Repository navigation
Investigate #68: pooled flood surface confirmed, max-per-watercourse fix rejected - #69
Merged
NewGraphEnvironment merged 9 commits intoOct 9, 2026
Conversation
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
Failing on main by design: the bundled tile loses up to 12,220 valley cells when a blue line is added, and the creek fixture loses 46 under the pooled IDW. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
fl_flood_depth(groups =), fl_flood_model(groups =) and fl_valley_confine(group_field =, groups =) interpolate each watercourse's waterline from its own cells and keep the highest. The #68 tests pass, including exact monotonicity in added blue lines. Not ready to merge: on the bundled tile the grouped extent is +67% to +143% over the pooled one and grouped ff2 exceeds pooled ff6. Recorded in findings.md; the default is held for a user decision. Four existing pins fail under the new default (attribute fixture, stac vignette cache). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
The user chose no code for #68: the per-blue-line max extent is +67% to +143% on the bundled tile and flood_factor nearly stops mattering. Implementation and tests remain in history at d71bbb5. The PR carries the measurements and the issue rewrite. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW
NewGraphEnvironment
deleted the
68-fl-flood-depth-averages-every-stream-s-f
branch
October 9, 2026 14:20
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
co_ff04. The 451.9 ha (1.27%) loss reproduces exactly on main. 78.6% of the lost cells are dropped by the all-streams flood mask, which is the pooled IDW blend at work.blue_line_keyand keeps the cell-wise max. It passed every property test, including exact monotonicity in added blue lines. But it grows the bundled floodplain +67% (ff6) to +143% (ff2), and ff2 exceeds today's ff6. Per-order grouping does the same (+52% to +101%). The implementation and its tests are in history at d71bbb5 and reverted on the branch.research/flood_surface_interpolation.md, with the durable verdict and^research$in.Rbuildignore, and a CLAUDE.md design decision so the fix is not re-tried.Related Issues
Test plan
R/,man/ortests/. The branch diff isplanning/,research/,CLAUDE.mdand.Rbuildignore.research/excluded from the build (^research$)Notes
scipy griddata(BlueGeo / bcfishpass).planning/archive/2026-10-issue-68-flood-surface-blend/.🤖 Generated with Claude Code
https://claude.ai/code/session_0186ZefY3qMPiW8pwnwwMuxW