Broaden official source portfolio
GeoIntel CI / docs-smoke (push) Canceled after 0s
GeoIntel CI / contract-smoke (push) Canceled after 0s

This commit is contained in:
Codex
2026-07-16 02:04:31 +02:00
parent 8628582caa
commit fdf3881592
10 changed files with 970 additions and 197 deletions
+147 -96
View File
@@ -1,130 +1,181 @@
# Datavindplaats Source Roadmap
This roadmap lists public Vlaamse Datavindplaats sources that are strong
additions for GeoIntel after the current Mol/Kempen foundation. Every source
must still receive a governed operator contract, source evidence retention,
CRS/unit validation and selection metrics before being shown as loaded data.
This roadmap was rebalanced on 2026-07-16 after a platform-wide review of
public Vlaamse Datavindplaats sources. Water is one analysis domain, not the
organizing principle of GeoIntel. The target product remains simple:
## P1 - Hydrology And Water Risk
1. select an area on the map;
2. choose a subject;
3. read source-correct measurements;
4. compare persisted observations when a valid time series exists;
5. inspect provenance and export evidence.
### Vlaamse Hydrografische Atlas - Waterlopen
Every source below is either already operational or has an official public
catalog/service record. A catalog entry is not the same as an implemented
provider. GeoIntel may label a source operational only after governed import,
CRS/unit validation, persisted provenance and selection-metric tests pass.
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/vlaamse-hydrografische-atlas-waterlopen-toestand-03-02-2026
- Service: https://www.vlaanderen.be/datavindplaats/catalogus/ogc-api-features-vlaamse-hydrografische-atlas-waterlopen
- Type: vector watercourse axes.
- Value: replace mixed GRB water context with authoritative watercourse
network metrics: length by category, navigable/non-navigable classes,
manager, basin/zone context and proximity to selected area.
- Caveat: line network is not water surface, discharge, depth or volume.
## Current Operational Foundation
### Afstromingskaart - meervoudige stroomlijnen
- Space and buildings: GRB building/road context, Buildings and Addresses
Register for Mol, official land-use snapshots and orthophotos.
- Nature and agriculture: regional BWK/Natura 2000 state 2025 and definitive
agricultural parcel editions.
- Soil and relief: bounded DHMV DTM/DSM acquisition with height, relief and
slope metrics.
- Climate and environment: VMM flood-hazard scenarios and station-bound
Waterinfo series.
- Historical context: official 1778, 1873 and 1969 mapped land-use classes and
modern land-use editions where persisted.
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/afstromingskaart-meervoudige-stroomlijnen
- Type: 5 m GeoTIFF, EPSG:31370.
- Value: runoff-risk and drainage-context layer for drawn rectangles and
parcels; useful beside DHMV and VMM flood scenarios.
- Caveat: modelled potential flow pattern, not observed water movement.
This is a strong base, but population/services, soil classes and modeled
accessibility are still genuine coverage gaps.
### Overstromingsgevoelige gebieden pluviaal/fluviaal
## Domain A - Space, Buildings And Economy
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/overstromingsgevoelige-gebieden-pluviaal
- Type: watertoets/informatieplicht map layer.
- Value: policy/legal flood-sensitivity overlay complementary to the current
VMM modelled water-depth scenario rasters.
- Caveat: do not merge with VMM depth rasters into one metric; it answers a
different policy question.
### Highest-value additions
## P2 - Soil, Infiltration And Erosion
- [Landgebruik Vlaanderen 2022](https://www.vlaanderen.be/datavindplaats/catalogus/landgebruik-vlaanderen-toestand-2022)
- 10 m GeoTIFF in EPSG:31370 with WMS/WCS distribution.
- Measure hectares and share by actual land-use class; do not present it as
legal zoning.
- [Ruimtebeslag Vlaanderen 2025](https://www.vlaanderen.be/datavindplaats/catalogus/ruimtebeslag-vlaanderen-toestand-2025)
- 10 m settlement/space-occupation raster.
