Real vs simulated activation history: convention + DRC/PHL/MMR retrospectives#289
Merged
Conversation
…pectives - INGESTION.md: new rule — never conflate real-life activations (fired under earlier trigger versions) with the SIMULATED would-have-activated record of the current version; cite the team-designated authoritative variant/artifact. - cod-infectious-disease/2026-06-10: authoritative simulated per-year history (ch08 Table 9.4, L2.5 = p98.9 same-province-only, total 26, ~2/yr; per Zack) + TODO to reconcile doc's 174/63+adjacent-mute (L3) with L2.5. - phl-storms/2025-10-03: TODO — 2026 revision in development moves scope to Regions 2/3/5/8 (Caraga dropped); per-region readiness hindcasts on trigger-revision-2026 branch; no committed per-region action hindcast; Drive deck pointers (Trigger Revision 2026, MHIBFEWS, Trigger Review 2025). - mmr-cyclones/development: longer-record ECMWF hindcast retrospective was completed (wind-only): Mala'06*, Nargis'08, Giri'10, Komen'15 (rain), Mora'17, Mocha'23 — six storms, six distinct years; deck pointers + 200-vs-175mm rainfall-threshold discrepancy. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Framework document adopted the L3 spatial constraint (same-province + neighbours), not the DS-recommended L2.5. Replace the L2.5 simulated table with the L3 table from ch08 §9.7 (commit c64323a, 2026-07-17): 2011-2025 = 2,2,1,1,2,1,4,1,1,2,1,0,4 — total 22 over 13 yrs (1.69/yr deterministic, bootstrap 1.97). Reproduced by running the chapter's exact L3 chunk against idsr_zf (gh-pages book not yet rebuilt); matches the chapter's stated 2017=4 correction. INGESTION origin note updated to L3. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
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
1 participant
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.
What
Answers to three portfolio questions (DRC cholera / PHL storms / MMR cyclones historical activations under the most recent trigger versions) surfaced a general gap and several page-level ones.
General convention (docs/INGESTION.md)
New rule: real-life vs simulated activation history — never conflate them. Real activations fired under whatever trigger was live at the time; triggers are almost always revised between versions, so "how often would the current framework have activated" must be answered from the current version's backtest, not the real-life list. Calibration analyses explore multiple variants — record which one the framework team designates as authoritative, with an exact artifact pointer.
frameworks/cod-infectious-disease/2026-06-10.md
Adds the authoritative simulated per-year history (designated by Zack): ch08 "Spatial Constraint Alternatives" Table 9.4,
idsrbranch — L2.5 setting (p98.9, same-province-only persistent muting; 2025 thresholds 182/67). Total 26 over 13 triggerable years (~2/yr; 2017: 5, 2025: 7, 2024: 0). TODO left: the endorsed PDF states 174/63 + same-or-adjacent muting = the L3 variant — reconcile which setting the live monitor implements.frameworks/phl-storms/2025-10-03.md
TODO recording that the 2026 revision in development moves scope to Regions 2/3/5/8 (Caraga dropped); per-region readiness hindcasts live on the
trigger-revision-2026branch; no committed per-region action-trigger hindcast exists (510 runs are in external files); Drive deck pointers (Trigger Revision 2026, MHIBFEWS review, Trigger Review 2025 — the latter's "Storm Impact by Region" table is an uncaptioned slide visual).frameworks/mmr-cyclones/development.md
TODO recording the completed longer-record ECMWF-hindcast retrospective the page was missing (wind-only, results CSV in Drive): would-have-triggered = Mala 2006 (observed basis), Nargis 2008, Giri 2010, Komen 2015 (rain), Mora 2017, Mocha 2023 — six storms, six distinct years. Also flags the 200 mm (proposal deck) vs 175 mm (deployed
constants.py) rainfall-threshold discrepancy.Stubs are deliberately TODO-comments per the capture-as-you-go convention; folding them into page prose can follow.
🤖 Generated with Claude Code