feat: add cross-domain Mol data profile

backend/README.md

@@ -1289,6 +1289,43 @@ Settings: `FLOOD_HAZARD_ENABLED`, `FLOOD_HAZARD_WCS_URL`,
 `FLOOD_HAZARD_MAX_SIDE_M`, `FLOOD_HAZARD_MAX_PIXELS`,
 `FLOOD_HAZARD_TIMEOUT_SECONDS` and `FLOOD_HAZARD_MAX_RESPONSE_MB`.
 
+## Cross-domain thematic rasters and DOV soil
+
+The governed thematic registry exposes five fixed MercatorNet products through
+`GET .../datasets/thematic-raster/products`. Acquisition uses
+`POST .../datasets/thematic-raster/acquire`; selection and PNG rendering use
+`POST .../raster/thematic/select` and `GET .../raster/thematic/image`.
+
+Provision every product for the exact persisted Mol Area:
+
+```bash
+docker exec geointel python /app/scripts/provision_thematic_rasters.py
+```
+
+Inspect the complete 28-municipality matrix without writes, then run it after
+the Mol source/runtime gate passes:
+
+```bash
+docker exec geointel python /app/scripts/provision_thematic_rasters.py \
+  --project-name "Kempen Regional Workbench" --all-municipalities --dry-run
+```
+
+Settings: `THEMATIC_RASTER_ENABLED`, `THEMATIC_RASTER_WCS_URL`,
+`THEMATIC_RASTER_MIN_SIDE_M`, `THEMATIC_RASTER_MAX_SIDE_M`,
+`THEMATIC_RASTER_MAX_PIXELS`, `THEMATIC_RASTER_TIMEOUT_SECONDS` and
+`THEMATIC_RASTER_MAX_RESPONSE_MB`.
+
+Provision the official DOV soil polygons for Mol through the existing vector
+upload path:
+
+```bash
+docker exec geointel python /app/scripts/provision_mol_soil_map.py
+```
+
+Use `--fetch-only` to retain and validate source evidence without importing.
+The operator never writes directly to PostGIS. Soil drainage and related map
+classes represent the 1949-1971 survey and are not current observations.
+
 ## Waterinfo station histories
 
 Run the explicit operator after the regional workspace and Mol Area exist:

backend/app/api/routes/datasets.py

@@ -26,6 +26,8 @@ from app.schemas import (
     TerrainSelectionRequest,
     FloodHazardAcquireRequest,
     FloodHazardSelectionRequest,
+    ThematicRasterAcquireRequest,
+    ThematicRasterSelectionRequest,
     VectorBBoxResponse,
     VectorBufferRequest,
     VectorClipRequest,
@@ -48,6 +50,8 @@ from app.services.dhmv_acquisition_service import DhmvAcquisitionService
 from app.services.terrain_analysis_service import TerrainAnalysisService
 from app.services.flood_hazard_acquisition_service import FloodHazardAcquisitionService
 from app.services.flood_hazard_analysis_service import FloodHazardAnalysisService
+from app.services.thematic_raster_acquisition_service import ThematicRasterAcquisitionService
+from app.services.thematic_raster_analysis_service import ThematicRasterAnalysisService
 from app.utils.response import envelope
 
 router = APIRouter(prefix="/projects/{project_id}", tags=["datasets"])
@@ -207,6 +211,30 @@ def list_flood_hazard_products(project_id: UUID, db: Session = Depends(get_db)):
     return envelope({"items": items, "total": len(items)})
 
 
+@router.post("/datasets/thematic-raster/acquire", response_model=dict)
+def acquire_bounded_thematic_raster(
+    project_id: UUID,
+    payload: ThematicRasterAcquireRequest,
+    db: Session = Depends(get_db),
+):
+    job = JobService.run_sync_job(
+        db=db,
+        project_id=project_id,
+        job_type="raster.thematic.acquire",
+        parameters=payload.model_dump(mode="json"),
+        operation=lambda: ThematicRasterAcquisitionService.acquire(db, project_id, payload),
+    )
+    return envelope(job)
+
+
+@router.get("/datasets/thematic-raster/products", response_model=dict)
+def list_thematic_raster_products(project_id: UUID, db: Session = Depends(get_db)):
+    if not db.get(Project, project_id):
+        raise AppError(code="PROJECT_NOT_FOUND", message="Project not found", status_code=404)
+    items = ThematicRasterAcquisitionService.list_products()
+    return envelope({"items": items, "total": len(items)})
+
+
 @router.get("/datasets", response_model=dict)
 def list_datasets(
     project_id: UUID,
@@ -560,6 +588,30 @@ def raster_flood_hazard_image(
     )
 
 
+@router.post("/datasets/{dataset_id}/raster/thematic/select", response_model=dict)
+def raster_thematic_selection(
+    project_id: UUID,
+    dataset_id: UUID,
+    payload: ThematicRasterSelectionRequest,
+    db: Session = Depends(get_db),
+):
+    return envelope(ThematicRasterAnalysisService.analyze(db, project_id, dataset_id, payload))
+
+
+@router.get("/datasets/{dataset_id}/raster/thematic/image")
+def raster_thematic_image(
+    project_id: UUID,
+    dataset_id: UUID,
+    db: Session = Depends(get_db),
+):
+    content = ThematicRasterAnalysisService.render_png(db, project_id, dataset_id)
+    return Response(
+        content=content,
+        media_type="image/png",
+        headers={"Cache-Control": "private, max-age=86400"},
+    )
+
+
 @router.get("/datasets/{dataset_id}/raster/stats", response_model=dict)
 def raster_stats(
     project_id: UUID,

backend/app/core/config.py

@@ -53,6 +53,16 @@ class Settings(BaseSettings):
     flood_hazard_max_pixels: int = Field(default=12_000_000, ge=1, validation_alias="FLOOD_HAZARD_MAX_PIXELS")
     flood_hazard_timeout_seconds: int = Field(default=300, ge=1, validation_alias="FLOOD_HAZARD_TIMEOUT_SECONDS")
     flood_hazard_max_response_mb: int = Field(default=160, ge=1, validation_alias="FLOOD_HAZARD_MAX_RESPONSE_MB")
+    thematic_raster_enabled: bool = Field(default=True, validation_alias="THEMATIC_RASTER_ENABLED")
+    thematic_raster_wcs_url: str = Field(
+        default="https://www.mercator.vlaanderen.be/raadpleegdienstenmercatorpubliek/wcs",
+        validation_alias="THEMATIC_RASTER_WCS_URL",
+    )
+    thematic_raster_min_side_m: float = Field(default=100.0, gt=0, validation_alias="THEMATIC_RASTER_MIN_SIDE_M")
+    thematic_raster_max_side_m: float = Field(default=20_000.0, gt=0, validation_alias="THEMATIC_RASTER_MAX_SIDE_M")
+    thematic_raster_max_pixels: int = Field(default=12_000_000, ge=1, validation_alias="THEMATIC_RASTER_MAX_PIXELS")
+    thematic_raster_timeout_seconds: int = Field(default=300, ge=1, validation_alias="THEMATIC_RASTER_TIMEOUT_SECONDS")
+    thematic_raster_max_response_mb: int = Field(default=160, ge=1, validation_alias="THEMATIC_RASTER_MAX_RESPONSE_MB")
     redis_url: str | None = Field(default=None, validation_alias="REDIS_URL")
     log_level: str = Field(default="INFO", validation_alias="GEOINTEL_LOG_LEVEL")
     database_statement_timeout_ms: int = Field(default=5_000, validation_alias="DATABASE_STATEMENT_TIMEOUT_MS")

backend/app/schemas/__init__.py

@@ -51,6 +51,15 @@ from .flood_hazard import (
     FloodHazardSelectionResponse,
     FloodHazardSelectionSummary,
 )
+from .thematic_raster import (
+    ThematicRasterAcquireRequest,
+    ThematicRasterAcquisitionResult,
+    ThematicRasterMetric,
+    ThematicRasterProductRead,
+    ThematicRasterSelectionRequest,
+    ThematicRasterSelectionResponse,
+    ThematicRasterSelectionSummary,
+)
 from .external import (
     ExternalFetchRequest,
     ExternalFetchResponse,
@@ -167,6 +176,13 @@ __all__ = [
     "FloodHazardSelectionRequest",
     "FloodHazardSelectionResponse",
     "FloodHazardSelectionSummary",
+    "ThematicRasterAcquireRequest",
+    "ThematicRasterAcquisitionResult",
+    "ThematicRasterMetric",
+    "ThematicRasterProductRead",
+    "ThematicRasterSelectionRequest",
+    "ThematicRasterSelectionResponse",
+    "ThematicRasterSelectionSummary",
     "VectorBBoxResponse",
     "VectorClipRequest",
     "VectorBufferRequest",

backend/app/schemas/thematic_raster.py

@@ -0,0 +1,96 @@
+from __future__ import annotations
+
+from uuid import UUID
+
+from pydantic import BaseModel
+
+from .operations import VectorSelectionBBox
+
+
+class ThematicRasterAcquireRequest(BaseModel):
+    bbox: VectorSelectionBBox
+    area_id: UUID | None = None
+    product_key: str
+    force_refresh: bool = False
+
+
+class ThematicRasterProductRead(BaseModel):
+    key: str
+    display_name: str
+    theme: str
+    metric_kind: str
+    coverage_id: str
+    native_resolution_m: float
+    source_crs: str
+    source_value_unit: str
+    observation_year: int
+    source_version: str
+    catalog_url: str
+    attribution: str
+    license_note: str
+    legend_min_label: str
+    legend_max_label: str
+    limitation_message: str
+
+
+class ThematicRasterAcquisitionResult(BaseModel):
+    output_dataset_id: UUID
+    reused: bool
+    provider: str
+    product_key: str
+    display_name: str
+    theme: str
+    metric_kind: str
+    coverage_id: str
+    resolution_m: float
+    width: int
+    height: int
+    valid_pixel_count: int
+    bbox_epsg4326: list[float]
+    bbox_epsg31370: list[float]
+    observation_year: int
+    source_value_unit: str
+    attribution: str
+    limitation_message: str
+
+
+class ThematicRasterSelectionRequest(BaseModel):
+    bbox: VectorSelectionBBox
+    area_id: UUID | None = None
+
+
+class ThematicRasterMetric(BaseModel):
+    metric_key: str
+    metric_label: str
+    metric_value: float
+    metric_unit: str
+    aggregation_method: str
+    derived: bool = True
+    is_estimate: bool = True
+
+
+class ThematicRasterSelectionSummary(BaseModel):
+    metric_label: str
+    metric_value: float
+    metric_unit: str
+    aggregation_method: str
+    primary_metric_key: str
+    metrics: list[ThematicRasterMetric]
+
+
+class ThematicRasterSelectionResponse(BaseModel):
+    dataset_id: UUID
+    product_key: str
+    theme: str
+    metric_kind: str
+    selection_bbox: VectorSelectionBBox
+    selection_area_id: UUID | None = None
+    selected_cell_count: int
+    valid_cell_count: int
+    coverage_ratio: float
+    resolution_m: float
+    observation_year: int
+    summary: ThematicRasterSelectionSummary
+    unsupported_metrics: list[str]
+    limitation_message: str
+    generated_at: str