- Measure occupied hectares and share beside open space and land use.
- [Open ruimte Vlaanderen 2022](https://www.vlaanderen.be/datavindplaats/catalogus/open-ruimte-vlaanderen-toestand-2022)
- 10 m derived open-space product.
- Keep its method separate from raw land-use classes.
- [Kernen, linten en verspreide bebouwing 2022](https://www.vlaanderen.be/datavindplaats/catalogus/kernen-linten-verspreide-bebouwing-in-vlaanderen-kernen-toestand-2022)
- Morphological vector typology, not a legal planning boundary.
- [WFS Bedrijventerreinen](https://www.vlaanderen.be/datavindplaats/catalogus/wfs-bedrijventerreinen)
- Production WFS for economic areas.
- Measure total/occupied area and retain VLAIO/Digitaal Vlaanderen source
status and edition.
### Digitale bodemkaart van het Vlaams Gewest: bodemtypes
## Domain B - Nature And Agriculture
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/digitale-bodemkaart-van-het-vlaams-gewest-bodemtypes
- Type: vector polygons, EPSG:31370.
- Value: soil texture/drainage/profile metrics by selected rectangle, strong
context for forest, agriculture, runoff and suitability analysis.
- Caveat: based on historical field data from 1949-1971; current drainage can
differ.
- [BWK / Natura 2000 state 2025](https://www.vlaanderen.be/datavindplaats/catalogus/biologische-waarderingskaart-en-natura-2000-habitatkaart-toestand-2025)
is operational regionally. It remains a current state, not an invented annual
nature series.
- [Open geodata agricultural parcels](https://www.vlaanderen.be/datavindplaats/catalogus/open-geodata-landbouwgebruikspercelen)
is operational as annual declared-use area. Parcel identities remain
intentionally unstable between editions.
- Open-space and land-use rasters complement these object layers. Metrics must
not add overlapping classification surfaces as if they were disjoint.
### Potentiele bodemerosiekaart per perceel
## Domain C - Soil And Relief
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/potentiele-bodemerosiekaart-per-perceel-2023
- Type: parcel-level erosion sensitivity.
- Value: erosion risk by selected area, especially useful with agricultural
parcels, slope/DHMV and runoff.
- Caveat: potential erosion modelling; not observed erosion damage.
- [Digital soil map: soil types](https://www.vlaanderen.be/datavindplaats/catalogus/digitale-bodemkaart-van-het-vlaams-gewest-bodemtypes)
- Vector soil polygons in EPSG:31370.
- Measure hectares by soil type, texture and drainage class.
- Historical field mapping from 1949-1971 is baseline evidence, not a claim
about current parcel drainage.
- [Potential soil erosion per parcel 2023](https://www.vlaanderen.be/datavindplaats/catalogus/potentiele-bodemerosiekaart-per-perceel-2023)
- Modelled potential erosion, not observed damage.
- Combine with agricultural use and DHMV slope only while keeping each
source and method explicit.
- [Physical system map](https://www.vlaanderen.be/datavindplaats/catalogus/fysische-systeemkaart)
- Coarse interpretive landscape zones for context, not parcel evidence.
### Fysische systeemkaart
## Domain D - Mobility And Accessibility
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/fysische-systeemkaart
- Type: generalized physical-system zones.
- Value: quick landscape-system interpretation: soil, relief, hydrological
setting, infiltration/kwel sensitivity.
- Caveat: coarser interpretive layer, not parcel-scale evidence.
- Current GRB roads provide map and length context where persisted.
- [Knooppuntwaarde per hectare 2022](https://www.vlaanderen.be/datavindplaats/catalogus/knooppuntwaarde-per-ha-toestand-2022)
- 1 ha modeled public-transport accessibility score.
- [Hoppinpunten and mobility hubs](https://www.vlaanderen.be/DataCatalogRecord/c262d7d0-d90d-4695-a8c6-6427ef4abb38)
- Planned and realized multimodal points; expose status and proximity.
