fix(map): honor raster coverage and temporal dates

2026-07-22 00:18:55 +02:00
parent 80df5e70c9
commit 919f5879e5
9 changed files with 329 additions and 73 deletions
@@ -0,0 +1,74 @@
+# GeoIntel data coverage status
+
+Status date: 2026-07-21
+
+This document is the operational interpretation of the source registry. It
+does not replace the legal/source provenance stored with each Dataset.
+
+## Status labels in the workbench
+
+- `Beschikbaar`: a persisted Dataset covers the selection and can be queried
+  immediately.
+- `Automatisch`: an audited official source contract exists. GeoIntel obtains
+  and persists only the bounded selection when analysis starts.
+- `Alleen huidig`: one official observation exists, so no honest evolution can
+  be calculated.
+- `Ontbreekt`: no operational source contract covers that theme and zone.
+
+`Automatisch` therefore does not mean unavailable or simulated. It means that
+the source is materialized on first use and reused afterwards. AI training
+cannot replace missing official GIS observations and is not used to fabricate
+coverage or historical dates.
+
+## Current operational coverage
+
+| Zone | Persisted or bounded operational sources | Historical comparison |
+| --- | --- | --- |
+| Belgium | NGI administrative boundaries; Statbel population/statistical sectors | Statbel population 2021-2025 |
+| Flanders | GRB buildings, roads, water and parcels; DHMV terrain/surface; VMM flood scenarios; BWK/Natura 2000; DOV soil; policy rasters for space, open space, accessibility and services; agriculture and orthophoto where governed | Population 2021-2025; land-use/land-cover series where retained; agriculture editions; historical maps/orthophotos where the selected product has a real observation date |
+| Wallonia | Bounded PICC buildings, roads and hydrography; governed SPW bed-elevation/bathymetry products | No general cross-theme regional history yet |
+| Brussels | Bounded UrbIS buildings, street axes and cadastral parcels | No general cross-theme regional history yet |
+| Belgian North Sea | RBINS reporting units; Marine Spatial Plan 2026-2034; governed MDK bathymetry only when runtime acquisition is explicitly configured | No multi-epoch bathymetry or marine-plan trend yet |
+
+Mol and the Kempen are golden regression areas. Their persisted partitions are
+not national coverage. A partition is only selected when its recorded
+`bbox_epsg4326` intersects the drawn rectangle; otherwise GeoIntel uses an
+applicable bounded official source or reports the theme as unsupported.
+
+## Priority coverage gaps
+
+1. Govern the public Walloon WALOUS land-cover editions (including the
+   published 2018/2020 change product) as a real regional time series. A class
+   crosswalk is required because the older COSW 2005/2007 methodology differs.
+   Official catalogue:
+   `https://geoportail.wallonie.be/catalogue-donnees?search-text=occupation+du+sol`.
+2. Govern the current public Walloon flood-hazard vector/raster products and
+   retain their model scenario semantics separately from observed floods.
+   Official record:
+   `https://geoportail.wallonie.be/catalogue/14084108-2c7b-4091-b62d-ff0fc235213a.html`.
+3. Add the public UrbIS Land Cover product (regional situation 2024) for
+   Brussels through its official WFS/download contract. Keep it separate from
+   cadastral parcels and buildings. Product specification:
+   `https://urbisdownload.datastore.brussels/UrbIS/TechSpec/LandCover_TechSpec_FR20240401.pdf`.
+4. Add a common Belgium-wide topographic baseline with normalized theme
+   semantics across NGI, Flanders, Wallonia and Brussels.
+5. Govern comparable Walloon and Brussels historical editions before exposing
+   evolution for buildings, roads, land cover, soil, elevation or flood risk.
+6. Add nationally comparable land-cover history with explicit class crosswalks
+   and uncertainty; never compare incompatible legends silently.
+7. Add multi-epoch marine bathymetry and survey-footprint metadata before
+   presenting seabed evolution.
+8. Expand persisted raster partition manifests beyond the regression regions
+   only where repeated use justifies caching; bounded acquisition remains the
+   default for one-off selections.
+9. Add source freshness probes only for publishers with stable official edition
+   contracts. Do not infer a new observation from an import or HTTP date.
+
+## Acceptance rules for a new source
+
+A source is visible as operational only after its licence, authority,
+observation time, CRS, spatial coverage, schema, units and limitations are
+validated. It must persist through DatasetService and the canonical feature or
+raster flow, retain provenance/checksums, return semantic metrics, and have a
+selection-level regression test. Historical support additionally requires at
+least two distinct, methodologically comparable official observation dates.