backend/app/services/geo_assistant_service.py

@@ -22,8 +22,11 @@ from app.schemas.assistant import (
     AssistantTemporalSeries,
 )
 from app.schemas.flood_hazard import FloodHazardSelectionRequest
+from app.schemas.thematic_raster import ThematicRasterSelectionRequest
 from app.services.flood_hazard_acquisition_service import FloodHazardAcquisitionService
 from app.services.flood_hazard_analysis_service import FloodHazardAnalysisService
+from app.services.thematic_raster_acquisition_service import ThematicRasterAcquisitionService
+from app.services.thematic_raster_analysis_service import ThematicRasterAnalysisService
 from app.services.vector_feature_service import VectorFeatureService
 
 
@@ -42,9 +45,17 @@ class GeoAssistantService:
     )
     ESTIMATE_TOPIC_TERMS = {
         "population": ("bevolk", "inwoner"),
+        "space_occupation": ("ruimtebeslag",),
+        "open_space": ("open ruimte",),
+        "accessibility": ("bereikbaar", "knooppunt"),
+        "services": ("voorziening",),
     }
     ESTIMATE_TOPIC_LABELS = {
         "population": "bevolkingswaarden",
+        "space_occupation": "ruimtebeslagoppervlakten",
+        "open_space": "openruimte-oppervlakten",
+        "accessibility": "bereikbaarheidsscores",
+        "services": "voorzieningenscores",
     }
 
     @classmethod
@@ -264,6 +275,19 @@ class GeoAssistantService:
             if dataset.dataset_type == "raster" and dataset.source_name == FloodHazardAcquisitionService.PROVIDER
             and (area is None or dataset.area_id is None or dataset.area_id == area.id)
         ]
+        thematic_candidates = [
+            dataset
+            for dataset in datasets
+            if dataset.dataset_type == "raster" and dataset.source_name == ThematicRasterAcquisitionService.PROVIDER
+            and (area is None or dataset.area_id is None or dataset.area_id == area.id)
+        ]
+        thematic_by_product: dict[str, Dataset] = {}
+        for dataset in thematic_candidates:
+            product_key = str((dataset.source_metadata or {}).get("product_key") or "")
+            current = thematic_by_product.get(product_key)
+            if product_key and (current is None or (dataset.imported_at or datetime.min.replace(tzinfo=timezone.utc)) > (current.imported_at or datetime.min.replace(tzinfo=timezone.utc))):
+                thematic_by_product[product_key] = dataset
+        thematic_datasets = list(thematic_by_product.values())
         warnings: list[str] = []
         context_metrics: list[AssistantContextMetric] = []
         source_dataset_ids: list[UUID] = []
@@ -323,6 +347,47 @@ class GeoAssistantService:
                     }
                 )
 
+            for dataset in sorted(thematic_datasets, key=lambda item: str((item.source_metadata or {}).get("product_key") or item.name)):
+                try:
+                    result = ThematicRasterAnalysisService.analyze(
+                        db,
+                        project_id,
+                        dataset.id,
+                        ThematicRasterSelectionRequest(bbox=bbox, area_id=area.id if area is not None else None),
+                        settings=self.settings,
+                    )
+                except AppError as exc:
+                    warnings.append(f"{dataset.name}: {exc.message}")
+                    continue
+                serialized_metrics: list[dict[str, Any]] = []
+                for metric in result["summary"]["metrics"]:
+                    item = AssistantContextMetric(
+                        theme=result["theme"],
+                        label=str(metric["metric_label"]),
+                        value=float(metric["metric_value"]),
+                        unit=str(metric["metric_unit"]),
+                        source=ThematicRasterAcquisitionService.ATTRIBUTION,
+                        dataset_id=dataset.id,
+                        observed_at=dataset.observed_at,
+                        is_estimate=bool(metric.get("is_estimate", True)),
+                    )
+                    context_metrics.append(item)
+                    serialized_metrics.append(item.model_dump(mode="json"))
+                    serialized_metrics[-1]["measurement_quality"] = "resolutiegebonden_bronmeting"
+                source_dataset_ids.append(dataset.id)
+                current_context.append(
+                    {
+                        "dataset_name": dataset.name,
+                        "dataset_id": str(dataset.id),
+                        "theme": result["theme"],
+                        "source": ThematicRasterAcquisitionService.ATTRIBUTION,
+                        "observed_at": dataset.observed_at.isoformat() if dataset.observed_at else None,
+                        "metrics": serialized_metrics,
+                        "unsupported_metrics": result["unsupported_metrics"],
+                        "warning": result["limitation_message"],
+                    }
+                )
+
             for dataset in sorted(
                 flood_hazard_datasets,
                 key=lambda item: str((item.source_metadata or {}).get("product_key") or item.name),
@@ -425,6 +490,7 @@ class GeoAssistantService:
                 "water_volume_available": False,
                 "water_volume_reason": "Geen bathymetrie gekoppeld voor de permanente inhoud van waterlichamen.",
                 "flood_hazard_scenarios_available": bool(flood_hazard_datasets),
+                "thematic_policy_rasters_available": bool(thematic_datasets),
                 "flood_depth_area_integral_is_concurrent_volume": False,
                 "object_counts_are_supporting_metrics": True,
                 "causal_explanations_available": False,

