook, Chapter 8: Irrigation. nrcs.usda.gov
6. USDA Agricultural Marketing Service. USDA Organic Integrity Database. organic.ams.usda.gov

Peer-Reviewed Literature

6. Audsley, E. et al. (2009). Estimation of the Greenhouse Gas Emissions from Agricultural Pesticide Manufacture and Use. Cranfield University.
7. Hillier, J. et al. (2009). The carbon footprints of food crop production. International Journal of Agricultural Sustainability, 7(2), 107–118.
8. Lal, R. (2004). Carbon emission from farm operations. Environment International, 30(7), 981–990.