- [Fietssnelwegen WFS](https://www.vlaanderen.be/datavindplaats/catalogus/fietssnelwegen-wfs)
- Region-wide network with realization, quality, lighting and route context.
## P3 - Spatial Planning And Built Environment
Accessibility scores are not live timetable reliability. Network length and
proximity are separate measurements from modeled accessibility.
### Ruimtebeslag - Vlaanderen - toestand 2025
## Domain E - Population And Services
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/ruimtebeslag-vlaanderen-toestand-2025
- Type: 10 m GeoTIFF, EPSG:31370.
- Value: built-up/settlement pressure metrics and evolution, complementary to
land-use and GRB building footprints.
- Caveat: settlement-area concept, not individual building footprint.
- [Statistical sectors, state 2024](https://www.vlaanderen.be/datavindplaats/catalogus/statistische-sectoren-van-belgie-toestand-01-01-2024)
- Spatial aggregation key for population and socio-economic indicators.
- [Population density per hectare 2019](https://www.vlaanderen.be/datavindplaats/catalogus/inwonersdichtheid-per-ha-vlaanderen-toestand-2019)
- 1 ha density surface corrected against Statbel sectors.
- Report observation year and estimation method; never expose address-level
population.
- [Total service level 2022](https://www.vlaanderen.be/datavindplaats/catalogus/totaal-voorzieningenniveau-toestand-2022)
- 1 ha modeled proximity score for daily services.
- [Synthesis node value and service level 2022](https://www.vlaanderen.be/datavindplaats/catalogus/synthese-knooppuntwaarde-en-voorzieningenniveau-toestand-2022)
- End-user classification after the two component scores are available.
- [Municipality indicators](https://www.vlaanderen.be/datavindplaats/catalogus/gemeente-in-cijfers)
- Context for municipal comparison; not a replacement for area-level map
measurements.
### Open ruimte - Vlaanderen - toestand 2022
## Domain F - Climate And Living Environment
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/open-ruimte-vlaanderen-toestand-2022
- Type: 10 m GeoTIFF, EPSG:31370.
- Value: open-space share per selection and regional fragmentation analysis.
- Caveat: derived from land-use/space-occupation/core logic; keep method
separate from raw land-use classes.
- VMM flood-hazard rasters and Waterinfo station series are already governed.
- [Urban heat islands](https://www.vlaanderen.be/datavindplaats/catalogus/hitte-eilanden-in-steden)
- Hittestress evidence with explicit year/method limitations.
- [PM2.5 concentration](https://www.vlaanderen.be/datavindplaats/catalogus/concentratie-fijn-stof-pm25-in-de-omgevingslucht)
and [GES air-quality map 2017](https://www.vlaanderen.be/datavindplaats/catalogus/ges-kaart-luchtkwaliteit-toestand-2017)
- Keep measured station trends separate from modeled area maps.
- [VHA watercourses OGC API](https://www.vlaanderen.be/datavindplaats/catalogus/ogc-api-features-vlaamse-hydrografische-atlas-waterlopen)
and [runoff map](https://www.vlaanderen.be/datavindplaats/catalogus/afstromingskaart-meervoudige-stroomlijnen)
remain useful, but they follow the broader cross-domain foundation.
Watercourse lines are not surfaces, discharge, depth or volume.
### Kernen, linten, verspreide bebouwing - toestand 2022
## Implementation Waves
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/kernen-linten-verspreide-bebouwing-in-vlaanderen-kernen-toestand-2022
- Type: vector typology.
- Value: readable urban morphology: core/lint/dispersed-building context for
Mol and Kempen.
- Caveat: algorithmic morphology; not a legal planning boundary.