backend/app/services/thematic_raster_acquisition_service.py

@@ -0,0 +1,597 @@
+from __future__ import annotations
+
+import hashlib
+import json
+import math
+import time
+from dataclasses import dataclass
+from datetime import UTC, datetime
+from pathlib import Path
+from typing import Any, Callable
+from urllib.error import HTTPError, URLError
+from urllib.parse import urlencode
+from urllib.request import Request, urlopen
+from uuid import UUID
+
+from geoalchemy2.shape import to_shape
+from pyproj import Transformer
+from shapely.geometry import box, mapping
+from shapely.ops import transform as shapely_transform
+
+from app.core.config import Settings, get_settings
+from app.core.errors import AppError
+from app.models import Area, Dataset, Project
+from app.schemas.thematic_raster import (
+    ThematicRasterAcquireRequest,
+    ThematicRasterAcquisitionResult,
+    ThematicRasterProductRead,
+)
+from app.services.dataset_service import DatasetService
+
+
+@dataclass(frozen=True)
+class ThematicRasterProduct:
+    key: str
+    display_name: str
+    theme: str
+    metric_kind: str
+    coverage_id: str
+    native_resolution_m: float
+    source_value_unit: str
+    observation_year: int
+    source_version: str
+    catalog_url: str
+    legend_min_label: str
+    legend_max_label: str
+    limitation_message: str
+
+
+class ThematicRasterAcquisitionService:
+    """Acquire bounded, allowlisted policy rasters from MercatorNet WCS."""
+
+    PROVIDER = "department_omgeving_thematic_raster"
+    SOURCE_CRS = "EPSG:31370"
+    WCS_VERSION = "1.0.0"
+    NODATA = -9999.0
+    WCS_TILE_SIDE_M = 10_000.0
+    WCS_REQUEST_INTERVAL_SECONDS = 0.5
+    ATTRIBUTION = "Bron: Departement Omgeving, MercatorNet"
+    LICENSE_NOTE = "Publieke GDI-Vlaanderen bron; bronvermelding en productspecifieke gebruiksvoorwaarden blijven van toepassing."
+
+    @staticmethod
+    def _products() -> dict[str, ThematicRasterProduct]:
+        products = (
+            ThematicRasterProduct(
+                key="space_occupation_2025",
+                display_name="Ruimtebeslag Vlaanderen 2025",
+                theme="space_occupation",
+                metric_kind="binary_area",
+                coverage_id="lu:lu_ruibes_vlaa_2025_v3",
+                native_resolution_m=10.0,
+                source_value_unit="class_0_1",
+                observation_year=2025,
+                source_version="Toestand 2025 versie 3",
+                catalog_url="https://www.vlaanderen.be/datavindplaats/catalogus/ruimtebeslag-vlaanderen-toestand-2025",
+                legend_min_label="Geen ruimtebeslag",
+                legend_max_label="Ruimtebeslag",
+                limitation_message=(
+                    "Binaire 10 m-kaart volgens de beleidsdefinitie van ruimtebeslag. Celgebaseerde oppervlakte is een "
+                    "resolutiegebonden schatting en is niet gelijk aan uitsluitend bebouwde oppervlakte of verharding."
+                ),
+            ),
+            ThematicRasterProduct(
+                key="open_space_2022",
+                display_name="Open ruimte Vlaanderen 2022",
+                theme="open_space",
+                metric_kind="binary_area",
+                coverage_id="lu:lu_openruimte_vlaa_2022_v3",
+                native_resolution_m=10.0,
+                source_value_unit="class_0_1",
+                observation_year=2022,
+                source_version="Toestand 2022 versie 3",
+                catalog_url="https://www.vlaanderen.be/datavindplaats/catalogus/open-ruimte-vlaanderen-toestand-2022",
+                legend_min_label="Geen open ruimte",
+                legend_max_label="Open ruimte",
+                limitation_message=(
+                    "Binaire 10 m-beleidskaart afgeleid uit landgebruik, ruimtebeslag en kernen. Open ruimte is niet "
+                    "synoniem met natuur, bos, publieke toegankelijkheid of planologische bestemming."
+                ),
+            ),
+            ThematicRasterProduct(
+                key="population_density_2019",
+                display_name="Inwonersdichtheid per hectare 2019",
+                theme="population",
+                metric_kind="population_density",
+                coverage_id="ni:ni_inw_ha_vlaa_2019",
+                native_resolution_m=100.0,
+                source_value_unit="inhabitants_per_hectare",
+                observation_year=2019,
+                source_version="Toestand 2019",
+                catalog_url="https://www.vlaanderen.be/datavindplaats/catalogus/inwonersdichtheid-per-ha-vlaanderen-toestand-2019",
+                legend_min_label="0 inwoners/ha",
+                legend_max_label="Hogere dichtheid",
+                limitation_message=(
+                    "Statistische 1 ha-rasterinschatting voor 2019, gecorrigeerd op statistische-sectorbasis. De som "
+                    "binnen een getekende grens is een rasterraming en geen actuele registertelling."
+                ),
+            ),
+            ThematicRasterProduct(
+                key="node_value_2022",
+                display_name="Knooppuntwaarde collectief vervoer 2022",
+                theme="accessibility",
+                metric_kind="index_score",
+                coverage_id="lu:lu_knptw_ha_2022_v3",
+                native_resolution_m=100.0,
+                source_value_unit="source_index_score",
+                observation_year=2022,
+                source_version="Toestand 2022 versie 3",
+                catalog_url="https://www.vlaanderen.be/datavindplaats/catalogus/knooppuntwaarde-per-ha-toestand-2022",
+                legend_min_label="Lagere knooppuntwaarde",
+                legend_max_label="Hogere knooppuntwaarde",
+                limitation_message=(
+                    "Bronindex per hectare op basis van collectief-vervoerknooppunten en afstandsverval. De score is "
+                    "geen percentage, reistijd, dienstregeling van vandaag of garantie op bereikbaarheid."
+                ),
+            ),
+            ThematicRasterProduct(
+                key="service_level_2022",
+                display_name="Totaal voorzieningenniveau 2022",
+                theme="services",
+                metric_kind="normalized_score",
+                coverage_id="lu:lu_totvznv_ha_2022_v3",
+                native_resolution_m=100.0,
+                source_value_unit="score_0_1",
+                observation_year=2022,
+                source_version="Toestand 2022 versie 3",
+                catalog_url="https://www.vlaanderen.be/datavindplaats/catalogus/totaal-voorzieningenniveau-toestand-2022",
+                legend_min_label="Lager voorzieningenniveau",
+                legend_max_label="Hoger voorzieningenniveau",
+                limitation_message=(
+                    "Genormaliseerde 0-1 nabijheidsscore voor basis-, regionale en metropolitane voorzieningen in "
+                    "referentiejaar 2022. Dit is geen objecttelling, openingsurencontrole of actuele reistijd."
+                ),
+            ),
+        )
+        return {product.key: product for product in products}
+
+    @staticmethod
+    def list_products() -> list[dict[str, Any]]:
+        return [
+            ThematicRasterProductRead(
+                key=product.key,
+                display_name=product.display_name,
+                theme=product.theme,
+                metric_kind=product.metric_kind,
+                coverage_id=product.coverage_id,
+                native_resolution_m=product.native_resolution_m,
+                source_crs=ThematicRasterAcquisitionService.SOURCE_CRS,
+                source_value_unit=product.source_value_unit,
+                observation_year=product.observation_year,
+                source_version=product.source_version,
+                catalog_url=product.catalog_url,
+                attribution=ThematicRasterAcquisitionService.ATTRIBUTION,
+                license_note=ThematicRasterAcquisitionService.LICENSE_NOTE,
+                legend_min_label=product.legend_min_label,
+                legend_max_label=product.legend_max_label,
+                limitation_message=product.limitation_message,
+            ).model_dump()
+            for product in ThematicRasterAcquisitionService._products().values()
+        ]
+
+    @staticmethod
+    def _product(product_key: str) -> ThematicRasterProduct:
+        product = ThematicRasterAcquisitionService._products().get(product_key.strip().lower())
+        if product is None:
+            raise AppError(
+                code="THEMATIC_RASTER_PRODUCT_NOT_SUPPORTED",
+                message="Select a product from the governed Flemish thematic raster registry",
+                details={"product_key": product_key},
+                status_code=422,
+            )
+        return product
+
+    @staticmethod
+    def _prepared_request(payload: ThematicRasterAcquireRequest, settings: Settings) -> dict[str, Any]:
+        if not settings.thematic_raster_enabled:
+            raise AppError(code="THEMATIC_RASTER_NOT_CONFIGURED", message="Official thematic raster acquisition is disabled", status_code=503)
+        product = ThematicRasterAcquisitionService._product(payload.product_key)
+        values = (payload.bbox.min_x, payload.bbox.min_y, payload.bbox.max_x, payload.bbox.max_y)
+        if payload.bbox.crs.upper() != "EPSG:4326":
+            raise AppError(code="INVALID_BBOX_CRS", message="Thematic raster selection requires EPSG:4326", status_code=400)
+        if not all(math.isfinite(value) for value in values) or values[0] >= values[2] or values[1] >= values[3]:
+            raise AppError(code="INVALID_BBOX", message="Thematic raster selection must be a finite non-empty rectangle", status_code=400)
+
+        transformer = Transformer.from_crs("EPSG:4326", ThematicRasterAcquisitionService.SOURCE_CRS, always_xy=True)
+        raw_bounds = transformer.transform_bounds(*values, densify_pts=21)
+        resolution = product.native_resolution_m
+        lambert_bounds = (
+            math.floor(raw_bounds[0] / resolution) * resolution,
+            math.floor(raw_bounds[1] / resolution) * resolution,
+            math.ceil(raw_bounds[2] / resolution) * resolution,
+            math.ceil(raw_bounds[3] / resolution) * resolution,
+        )
+        width_m = lambert_bounds[2] - lambert_bounds[0]
+        height_m = lambert_bounds[3] - lambert_bounds[1]
+        if width_m < settings.thematic_raster_min_side_m or height_m < settings.thematic_raster_min_side_m:
+            raise AppError(
+                code="THEMATIC_RASTER_SELECTION_TOO_SMALL",
+                message=f"Select an area of at least {settings.thematic_raster_min_side_m:g} by {settings.thematic_raster_min_side_m:g} metres",
+                status_code=422,
+            )
+        if width_m > settings.thematic_raster_max_side_m or height_m > settings.thematic_raster_max_side_m:
+            raise AppError(
+                code="THEMATIC_RASTER_SELECTION_TOO_LARGE",
+                message=f"Select an area no larger than {settings.thematic_raster_max_side_m:g} by {settings.thematic_raster_max_side_m:g} metres",
+                details={"width_m": width_m, "height_m": height_m},
