Methods and limits
How the public Tyee completion tracker is built, what it will not say, and where official products remain authoritative.
Source lineage
The official 2002+ daily series is the DFO Tyee Test Fishery daily report. Method context: Tyee parameter page. Source id: dfo-tyee-daily. Complete title: DFO Tyee Test Fishery daily report. Author: Fisheries and Oceans Canada. Source version URL: https://www-ops2.pac.dfo-mpo.gc.ca/fos2_Internet/Testfish/rptDTFDTyee.cfm.
Owner review of the official DFO terms of use on 2026-08-20 found that published DFO website information may be reproduced for personal or public non-commercial use without charge or further permission, unless otherwise specified. Reusers must exercise due diligence, name the complete title and author when available, and identify the source version URL. Commercial redistribution requires prior written permission from DFO's copyright administrator unless otherwise specified. Government of Canada trademarks and official symbols are not licensed for reuse. That finding applies to the public DFO Tyee pages used here unless a page-specific notice says otherwise. It does not establish terms for other sources. The Skeena Knowledge Trust historical compilation remains not owner-verified for reuse.
Operational missingness
Non-fishing days are omitted official rows. They are operationally missing, never biological zeroes. A cumulative total may carry forward for display; that carry-forward does not create an operated observation.
Gear era
DFO changed the Tyee net in 2002. The public champion uses only the 2002+ regime. 1998 and other pre-2002 seasons stay labeled context and stay out of the primary donor pool.
Uncertainty
The displayed band is the central 80% of post-2002 same-date completion outcomes. It is descriptive historical context, not a predictive interval. Range type: descriptive_historical_range_not_predictive_interval. central 80% of post-2002 same-date completion outcomes.
AI boundary
Deterministic software ingests, validates, models, scores, and publishes. Language models may author code, review code, and explain outputs. They may not invent observations, choose a favorable result after seeing an outcome, alter an issued record, or make a management decision.
Official products
This tracker is independent public-interest research. It is not an operational fishery forecast, not an official management product, and not endorsed by DFO, the Province of BC, First Nations, or data partners. Official agency assessments remain authoritative.
Issued limits
- The historical completion range is descriptive, not a calibrated predictive interval.
- 1998 is a pre-2002 gear-regime comparison and is not used in the primary model.
- No index-to-fish abundance conversion is published because conversion uncertainty is unmodeled.
- No Aug 30 or late-September timing probability is published because timing paths are not independently validated.
Checked daily means the refresh is inspected every day and published after validation. It does not promise that the official observation will change every day.