Traceable estimates, explicit limits.

A useful transport-chain accounting framework — not a certification claim.

ISO 14083:2023 specifies a common methodology for quantifying and reporting greenhouse-gas emissions from passenger and freight transport-chain operations. Farleda uses its public scope as a design reference for leg-level activity, data provenance and reproducibility.

Farleda has not purchased or reproduced the full standard on this page, has not undergone an ISO conformity assessment, and does not claim certification or clause-by-clause conformance. Farleda output is also not automatically CSRD, ESRS E1, EU Taxonomy or CountEmissionsEU compliant: those assessments require organisational boundaries, controls, assurance and the law applicable at the reporting date.

Current rounded factors used by Farleda methodology v1.1.

The primary table is the revised UK DESNZ 2026 conversion-factor flat file published 31 July 2026. DESNZ states that these factors are for UK activities and that other uses are at the user's risk. Farleda therefore presents them as disclosed planning proxies where operator-specific European factors are unavailable.

The current catalogue primarily represents direct operational factors. Flight factors include the DESNZ radiative-forcing uplift. Separate well-to-tank rows in the DESNZ workbook are not currently added. The displayed factor ranges describe scenarios or classes, not statistical confidence intervals. Walking and cycling are zero direct operational emissions only; lifecycle and food-energy effects are out of scope.

Measured, current, historical and modelled values are deliberately labelled differently.

Farleda converts grid intensity to rail intensity with 0.09 kWh per passenger-kilometre. This is a product modelling assumption, not a measured energy value for every train or occupancy. The static 28 g/pkm factor is replaced when country context is available.

For a leg with a departure time, the planner samples the matching UTC hour of a 24-hour profile. With an electricityMaps key, the profile is derived from the provider's recent history and labelled hourly-live. Without a key, Farleda applies a declared, country-specific shape to the annual snapshot and labels it hourly-modelled. This is a heuristic, not a forecast of the future grid.

A cross-border leg currently averages its endpoint-country factors. It is not yet weighted by rail kilometres within each country. electricityMaps documents consumption-based lifecycle intensity, while the embedded annual table is labelled production-based; the two accounting bases are not treated as an interchangeable validation series.

Transitous/MOTIS supplies scheduled public-transport alternatives when a suitable feed and journey are available. Geometry quality varies by source feed; Farleda does not describe every returned line as surveyed track geometry. The current OSRM integration uses a driving profile. Walking and cycling reuse this road geometry as an approximation; dedicated pedestrian and bicycle routing is not integrated.

Flights use the nearest eligible OurAirports pair and a great-circle path. That is an approximation between airport coordinates, not the actual flown track. When routed geometry is unavailable, Farleda uses disclosed great-circle detour factors. Source and quality are carried with the leg.

Candidate public-transport routes are deduplicated by a route signature, evaluated on emissions, duration and transfers, and ranked with weights for carbon, time, balanced or cost preference. The current cost preference uses operational simplicity because Transitous does not provide live bookable fares; it is not a price comparison.

• Weather: Open-Meteo forecast data is used for dates up to 14 days ahead. Farther dates currently show an average of daily values for the corresponding month in 2023 — not an ERA5 multi-year climate normal. The current keyless endpoint is limited by Open-Meteo to non-commercial use; production launch requires a paid customer endpoint or compliant self-hosting.
• Air quality: the current European Air Quality Index and pollutants are fetched from Open-Meteo when requested; it is contextual information, not a medical recommendation.
• Places and stays: OpenStreetMap/Overpass candidates depend on upstream completeness and do not imply availability, certification, price or sustainability verification.
• Klimaticket: the present UI approximates coverage when both endpoints are in Austria. It does not verify service-level tariff validity or every operator's conditions.

The server signs a canonical passport payload with HMAC-SHA256 and the declared methodology version. The public passport verifier checks format, signature and payload consistency against the configured server secret. A valid result shows that the payload has not changed since Farleda signed it. The payload retains rounded leg summaries and declared versions, not a complete frozen set of upstream responses and per-leg factors for independent replay.

It does not independently prove that the traveller completed the journey, that an operator measured the energy, or that the estimate meets an external assurance standard. HMAC verification is controlled by Farleda because the verification secret is server-side; it is not the same as a third-party public-key attestation.

Public metrics are aggregate database results. They expose counts and sums, not individual trip rows, user identifiers or IP addresses. Internally, the underlying product still processes account and trip data as described in the privacy notice; an aggregate endpoint does not make the entire service anonymous.

• Documentation · 2026-10-10: routing-profile limits, passport replay boundary and interpretation of factor ranges clarified. Model version and factors unchanged.
• v1.1 · 2026-08-11: UK DESNZ 2026 revised factors; correct car vehicle-km allocation; grid snapshot/live/modelled labels; explicit accounting and assurance limits.
• v1.0 · 2026-04: initial public methodology and factor API.