+                status_code=422,
+            )
+        width = max(1, round(width_m / resolution))
+        height = max(1, round(height_m / resolution))
+        if width * height > settings.thematic_raster_max_pixels:
+            raise AppError(
+                code="THEMATIC_RASTER_SELECTION_TOO_LARGE",
+                message="Thematic raster selection exceeds the configured cell limit",
+                details={"pixel_count": width * height, "max_pixels": settings.thematic_raster_max_pixels},
+                status_code=422,
+            )
+        request_identity = {
+            "provider": ThematicRasterAcquisitionService.PROVIDER,
+            "coverage_id": product.coverage_id,
+            "bbox_epsg4326": [round(float(value), 8) for value in values],
+            "bbox_epsg31370": [round(float(value), 3) for value in lambert_bounds],
+            "resolution_m": resolution,
+            "area_id": str(payload.area_id) if payload.area_id else None,
+        }
+        request_hash = hashlib.sha256(json.dumps(request_identity, sort_keys=True).encode()).hexdigest()
+        return {
+            **request_identity,
+            "product": product,
+            "request_hash": request_hash,
+            "width": width,
+            "height": height,
+        }
+
+    @staticmethod
+    def _wcs_request_url(settings: Settings, product: ThematicRasterProduct, bounds: tuple[float, float, float, float]) -> str:
+        query = {
+            "SERVICE": "WCS",
+            "VERSION": ThematicRasterAcquisitionService.WCS_VERSION,
+            "REQUEST": "GetCoverage",
+            "COVERAGE": product.coverage_id,
+            "CRS": ThematicRasterAcquisitionService.SOURCE_CRS,
+            "BBOX": ",".join(f"{value:.3f}" for value in bounds),
+            "RESX": f"{product.native_resolution_m:g}",
+            "RESY": f"{product.native_resolution_m:g}",
+            "FORMAT": "image/tiff",
+            "RESPONSE_CRS": ThematicRasterAcquisitionService.SOURCE_CRS,
+        }
+        return f"{settings.thematic_raster_wcs_url}?{urlencode(query)}"
+
+    @staticmethod
+    def _tile_bounds(prepared: dict[str, Any]) -> list[tuple[float, float, float, float]]:
+        min_x, min_y, max_x, max_y = prepared["bbox_epsg31370"]
+        resolution = prepared["product"].native_resolution_m
+        side = max(resolution, math.floor(ThematicRasterAcquisitionService.WCS_TILE_SIDE_M / resolution) * resolution)
+        tiles: list[tuple[float, float, float, float]] = []
+        y = min_y
+        while y < max_y:
+            tile_max_y = min(y + side, max_y)
+            x = min_x
+            while x < max_x:
+                tile_max_x = min(x + side, max_x)
+                tiles.append((x, y, tile_max_x, tile_max_y))
+                x = tile_max_x
+            y = tile_max_y
+        return tiles
+
+    @staticmethod
+    def _scope_geometry(db, project_id: UUID, area_id: UUID | None, bbox_epsg4326: list[float]):
+        if not db.get(Project, project_id):
+            raise AppError(code="PROJECT_NOT_FOUND", message="Project not found", status_code=404)
+        selection = box(*bbox_epsg4326)
+        if area_id is None:
+            return selection
+        area = db.get(Area, area_id)
+        if not area:
+            raise AppError(code="AREA_NOT_FOUND", message="Area not found", status_code=404)
+        if area.project_id != project_id:
+            raise AppError(code="INVALID_DATASET_SCOPE", message="Area does not belong to this project", status_code=400)
+        intersection = to_shape(area.geometry).intersection(selection)
+        if intersection.is_empty or intersection.area <= 0:
+            raise AppError(code="THEMATIC_RASTER_SELECTION_OUTSIDE_AREA", message="Selection does not overlap the selected work area", status_code=422)
+        return intersection
+
+    @staticmethod
+    def _fetch(request_url: str, settings: Settings, opener: Callable[..., Any] | None = None) -> tuple[bytes, str]:
+        request = Request(request_url, headers={"Accept": "image/tiff,*/*", "User-Agent": "GeoIntel/0.1 bounded-thematic-raster"})
+        max_bytes = settings.thematic_raster_max_response_mb * 1024 * 1024
+        try:
+            with (opener or urlopen)(request, timeout=settings.thematic_raster_timeout_seconds) as response:
+                content_type = str(response.headers.get("Content-Type", ""))
+                content_length = response.headers.get("Content-Length")
+                if content_length and int(content_length) > max_bytes:
+                    raise AppError(code="THEMATIC_RASTER_RESPONSE_TOO_LARGE", message="Official raster response exceeds the configured size limit", status_code=502)
+                content = response.read(max_bytes + 1)
+        except AppError:
+            raise
+        except HTTPError as exc:
+            preview = exc.read(300).decode("utf-8", errors="replace")
+            raise AppError(
+                code="THEMATIC_RASTER_PROVIDER_UNAVAILABLE",
+                message="The official MercatorNet WCS could not complete the bounded request",
+                details={"reason": str(exc), "provider_status_code": int(exc.code), "response_preview": preview},
+                status_code=502,
+            ) from exc
+        except (URLError, TimeoutError, OSError) as exc:
+            raise AppError(
+                code="THEMATIC_RASTER_PROVIDER_UNAVAILABLE",
+                message="The official MercatorNet WCS could not complete the bounded request",
+                details={"reason": str(exc)},
+                status_code=502,
+            ) from exc
+        if len(content) > max_bytes:
+            raise AppError(code="THEMATIC_RASTER_RESPONSE_TOO_LARGE", message="Official raster response exceeds the configured size limit", status_code=502)
+        if not content.startswith((b"II*\x00", b"MM\x00*")):
+            preview = content[:300].decode("utf-8", errors="replace")
+            raise AppError(
+                code="THEMATIC_RASTER_PROVIDER_INVALID_RESPONSE",
+                message="The official MercatorNet service did not return a GeoTIFF coverage",
+                details={"content_type": content_type, "response_preview": preview},
+                status_code=502,
+            )
+        return content, content_type
+
+    @staticmethod
+    def _mosaic(coverages: list[bytes], product: ThematicRasterProduct) -> bytes:
+        if len(coverages) == 1:
+            return coverages[0]
+        try:
+            import rasterio
+            from rasterio.io import MemoryFile
+            from rasterio.merge import merge
+        except ImportError as exc:
+            raise AppError(code="RASTER_PROCESSING_UNAVAILABLE", message="Rasterio is required to assemble thematic raster tiles", status_code=503) from exc
+        memories = [MemoryFile(content) for content in coverages]
+        sources = []
+        try:
+            sources = [memory.open() for memory in memories]
+            for source in sources:
+                if source.crs is None or source.crs.to_epsg() != 31370 or source.count != 1:
+                    raise AppError(code="THEMATIC_RASTER_TILE_MISMATCH", message="Thematic raster tiles have incompatible CRS or bands", status_code=502)
+                if not all(math.isclose(abs(float(value)), product.native_resolution_m, abs_tol=0.05) for value in source.res):
+                    raise AppError(code="THEMATIC_RASTER_TILE_MISMATCH", message="Thematic raster tile resolution differs from the registry", status_code=502)
+            mosaic, transform = merge(sources, res=(product.native_resolution_m, product.native_resolution_m), nodata=ThematicRasterAcquisitionService.NODATA, dtype="float32")
+            profile = sources[0].profile.copy()
+            profile.pop("blockxsize", None)
+            profile.pop("blockysize", None)
+            profile.update(driver="GTiff", width=mosaic.shape[2], height=mosaic.shape[1], count=1, dtype="float32", crs=ThematicRasterAcquisitionService.SOURCE_CRS, transform=transform, nodata=ThematicRasterAcquisitionService.NODATA, compress="deflate", predictor=3)
+            with MemoryFile() as output_memory:
+                with output_memory.open(**profile) as output:
+                    output.write(mosaic)
+                return output_memory.read()
+        except AppError:
+            raise
+        except Exception as exc:
+            raise AppError(code="THEMATIC_RASTER_TILE_MOSAIC_FAILED", message="Thematic raster tiles could not be assembled", details={"reason": str(exc)}, status_code=502) from exc
+        finally:
+            for source in sources:
+                source.close()
+            for memory in memories:
+                memory.close()
+
+    @staticmethod
+    def _fetch_coverage(prepared: dict[str, Any], settings: Settings, opener: Callable[..., Any] | None = None) -> tuple[bytes, dict[str, Any]]:
+        product: ThematicRasterProduct = prepared["product"]
+        request_urls = [ThematicRasterAcquisitionService._wcs_request_url(settings, product, bounds) for bounds in ThematicRasterAcquisitionService._tile_bounds(prepared)]
+        coverages: list[bytes] = []
+        digest = hashlib.sha256()
+        content_types: list[str] = []
+        for index, request_url in enumerate(request_urls):
+            if index and opener is None:
+                time.sleep(ThematicRasterAcquisitionService.WCS_REQUEST_INTERVAL_SECONDS)
+            content, content_type = ThematicRasterAcquisitionService._fetch(request_url, settings, opener)
+            digest.update(len(content).to_bytes(8, "big"))
+            digest.update(content)
+            coverages.append(content)
+            content_types.append(content_type)
+        return ThematicRasterAcquisitionService._mosaic(coverages, product), {
+            "tile_count": len(request_urls),
+            "request_urls": request_urls,
+            "response_content_types": content_types,
+            "coverage_sha256": digest.hexdigest(),
+        }
+
+    @staticmethod
+    def _validate_values(values, product: ThematicRasterProduct) -> None:
+        import numpy as np
+
+        if values.size == 0:
+            raise AppError(code="THEMATIC_RASTER_NO_VALID_DATA", message="The official product contains no valid cells in this selection", status_code=422)
+        minimum = float(values.min())
+        maximum = float(values.max())
+        if minimum < 0:
+            raise AppError(code="THEMATIC_RASTER_INVALID_VALUES", message="Official thematic raster contains unexpected negative values", details={"minimum": minimum}, status_code=502)
+        if product.metric_kind == "binary_area" and not set(np.unique(values).tolist()).issubset({0.0, 1.0}):
+            raise AppError(code="THEMATIC_RASTER_INVALID_VALUES", message="Binary thematic raster contains classes outside 0 and 1", status_code=502)
+        if product.metric_kind == "normalized_score" and maximum > 1.0001:
+            raise AppError(code="THEMATIC_RASTER_INVALID_VALUES", message="Normalized thematic score falls outside the documented 0-1 range", details={"maximum": maximum}, status_code=502)
+
+    @staticmethod
+    def _normalize_raster(content: bytes, scope_geometry_4326, prepared: dict[str, Any]) -> tuple[bytes, dict[str, Any]]:
+        try:
+            import numpy as np
+            from rasterio.io import MemoryFile
+            from rasterio.mask import mask
+        except ImportError as exc:
+            raise AppError(code="RASTER_PROCESSING_UNAVAILABLE", message="Rasterio and numpy are required for thematic raster validation", status_code=503) from exc
+        product: ThematicRasterProduct = prepared["product"]
+        try:
+            with MemoryFile(content) as source_memory, source_memory.open() as source:
+                if source.crs is None or source.crs.to_epsg() != 31370:
+                    raise AppError(code="THEMATIC_RASTER_INVALID_CRS", message="Official thematic raster must use EPSG:31370", status_code=502)
+                if source.count != 1:
+                    raise AppError(code="THEMATIC_RASTER_INVALID_BANDS", message="Official thematic raster must contain one band", status_code=502)
+                if not all(math.isclose(abs(float(value)), product.native_resolution_m, abs_tol=0.05) for value in source.res):
+                    raise AppError(code="THEMATIC_RASTER_INVALID_RESOLUTION", message="Official thematic raster resolution differs from the registry", status_code=502)
+                transformer = Transformer.from_crs("EPSG:4326", ThematicRasterAcquisitionService.SOURCE_CRS, always_xy=True)
+                scope_metric = shapely_transform(transformer.transform, scope_geometry_4326)
+                clipped, transform = mask(source, [mapping(scope_metric)], crop=True, filled=False, indexes=[1])
+                band = np.ma.asarray(clipped[0], dtype="float32")
+                raw = np.asarray(band.filled(np.nan), dtype="float32")
+                invalid = np.ma.getmaskarray(band) | ~np.isfinite(raw)
+                if source.nodata is not None:
+                    invalid |= np.isclose(raw, float(source.nodata))
+                normalized = np.ma.array(raw, mask=invalid)
+                values = normalized.compressed().astype("float64")
+                ThematicRasterAcquisitionService._validate_values(values, product)
+                profile = source.profile.copy()
+                profile.pop("blockxsize", None)
+                profile.pop("blockysize", None)
+                profile.update(driver="GTiff", width=normalized.shape[1], height=normalized.shape[0], count=1, dtype="float32", crs=ThematicRasterAcquisitionService.SOURCE_CRS, transform=transform, nodata=ThematicRasterAcquisitionService.NODATA, compress="deflate", predictor=3)
+                with MemoryFile() as output_memory:
+                    with output_memory.open(**profile) as output:
+                        output.write(normalized.filled(ThematicRasterAcquisitionService.NODATA), 1)
+                    normalized_content = output_memory.read()
+                return normalized_content, {
+                    "width": int(normalized.shape[1]),
+                    "height": int(normalized.shape[0]),
+                    "valid_pixel_count": int(values.size),
+                    "nodata_value": ThematicRasterAcquisitionService.NODATA,
+                    "resolution_m": product.native_resolution_m,
+                    "minimum_value": float(values.min()),
+                    "maximum_value": float(values.max()),
+                    "p02_value": float(np.percentile(values, 2)),
+                    "p98_value": float(np.percentile(values, 98)),
+                }
+        except AppError:
+            raise
+        except Exception as exc:
+            raise AppError(code="THEMATIC_RASTER_INVALID", message="The official thematic raster could not be validated", details={"reason": str(exc)}, status_code=502) from exc
+
+    @staticmethod
+    def _cached_dataset(db, project_id: UUID, filename: str) -> Dataset | None:
+        candidate = (
+            db.query(Dataset)
+            .filter(Dataset.project_id == project_id, Dataset.name == filename, Dataset.source_name == ThematicRasterAcquisitionService.PROVIDER, Dataset.status == "ready")
+            .order_by(Dataset.imported_at.desc())
+            .first()
+        )
+        return candidate if candidate and candidate.storage_path and Path(candidate.storage_path).is_file() else None
+
+    @staticmethod
+    def acquire(db, project_id: UUID, payload: ThematicRasterAcquireRequest, *, settings: Settings | None = None, opener: Callable[..., Any] | None = None) -> dict[str, Any]:
+        resolved_settings = settings or get_settings()
+        prepared = ThematicRasterAcquisitionService._prepared_request(payload, resolved_settings)
+        product: ThematicRasterProduct = prepared["product"]
+        scope_geometry = ThematicRasterAcquisitionService._scope_geometry(db, project_id, payload.area_id, prepared["bbox_epsg4326"])
+        filename = f"thematic_{product.key}_{prepared['request_hash'][:12]}.tif"
+        if not payload.force_refresh:
+            cached = ThematicRasterAcquisitionService._cached_dataset(db, project_id, filename)
+            if cached is not None:
+                metadata = cached.source_metadata or {}
+                raster_metadata = cached.metadata_json or {}
+                return ThematicRasterAcquisitionResult(
+                    output_dataset_id=cached.id,
+                    reused=True,
+                    provider=ThematicRasterAcquisitionService.PROVIDER,
+                    product_key=product.key,
+                    display_name=product.display_name,
+                    theme=product.theme,
+                    metric_kind=product.metric_kind,
+                    coverage_id=product.coverage_id,
+                    resolution_m=product.native_resolution_m,
+                    width=int(raster_metadata.get("width", prepared["width"])),
+                    height=int(raster_metadata.get("height", prepared["height"])),
+                    valid_pixel_count=int(metadata.get("valid_pixel_count", 0)),
+                    bbox_epsg4326=prepared["bbox_epsg4326"],
+                    bbox_epsg31370=prepared["bbox_epsg31370"],
+                    observation_year=product.observation_year,
+                    source_value_unit=product.source_value_unit,
+                    attribution=ThematicRasterAcquisitionService.ATTRIBUTION,
+                    limitation_message=product.limitation_message,
+                ).model_dump(mode="json")
+
+        coverage, transfer = ThematicRasterAcquisitionService._fetch_coverage(prepared, resolved_settings, opener)
+        normalized, validation = ThematicRasterAcquisitionService._normalize_raster(coverage, scope_geometry, prepared)
+        acquired_at = datetime.now(UTC)
+        observed_at = datetime(product.observation_year, 12, 31, 23, 59, 59, tzinfo=UTC)
+        scope_key = str(payload.area_id) if payload.area_id else prepared["request_hash"][:24]
+        dataset = DatasetService.import_raster_bytes(
+            db,
+            project_id=project_id,
+            area_id=payload.area_id,
+            filename=filename,
+            content=normalized,
+            source=f"Departement Omgeving MercatorNet WCS {product.coverage_id}",
+            source_name=ThematicRasterAcquisitionService.PROVIDER,
+            temporal_series_key=f"department-omgeving:thematic-raster:{product.key}:{scope_key}",
+            observed_at=observed_at,
+            valid_from=datetime(product.observation_year, 1, 1, tzinfo=UTC),
+            valid_to=observed_at,
+            temporal_granularity="year",
+            source_version=product.source_version,
+            source_metadata={
+                "provider": ThematicRasterAcquisitionService.PROVIDER,
+                "service": "WCS",
+                "service_version": ThematicRasterAcquisitionService.WCS_VERSION,
+                "product_key": product.key,
+                "product_display_name": product.display_name,
+                "theme": product.theme,
+                "metric_kind": product.metric_kind,
+                "coverage_id": product.coverage_id,
+                "native_resolution_m": product.native_resolution_m,
+                "analysis_resolution_m": product.native_resolution_m,
+                "source_crs": ThematicRasterAcquisitionService.SOURCE_CRS,
+                "source_value_unit": product.source_value_unit,
+                "observation_year": product.observation_year,
+                "observation_date_precision": "year",
+                "valid_pixel_count": validation["valid_pixel_count"],
+                "minimum_value": validation["minimum_value"],
+                "maximum_value": validation["maximum_value"],
+                "render_min_value": validation["p02_value"],
+                "render_max_value": validation["p98_value"],
+                "bbox_epsg4326": prepared["bbox_epsg4326"],
+                "bbox_epsg31370": prepared["bbox_epsg31370"],
+                "catalog_url": product.catalog_url,
+                "attribution": ThematicRasterAcquisitionService.ATTRIBUTION,
+                "license_note": ThematicRasterAcquisitionService.LICENSE_NOTE,
+                "legend_min_label": product.legend_min_label,
+                "legend_max_label": product.legend_max_label,
+                "coverage_scope": "municipality" if payload.area_id else "bounded_selection",
+            },
+            provenance_metadata={
+                "acquisition": "explicit_bounded_tiled_wcs_coverage",
+                "acquired_at": acquired_at.isoformat(),
+                "request_hash": prepared["request_hash"],
+                "tile_count": transfer["tile_count"],
+                "tile_request_urls": transfer["request_urls"],
+                "response_content_types": transfer["response_content_types"],
+                "coverage_sha256": transfer["coverage_sha256"],
+                "normalized_sha256": hashlib.sha256(normalized).hexdigest(),
+                "bbox_epsg4326": prepared["bbox_epsg4326"],
+                "bbox_epsg31370": prepared["bbox_epsg31370"],
+                "clipped_to_area_id": str(payload.area_id) if payload.area_id else None,
+                "validation": validation,
+                "limitation_message": product.limitation_message,
+            },
+        )
+        return ThematicRasterAcquisitionResult(
+            output_dataset_id=dataset.id,
+            reused=False,
+            provider=ThematicRasterAcquisitionService.PROVIDER,
+            product_key=product.key,
+            display_name=product.display_name,
+            theme=product.theme,
+            metric_kind=product.metric_kind,
+            coverage_id=product.coverage_id,
+            resolution_m=product.native_resolution_m,
+            width=validation["width"],
+            height=validation["height"],
+            valid_pixel_count=validation["valid_pixel_count"],
+            bbox_epsg4326=prepared["bbox_epsg4326"],
+            bbox_epsg31370=prepared["bbox_epsg31370"],
+            observation_year=product.observation_year,
+            source_value_unit=product.source_value_unit,
+            attribution=ThematicRasterAcquisitionService.ATTRIBUTION,
+            limitation_message=product.limitation_message,
+        ).model_dump(mode="json")