### Wave 1 - Cross-domain area profile
### Verstedelijkt, randstedelijk en landelijk Vlaanderen - toestand 2022
Implement one governed Mercator thematic-raster registry instead of separate
ad-hoc provider code. Its fixed allowlist should first cover:
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/verstedelijkt-randstedelijk-en-landelijk-vlaanderen-indeling-op-basis-van-statistische-sectoren-toestand-2022
- Type: statistical-sector typology.
- Value: high-level area profile and comparison between municipalities.
- Caveat: analysis framework, not direct zoning policy.
1. space occupation 2025;
2. open space 2022;
3. population density 2019;
4. node value 2022;
5. total service level 2022.
## P4 - Accessibility And Services
The existing Dataset, DatasetVersion, Job and raster-selection architecture is
sufficient. Each product needs a fixed coverage id, native unit/classes,
observation date, resolution, legend, permitted aggregations and caveat. No
arbitrary WCS URL or layer name may enter the API.
### Knooppuntwaarde per ha - toestand 2022
In parallel, add the digital soil map through the existing canonical vector
DatasetService/VectorFeatureService path. This immediately gives every drawn
rectangle a balanced profile across space, soil, people, services and
accessibility.
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/knooppuntwaarde-per-ha-toestand-2022
- Type: 1 ha raster/grid score.
- Value: public-transport accessibility score for selected areas.
- Caveat: modelled accessibility score, not live timetable reliability.
### Wave 2 - Economic and mobility objects
### Totaal voorzieningenniveau - toestand 2022
Add governed vector operators for business parks, Hoppinpoints and cycle
highways. Persist provider output only through DatasetService and
VectorFeatureService. Selection metrics should be area for business parks,
counts/proximity for Hoppinpoints and length/status for cycle highways.
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/totaal-voorzieningenniveau-toestand-2022
- Type: 1 ha raster/grid score.
- Value: services proximity score for living-quality and spatial-development
analysis.
- Caveat: score is method-dependent and should be shown with source method.
### Wave 3 - Living environment
### Synthese knooppuntwaarde en voorzieningenniveau - toestand 2022
Add heat and air-quality sources only after their spatial scale and temporal
comparability are explicit. Then add VHA/runoff as hydrological context. Do not
derive missing volume, exposure or health outcomes.
- Catalogus: https://www.vlaanderen.be/datavindplaats/catalogus/synthese-knooppuntwaarde-en-voorzieningenniveau-toestand-2022
- Type: synthesized spatial-opportunity classification.
- Value: simple end-user layer for "where is development/accessibility strong?"
- Caveat: derived classification; keep underlying component scores available.
### Wave 4 - Historical completeness
## Recommended Implementation Order
Search for comparable earlier editions of every accepted product. A valid
series requires compatible classes and methods, explicit observation dates
and persisted versions. When methods differ, show side-by-side evidence rather
than a fabricated trend.
1. VHA waterlopen plus afstromingskaart: highest immediate GIS value and
directly improves water/runoff analysis without pretending to know volume.
2. Bodemkaart plus potential erosion: unlocks soil, drainage, infiltration and
agriculture/nature interpretation.
3. Ruimtebeslag/open ruimte/kernen-linten: makes map selection much more
understandable for end users and supports evolution narratives.
4. Knooppuntwaarde/voorzieningenniveau/synthese: adds planning-quality context
after the physical/environmental layers are stable.
## Recommended Next Sprint
Build **Governed Flemish Thematic Raster Registry - Wave 1**. It is the best
next platform step because one bounded, allowlisted acquisition architecture
unlocks spatial planning, open space, population, accessibility and services
without changing the database schema or creating one-off provider paths.
Acceptance requires:
- product registry and honest capability metadata;
- bounded WCS acquisition through the existing Job/Dataset flow;
- exact Area/rectangle selection with product-specific metrics;
- MapLibre image overlay and simple legend;
- source date, unit, method and caveat in every result;
- no startup/browser fetch and no fabricated historical observations;
- focused GIS, API, persistence and frontend tests;
- one Mol live smoke before regional provisioning.