ADR-0057: Warn outside the Staley study area range
Status: Accepted
Date: 2026-09-09
Context
The supplied Staley accepted manuscript, printed page 12, reports study catchments of 0.2–8 km². The earlier WEPPpy 0.02 km² lower bound lacked source support. The owner selected the manuscript as authoritative, then requested warnings rather than hard failures outside its reported range.
Decision
For both M1 and M3, warn below 0.2 km² or above 8 km². Both endpoints are inside the range. Compare full delineated watershed area in unrounded m² against 200,000 and 8,000,000; display conversion/rounding does not control the decision. Present the warning with results and reports, while allowing assessment and preserving numerical outputs. Area alone must not cause a readiness failure, execution rejection or unavailable result.
The specification defines the canonical policy and warning wording. Integration belongs to future project results/UI work; the scalar engine has no area parameter.
Decision Provenance
Decision Venue: repository conversation, 2026-09-09 UTC (2026-09-08 America/Los_Angeles).
Participants Present: repository owner and Codex.
Decision Owner: repository owner, explicitly selecting manuscript authority and a warning without hard failure outside its range.
Implementer: Codex (current contracts); runtime integration pending.
Change Summary
Replace an unsupported 0.02 km² source transcription with 0.2 km² and resolve the previously open area policy to a nonblocking warning outside 0.2–8 km². No equation, probability, inverse calculation, or runtime code changes here.
Rationale and Alternatives Considered
Study coverage is evidence context, not a demonstrated physical validity boundary. A warning communicates extrapolation while preserving the user's ability to assess a basin. Hard rejection was explicitly rejected; silence would hide the study-domain limitation. Do not infer a warning threshold from pfdf tutorial delineation filters.
Evidence
- Supplied manuscript:
wepppy/nodb/mods/postfire_debris_flow/docs/pdfs/staley_2017.pdf, printed page 12; source SHA-256 and reading evidence in the publication audit. - Explicit owner directions in the repository conversation.
Consequences and Risk / Rollback Notes
Out-of-range predictions remain extrapolations, not newly validated results. Future integration must test below, at and above both endpoints, unchanged numerical outputs, and consistent warning presentation across display units. Revise this ADR if the owner changes the accepted source or warning policy; there is no persisted-state migration or runtime rollback in this docs change.