backend/app/services/thematic_raster_analysis_service.py

@@ -0,0 +1,260 @@
+from __future__ import annotations
+
+import io
+import math
+from datetime import UTC, datetime
+from pathlib import Path
+from uuid import UUID
+
+from geoalchemy2.shape import to_shape
+from pyproj import Transformer
+from shapely.geometry import box, mapping
+from shapely.ops import transform as shapely_transform
+
+from app.core.config import Settings, get_settings
+from app.core.errors import AppError
+from app.models import Area, Dataset
+from app.schemas.thematic_raster import (
+    ThematicRasterMetric,
+    ThematicRasterSelectionRequest,
+    ThematicRasterSelectionResponse,
+    ThematicRasterSelectionSummary,
+)
+from app.services.thematic_raster_acquisition_service import (
+    ThematicRasterAcquisitionService,
+    ThematicRasterProduct,
+)
+
+
+class ThematicRasterAnalysisService:
+    @staticmethod
+    def _load_dataset(db, project_id: UUID, dataset_id: UUID) -> Dataset:
+        dataset = db.get(Dataset, dataset_id)
+        if not dataset or dataset.project_id != project_id:
+            raise AppError(code="DATASET_NOT_FOUND", message="Dataset not found", status_code=404)
+        if dataset.dataset_type != "raster" or dataset.source_name != ThematicRasterAcquisitionService.PROVIDER:
+            raise AppError(
+                code="INVALID_THEMATIC_RASTER_DATASET",
+                message="Thematic analysis requires a governed Departement Omgeving raster dataset",
+                status_code=400,
+            )
+        if dataset.status != "ready" or not dataset.storage_path or not Path(dataset.storage_path).is_file():
+            raise AppError(code="DATASET_FILE_MISSING", message="Persisted thematic raster file is unavailable", status_code=404)
+        return dataset
+
+    @staticmethod
+    def _product(dataset: Dataset) -> ThematicRasterProduct:
+        source_metadata = dataset.source_metadata or {}
+        product = ThematicRasterAcquisitionService._products().get(str(source_metadata.get("product_key") or ""))
+        if product is None or source_metadata.get("coverage_id") != product.coverage_id:
+            raise AppError(code="INVALID_THEMATIC_RASTER_METADATA", message="Thematic raster provenance is incomplete", status_code=409)
+        return product
+
+    @staticmethod
+    def _selection_geometry(db, project_id: UUID, payload: ThematicRasterSelectionRequest):
+        selection = box(payload.bbox.min_x, payload.bbox.min_y, payload.bbox.max_x, payload.bbox.max_y)
+        if payload.area_id is None:
+            return selection
+        area = db.get(Area, payload.area_id)
+        if not area:
+            raise AppError(code="AREA_NOT_FOUND", message="Area not found", status_code=404)
+        if area.project_id != project_id:
+            raise AppError(code="INVALID_DATASET_SCOPE", message="Area does not belong to this project", status_code=400)
+        selection = selection.intersection(to_shape(area.geometry))
+        if selection.is_empty or selection.area <= 0:
+            raise AppError(code="THEMATIC_RASTER_SELECTION_OUTSIDE_AREA", message="Selection does not overlap the selected work area", status_code=422)
+        return selection
+
+    @staticmethod
+    def _unsupported_metrics(product: ThematicRasterProduct) -> list[str]:
+        if product.metric_kind == "binary_area":
+            return ["object_count", "parcel_area", "current_land_use"]
+        if product.metric_kind == "population_density":
+            return ["current_population", "household_count", "address_level_population"]
+        if product.metric_kind == "index_score":
+            return ["travel_time_minutes", "current_timetable", "stop_count"]
+        return ["facility_count", "opening_hours", "current_service_availability"]
+
+    @staticmethod
+    def analyze(
+        db,
+        project_id: UUID,
+        dataset_id: UUID,
+        payload: ThematicRasterSelectionRequest,
+        *,
+        settings: Settings | None = None,
+    ) -> dict:
+        resolved_settings = settings or get_settings()
+        dataset = ThematicRasterAnalysisService._load_dataset(db, project_id, dataset_id)
+        product = ThematicRasterAnalysisService._product(dataset)
+        selection_4326 = ThematicRasterAnalysisService._selection_geometry(db, project_id, payload)
+        try:
+            import numpy as np
+            import rasterio
+            from rasterio.features import geometry_mask
+            from rasterio.mask import mask
+        except ImportError as exc:
+            raise AppError(code="RASTER_PROCESSING_UNAVAILABLE", message="Rasterio and numpy are required for thematic raster analysis", status_code=503) from exc
+
+        try:
+            with rasterio.open(dataset.storage_path) as source:
+                if source.crs is None or source.crs.to_epsg() != 31370:
+                    raise AppError(code="INVALID_DATASET_CRS", message="Thematic raster CRS must be EPSG:31370", status_code=409)
+                transformer = Transformer.from_crs("EPSG:4326", source.crs, always_xy=True)
+                selection_metric = shapely_transform(transformer.transform, selection_4326)
+                analysis_geometry = selection_metric.intersection(box(*source.bounds))
+                if analysis_geometry.is_empty or analysis_geometry.area <= 0:
+                    raise AppError(code="THEMATIC_RASTER_SELECTION_OUTSIDE_DATASET", message="Selection does not overlap the persisted thematic raster", status_code=422)
+                min_x, min_y, max_x, max_y = analysis_geometry.bounds
+                expected_cells = math.ceil((max_x - min_x) / abs(source.res[0])) * math.ceil((max_y - min_y) / abs(source.res[1]))
+                if expected_cells > resolved_settings.thematic_raster_max_pixels:
+                    raise AppError(
+                        code="THEMATIC_RASTER_SELECTION_TOO_LARGE",
+                        message="Thematic raster analysis exceeds the configured cell limit",
+                        details={"pixel_count": expected_cells, "max_pixels": resolved_settings.thematic_raster_max_pixels},
+                        status_code=422,
+                    )
+                clipped, clipped_transform = mask(source, [mapping(analysis_geometry)], crop=True, filled=False, indexes=[1])
+                band = np.ma.asarray(clipped[0], dtype="float64")
+                raw = band.filled(np.nan)
+                selected = geometry_mask([mapping(analysis_geometry)], out_shape=band.shape, transform=clipped_transform, invert=True)
+                valid = selected & ~np.ma.getmaskarray(band) & np.isfinite(raw)
+                if source.nodata is not None:
+                    valid &= ~np.isclose(raw, float(source.nodata))
+                values = raw[valid]
+                ThematicRasterAcquisitionService._validate_values(values, product)
+                selected_cell_count = int(selected.sum())
+                valid_cell_count = int(values.size)
+                resolution_x = abs(float(source.res[0]))
+                resolution_y = abs(float(source.res[1]))
+                cell_area_m2 = resolution_x * resolution_y
+        except AppError:
+            raise
+        except Exception as exc:
+            raise AppError(
+                code="THEMATIC_RASTER_ANALYSIS_FAILED",
+                message="The persisted thematic raster could not be analysed",
+                details={"reason": str(exc)},
+                status_code=500,
+            ) from exc
+
+        def metric(key: str, label: str, value: float, unit: str, method: str, *, estimate: bool = True) -> ThematicRasterMetric:
+            return ThematicRasterMetric(
+                metric_key=key,
+                metric_label=label,
+                metric_value=round(float(value), 4),
+                metric_unit=unit,
+                aggregation_method=method,
+                is_estimate=estimate,
+            )
+
+        if product.metric_kind == "binary_area":
+            positive_count = int(np.count_nonzero(values >= 0.5))
+            positive_area_ha = positive_count * cell_area_m2 / 10_000.0
+            positive_share = positive_count / max(1, valid_cell_count) * 100.0
+            label = "Ruimtebeslag" if product.theme == "space_occupation" else "Open ruimte"
+            metrics = [
+                metric(f"{product.theme}_area_ha", f"{label} in selectie", positive_area_ha, "ha", "positive_source_cells_times_cell_area"),
+                metric(f"{product.theme}_share_pct", f"Aandeel {label.lower()}", positive_share, "%", "positive_source_cells_divided_by_valid_selected_cells"),
+                metric("valid_raster_area_ha", "Rasteroppervlakte met bronwaarde", valid_cell_count * cell_area_m2 / 10_000.0, "ha", "valid_selected_cells_times_cell_area"),
+            ]
+        elif product.metric_kind == "population_density":
+            estimated_population = float(values.sum() * (cell_area_m2 / 10_000.0))
+            metrics = [
+                metric("estimated_inhabitants", "Geraamd aantal inwoners (2019)", estimated_population, "inwoners", "sum_density_times_selected_cell_area_hectares"),
+                metric("population_density_mean_per_ha", "Gemiddelde inwonersdichtheid", values.mean(), "inwoners/ha", "mean_valid_one_hectare_source_cells"),
+                metric("population_density_p90_per_ha", "90e percentiel inwonersdichtheid", np.percentile(values, 90), "inwoners/ha", "percentile_90_valid_source_cells"),
+            ]
+        else:
+            unit = "score" if product.metric_kind == "index_score" else "score (0-1)"
+            label = "Knooppuntwaarde" if product.metric_kind == "index_score" else "Voorzieningenniveau"
+            metrics = [
+                metric(f"{product.theme}_mean", f"Gemiddelde {label.lower()}", values.mean(), unit, "mean_valid_source_cells"),
+                metric(f"{product.theme}_p10", f"10e percentiel {label.lower()}", np.percentile(values, 10), unit, "percentile_10_valid_source_cells"),
+                metric(f"{product.theme}_median", f"Mediaan {label.lower()}", np.percentile(values, 50), unit, "median_valid_source_cells"),
+                metric(f"{product.theme}_p90", f"90e percentiel {label.lower()}", np.percentile(values, 90), unit, "percentile_90_valid_source_cells"),
+            ]
+
+        primary = metrics[0]
+        response = ThematicRasterSelectionResponse(
+            dataset_id=dataset.id,
+            product_key=product.key,
+            theme=product.theme,
+            metric_kind=product.metric_kind,
+            selection_bbox=payload.bbox,
+            selection_area_id=payload.area_id,
+            selected_cell_count=selected_cell_count,
+            valid_cell_count=valid_cell_count,
+            coverage_ratio=round(valid_cell_count / max(1, selected_cell_count), 6),
+            resolution_m=round(max(resolution_x, resolution_y), 4),
+            observation_year=product.observation_year,
+            summary=ThematicRasterSelectionSummary(
+                metric_label=primary.metric_label,
+                metric_value=primary.metric_value,
+                metric_unit=primary.metric_unit,
+                aggregation_method=primary.aggregation_method,
+                primary_metric_key=primary.metric_key,
+                metrics=metrics,
+            ),
+            unsupported_metrics=ThematicRasterAnalysisService._unsupported_metrics(product),
+            limitation_message=product.limitation_message,
+            generated_at=datetime.now(UTC).isoformat(),
+        )
+        return response.model_dump(mode="json")
+
+    @staticmethod
+    def render_png(db, project_id: UUID, dataset_id: UUID, *, max_dimension: int = 1800) -> bytes:
+        dataset = ThematicRasterAnalysisService._load_dataset(db, project_id, dataset_id)
+        product = ThematicRasterAnalysisService._product(dataset)
+        try:
+            import numpy as np
+            import rasterio
+            from PIL import Image
+            from rasterio.enums import Resampling
+        except ImportError as exc:
+            raise AppError(code="RASTER_PROCESSING_UNAVAILABLE", message="Rasterio, numpy and Pillow are required for thematic raster rendering", status_code=503) from exc
+        palettes = {
+            "space_occupation": np.asarray([[251, 231, 211], [190, 62, 51]], dtype="float64"),
+            "open_space": np.asarray([[221, 238, 219], [38, 122, 70]], dtype="float64"),
+            "population": np.asarray([[238, 231, 246], [103, 58, 151]], dtype="float64"),
+            "accessibility": np.asarray([[233, 241, 244], [15, 118, 110]], dtype="float64"),
+            "services": np.asarray([[255, 244, 191], [182, 109, 22]], dtype="float64"),
+        }
+        try:
+            with rasterio.open(dataset.storage_path) as source:
+                scale = min(1.0, max_dimension / max(source.width, source.height))
+                width = max(1, round(source.width * scale))
+                height = max(1, round(source.height * scale))
+                resampling = Resampling.nearest if product.metric_kind == "binary_area" else Resampling.bilinear
+                data = source.read(1, out_shape=(height, width), masked=True, resampling=resampling)
+                values = np.asarray(data.filled(np.nan), dtype="float64")
+                valid = np.isfinite(values) & ~np.ma.getmaskarray(data)
+                if source.nodata is not None:
+                    valid &= ~np.isclose(values, float(source.nodata))
+                if product.metric_kind == "binary_area":
+                    valid &= values >= 0.5
+                    normalized = np.where(valid, 1.0, 0.0)
+                else:
+                    source_metadata = dataset.source_metadata or {}
+                    lower = float(source_metadata.get("render_min_value", np.nanpercentile(values[valid], 2) if valid.any() else 0.0))
+                    upper = float(source_metadata.get("render_max_value", np.nanpercentile(values[valid], 98) if valid.any() else 1.0))
+                    if upper <= lower:
+                        upper = lower + 1.0
+                    normalized = np.clip((values - lower) / (upper - lower), 0.0, 1.0)
+                colors = palettes[product.theme]
+                rgba = np.zeros((height, width, 4), dtype="uint8")
+                for channel in range(3):
+                    rgba[:, :, channel] = (colors[0, channel] + normalized * (colors[1, channel] - colors[0, channel])).astype("uint8")
+                rgba[:, :, 3] = np.where(valid, 205, 0).astype("uint8")
+                output = io.BytesIO()
+                Image.fromarray(rgba).save(output, format="PNG", optimize=True)
+                return output.getvalue()
+        except AppError:
+            raise
+        except Exception as exc:
+            raise AppError(
+                code="THEMATIC_RASTER_PREVIEW_FAILED",
+                message="The persisted thematic raster could not be rendered",
+                details={"reason": str(exc)},
+                status_code=500,
+            ) from exc

backend/app/services/vector_feature_service.py

@@ -29,6 +29,7 @@ FULL_AREA_CLIPPED_OPERATOR_TOOLS = {
     "provision_regional_bwk_natura2000.py",
     "provision_agricultural_parcel_history.py",
     "provision_buildings_addresses_register.py",
+    "provision_mol_soil_map.py",
 }
 
 SEMANTIC_METRICS_DISABLED_OPERATOR_TOOLS = {
@@ -101,6 +102,16 @@ SEMANTIC_SELECTION_METRICS: dict[str, tuple[dict[str, Any], ...]] = {
     ),
     "nature_value": (),
     "agriculture": (),
+    "soil": (
+        {
+            "metric_key": "soil_mapped_area",
+            "method": "intersection_area",
+            "label": "Bodemkaartoppervlakte",
+            "unit": "ha",
+            "geometry_dimension": 2,
+            "warning": "Historische bodemkartering op schaal 1:20.000; actuele lokale bodem- en drainagetoestand kan afwijken.",
+        },
+    ),
 }
 
 SEMANTIC_COUNT_LABELS = {
@@ -112,6 +123,7 @@ SEMANTIC_COUNT_LABELS = {
     "parcels": "Percelen",
     "nature_value": "BWK-kaartvlakken",
     "agriculture": "Landbouwgebruikspercelen",
+    "soil": "Bodemkaartvlakken",
 }
 
 # Sprint 205 initially normalized two official comma-separated ALZ group labels
@@ -158,6 +170,8 @@ class VectorFeatureService:
             "landbouw": "agriculture",
             "landbouwgebruik": "agriculture",
             "building_registry": "buildings",
+            "soil_map": "soil",
+            "bodem": "soil",
         }
         for candidate in candidates:
             if not isinstance(candidate, str) or not candidate.strip():

backend/tests/test_sprint213_thematic_rasters.py

@@ -0,0 +1,339 @@
+from __future__ import annotations
+
+from pathlib import Path
+from types import SimpleNamespace
+from uuid import uuid4
+
+import numpy as np
+import pytest
+import rasterio
+from fastapi.testclient import TestClient
+from pyproj import Transformer
+from rasterio.io import MemoryFile
+from rasterio.transform import from_origin
+
+from app.core.config import Settings
+from app.core.errors import AppError
+from app.db.session import get_db
+from app.main import app
+from app.models import Dataset, Job, Project
+from app.schemas.thematic_raster import ThematicRasterAcquireRequest, ThematicRasterSelectionRequest
+from app.schemas.assistant import AssistantQueryRequest
+from app.services.geo_assistant_service import GeoAssistantService
+from app.services.thematic_raster_acquisition_service import ThematicRasterAcquisitionService
+from app.services.thematic_raster_analysis_service import ThematicRasterAnalysisService
+from app.services.dataset_service import DatasetService
+
+
+ROOT = Path(__file__).resolve().parents[2]
+
+
+class FakeQuery:
+    def __init__(self, result=None):
+        self.result = result
+
+    def filter(self, *_args):
+        return self
+
+    def order_by(self, *_args):
+        return self
+
+    def first(self):
+        return self.result
+
+    def all(self):
+        return self.result if isinstance(self.result, list) else []
+
+
+class FakeSession:
+    def __init__(self, rows=None, query_result=None):
+        self.rows = rows or {}
+        self.query_result = query_result
+        self.added = []
+
+    def get(self, model, row_id):
+        row = self.rows.get((model, row_id))
+        if row is not None:
+            return row
+        return next((item for item in self.added if isinstance(item, model) and item.id == row_id), None)
+
+    def add(self, row):
+        self.added.append(row)
+
+    def commit(self):
+        return None
+
+    def rollback(self):
+        return None
+
+    def refresh(self, row):
+        return row
+
+    def query(self, _model):
+        return FakeQuery(self.query_result)
+
+
+class FakeResponse:
+    def __init__(self, content: bytes):
+        self.content = content
+        self.headers = {"Content-Type": "image/tiff", "Content-Length": str(len(content))}
+
+    def __enter__(self):
+        return self
+
+    def __exit__(self, *_args):
+        return None
+
+    def read(self, limit: int):
+        return self.content[:limit]
+
+
+def payload(product_key: str = "space_occupation_2025", *, side_m: float = 1000.0) -> ThematicRasterAcquireRequest:
+    transformer = Transformer.from_crs("EPSG:31370", "EPSG:4326", always_xy=True)
+    min_x, min_y = transformer.transform(200_000, 210_000)
+    max_x, max_y = transformer.transform(200_000 + side_m, 210_000 + side_m)
+    return ThematicRasterAcquireRequest(
+        bbox={"min_x": min_x, "min_y": min_y, "max_x": max_x, "max_y": max_y, "crs": "EPSG:4326"},
+        product_key=product_key,
+        force_refresh=True,
+    )
+
+
+def raster_bytes(values: np.ndarray, resolution: float, *, nodata: float = -9999.0) -> bytes:
+    with MemoryFile() as memory:
+        with memory.open(
+            driver="GTiff",
+            width=values.shape[1],
+            height=values.shape[0],
+            count=1,
+            dtype=str(values.dtype),
+            crs="EPSG:31370",
+            transform=from_origin(200_000, 210_000 + values.shape[0] * resolution, resolution, resolution),
+            nodata=nodata,
+        ) as output:
+            output.write(values, 1)
+        return memory.read()
+
+
+def test_registry_contains_five_governed_non_water_policy_products() -> None:
+    products = ThematicRasterAcquisitionService.list_products()
+
+    assert [item["key"] for item in products] == [
+        "space_occupation_2025",
+        "open_space_2022",
+        "population_density_2019",
+        "node_value_2022",
+        "service_level_2022",
+    ]
+    assert {item["theme"] for item in products} == {"space_occupation", "open_space", "population", "accessibility", "services"}
+    assert {item["native_resolution_m"] for item in products} == {10.0, 100.0}
+    assert all(item["coverage_id"].startswith(("lu:", "ni:")) for item in products)
+    assert all(item["source_crs"] == "EPSG:31370" for item in products)
+    assert all(item["attribution"] and item["license_note"] and item["limitation_message"] for item in products)
+
+
+def test_request_is_bounded_allowlisted_and_uses_native_wcs_resolution() -> None:
+    settings = Settings(_env_file=None)
+    prepared = ThematicRasterAcquisitionService._prepared_request(payload("population_density_2019"), settings)
+    url = ThematicRasterAcquisitionService._wcs_request_url(settings, prepared["product"], tuple(prepared["bbox_epsg31370"]))
+
+    assert "VERSION=1.0.0" in url
+    assert "COVERAGE=ni%3Ani_inw_ha_vlaa_2019" in url
+    assert "RESX=100" in url and "RESY=100" in url
+    assert prepared["width"] * prepared["height"] <= settings.thematic_raster_max_pixels
+
+    with pytest.raises(AppError) as exc_info:
+        ThematicRasterAcquisitionService._prepared_request(payload("arbitrary_remote_layer"), settings)
+    assert exc_info.value.code == "THEMATIC_RASTER_PRODUCT_NOT_SUPPORTED"
+
+
+def test_binary_and_normalized_products_fail_closed_on_invalid_values() -> None:
+    binary = ThematicRasterAcquisitionService._product("space_occupation_2025")
+    score = ThematicRasterAcquisitionService._product("service_level_2022")
+
+    with pytest.raises(AppError, match="Binary"):
+        ThematicRasterAcquisitionService._validate_values(np.asarray([0.0, 2.0]), binary)
+    with pytest.raises(AppError, match="0-1"):
+        ThematicRasterAcquisitionService._validate_values(np.asarray([0.2, 1.2]), score)
+
+
+def test_acquisition_clips_validates_and_delegates_persistence(monkeypatch) -> None:
+    project_id, output_dataset_id = uuid4(), uuid4()
+    db = FakeSession({(Project, project_id): Project(id=project_id, name="Mol")})
+    content = raster_bytes(np.ones((100, 100), dtype="float32"), 10.0)
+    captured: dict = {}
+
+    def fake_import(_db, **kwargs):
+        captured.update(kwargs)
+        return SimpleNamespace(id=output_dataset_id)
+
+    monkeypatch.setattr(DatasetService, "import_raster_bytes", fake_import)
+    result = ThematicRasterAcquisitionService.acquire(
+        db,
+        project_id,
+        payload(side_m=1000.0),
+        settings=Settings(_env_file=None),
+        opener=lambda *_args, **_kwargs: FakeResponse(content),
+    )
+
+    assert result["output_dataset_id"] == str(output_dataset_id)
+    assert captured["source_name"] == ThematicRasterAcquisitionService.PROVIDER
+    assert captured["source_metadata"]["product_key"] == "space_occupation_2025"
+    assert captured["source_metadata"]["metric_kind"] == "binary_area"
+    assert captured["source_metadata"]["valid_pixel_count"] > 9_800
+    assert captured["provenance_metadata"]["acquisition"] == "explicit_bounded_tiled_wcs_coverage"
+    assert len(captured["provenance_metadata"]["normalized_sha256"]) == 64
+
+
+def test_binary_area_analysis_returns_hectares_and_share(tmp_path) -> None:
+    project_id, dataset_id = uuid4(), uuid4()
+    values = np.zeros((10, 10), dtype="float32")
+    values[:, :5] = 1.0
+    path = tmp_path / "space.tif"
+    path.write_bytes(raster_bytes(values, 10.0))
+    dataset = Dataset(
+        id=dataset_id,
+        project_id=project_id,
+        name="space.tif",
+        dataset_type="raster",
+        source="official",
+        source_name=ThematicRasterAcquisitionService.PROVIDER,
+        source_metadata={"product_key": "space_occupation_2025", "coverage_id": "lu:lu_ruibes_vlaa_2025_v3"},
+        status="ready",
+        storage_path=str(path),
+    )
+    db = FakeSession({(Dataset, dataset_id): dataset})
+    result = ThematicRasterAnalysisService.analyze(
+        db,
+        project_id,
+        dataset_id,
+        ThematicRasterSelectionRequest(bbox=payload(side_m=100.0).bbox),
+    )
+    metrics = {item["metric_key"]: item for item in result["summary"]["metrics"]}
+
+    assert result["valid_cell_count"] == 100
+    assert metrics["space_occupation_area_ha"]["metric_value"] == pytest.approx(0.5)
+    assert metrics["space_occupation_share_pct"]["metric_value"] == pytest.approx(50.0)
+    assert "object_count" in result["unsupported_metrics"]
+
+
+def test_population_analysis_sums_one_hectare_density_cells_without_claiming_current_counts(tmp_path) -> None:
+    project_id, dataset_id = uuid4(), uuid4()
+    values = np.asarray([[10.0, 20.0], [30.0, 40.0]], dtype="float32")
+    path = tmp_path / "population.tif"
+    path.write_bytes(raster_bytes(values, 100.0))
+    dataset = Dataset(
+        id=dataset_id,
+        project_id=project_id,
+        name="population.tif",
+        dataset_type="raster",
+        source="official",
+        source_name=ThematicRasterAcquisitionService.PROVIDER,
+        source_metadata={"product_key": "population_density_2019", "coverage_id": "ni:ni_inw_ha_vlaa_2019"},
+        status="ready",
+        storage_path=str(path),
+    )
+    db = FakeSession({(Dataset, dataset_id): dataset})
+    result = ThematicRasterAnalysisService.analyze(
+        db,
+        project_id,
+        dataset_id,
+        ThematicRasterSelectionRequest(bbox=payload("population_density_2019", side_m=200.0).bbox),
+    )
+    metrics = {item["metric_key"]: item for item in result["summary"]["metrics"]}
+
+    assert metrics["estimated_inhabitants"]["metric_value"] == pytest.approx(100.0)
+    assert metrics["population_density_mean_per_ha"]["metric_value"] == pytest.approx(25.0)
+    assert metrics["estimated_inhabitants"]["is_estimate"] is True
+    assert "current_population" in result["unsupported_metrics"]
+
+
+def test_assistant_context_receives_persisted_thematic_metrics(tmp_path) -> None:
+    project_id, dataset_id = uuid4(), uuid4()
+    values = np.asarray([[10.0, 20.0], [30.0, 40.0]], dtype="float32")
+    path = tmp_path / "assistant-population.tif"
+    path.write_bytes(raster_bytes(values, 100.0))
+    project = Project(id=project_id, name="Mol", region="Mol")
+    dataset = Dataset(
+        id=dataset_id,
+        project_id=project_id,
+        name="population.tif",
+        dataset_type="raster",
+        source="official",
+        source_name=ThematicRasterAcquisitionService.PROVIDER,
+        source_metadata={"product_key": "population_density_2019", "coverage_id": "ni:ni_inw_ha_vlaa_2019"},
+        status="ready",
+        storage_path=str(path),
+    )
+    db = FakeSession({(Project, project_id): project, (Dataset, dataset_id): dataset}, query_result=[dataset])
+    context, metrics, _series, dataset_ids, _warnings, _scope = GeoAssistantService(Settings(_env_file=None))._build_context(
+        db,
+        project_id=project_id,
+        payload=AssistantQueryRequest(question="Hoeveel inwoners?", bbox=payload("population_density_2019", side_m=200.0).bbox),
+    )
+
+    assert any(metric.theme == "population" and metric.label.startswith("Geraamd aantal") for metric in metrics)
+    assert dataset_id in dataset_ids
+    assert context["rules"]["thematic_policy_rasters_available"] is True
+
+
+def test_index_renderer_returns_browser_png(tmp_path) -> None:
+    project_id, dataset_id = uuid4(), uuid4()
+    values = np.linspace(0.1, 4.0, 100, dtype="float32").reshape((10, 10))
+    path = tmp_path / "node.tif"
+    path.write_bytes(raster_bytes(values, 100.0))
+    dataset = Dataset(
+        id=dataset_id,
+        project_id=project_id,
+        name="node.tif",
+        dataset_type="raster",
+        source="official",
+        source_name=ThematicRasterAcquisitionService.PROVIDER,
+        source_metadata={
+            "product_key": "node_value_2022",
+            "coverage_id": "lu:lu_knptw_ha_2022_v3",
+            "render_min_value": 0.1,
+            "render_max_value": 4.0,
+        },
+        status="ready",
+        storage_path=str(path),
+    )
+    db = FakeSession({(Dataset, dataset_id): dataset})
+
+    assert ThematicRasterAnalysisService.render_png(db, project_id, dataset_id).startswith(b"\x89PNG\r\n\x1a\n")
+
+
+def test_api_uses_canonical_envelopes(monkeypatch) -> None:
+    project_id, dataset_id = uuid4(), uuid4()
+    db = FakeSession({(Project, project_id): Project(id=project_id, name="Mol")})
+    monkeypatch.setattr(
+        ThematicRasterAcquisitionService,
+        "acquire",
+        lambda *_args, **_kwargs: {"output_dataset_id": str(dataset_id), "provider": ThematicRasterAcquisitionService.PROVIDER},
+    )
+    monkeypatch.setattr(
+        ThematicRasterAnalysisService,
+        "analyze",
+        lambda *_args, **_kwargs: {"dataset_id": str(dataset_id), "theme": "population", "summary": {"metric_value": 10.0}},
+    )
+    app.dependency_overrides[get_db] = lambda: db
+    try:
+        client = TestClient(app)
+        products = client.get(f"/api/v1/projects/{project_id}/datasets/thematic-raster/products")
+        acquisition = client.post(
+            f"/api/v1/projects/{project_id}/datasets/thematic-raster/acquire",
+            json=payload().model_dump(mode="json"),
+        )
+        selection = client.post(
+            f"/api/v1/projects/{project_id}/datasets/{dataset_id}/raster/thematic/select",
+            json={"bbox": payload().bbox.model_dump()},
+        )
+    finally:
+        app.dependency_overrides.clear()
+
+    assert products.status_code == 200 and set(products.json()) == {"data"}
+    assert products.json()["data"]["total"] == 5
+    assert acquisition.status_code == 200 and set(acquisition.json()) == {"data"}
+    assert acquisition.json()["data"]["job_type"] == "raster.thematic.acquire"
+    assert selection.status_code == 200 and selection.json()["data"]["theme"] == "population"
+    assert any(isinstance(item, Job) for item in db.added)

backend/tests/test_sprint214_dov_soil_map.py

@@ -0,0 +1,183 @@
+from __future__ import annotations
+
+import importlib.util
+from pathlib import Path
+from uuid import uuid4
+
+import pytest
+from shapely.geometry import box, mapping, shape
+from shapely.ops import transform as transform_geometry
+
+from app.models import Dataset
+from app.services.vector_feature_service import VectorFeatureService
+
+
+ROOT = Path(__file__).resolve().parents[2]
+
+
+def load_operator():
+    path = ROOT / "scripts" / "provision_mol_soil_map.py"
+    spec = importlib.util.spec_from_file_location("dov_soil_map_operator", path)
+    assert spec and spec.loader
+    module = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(module)
+    return module
+
+
+class FakeResponse:
+    def __init__(self, payload: dict, url: str):
+        self._payload = payload
+        self.url = url
+        self.content = b'{"type":"FeatureCollection"}'
+
+    def raise_for_status(self) -> None:
+        return None
+
+    def json(self) -> dict:
+        return self._payload
+
+
+class FakeSession:
+    def __init__(self, pages: list[dict]):
+        self.pages = pages
+        self.calls: list[dict] = []
+
+    def get(self, _url: str, *, params: dict, timeout: int):
+        self.calls.append({"params": dict(params), "timeout": timeout})
+        return FakeResponse(self.pages[len(self.calls) - 1], f"https://example.test/page/{len(self.calls)}")
+
+
+def soil_feature(module, feature_id: str = "bodemtypes.1") -> dict:
+    return {
+        "type": "Feature",
+        "id": feature_id,
+        "geometry": mapping(box(5.0, 51.0, 5.02, 51.02)),
+        "properties": {
+            "gid": 1,
+            "id_kaartvlak": 10,
+            "Bodemtype": "Zeg",
+            "Unibodemtype": "Zeg",
+            "Bodemserie": "Zeg",
+            "Beknopte_omschrijving_bodemserie": "Natte zandbodem",
+            "Gegeneraliseerde_legende": "Nat zand",
+            "Textuurklasse_code": "Z",
+            "Textuurklasse": "zand",
+            "Drainageklasse_code": "e",
+            "Drainageklasse": "nat",
+            "Profielontwikkelingsgroep_code": "g",
+            "Profielontwikkelingsgroep": "humus B horizont",
+            "Eenduidige_legende_titel": "bodemserie Zeg",
+        },
+    }
+
+
+def test_wfs_pagination_is_bounded_complete_and_deterministic() -> None:
+    module = load_operator()
+    feature = soil_feature(module)
+    pages = [
+        {
+            "type": "FeatureCollection",
+            "numberMatched": 3,
+            "numberReturned": 2,
+            "features": [feature, {**feature, "id": "bodemtypes.2"}],
+        },
+        {
+            "type": "FeatureCollection",
+            "numberMatched": 3,
+            "numberReturned": 1,
+            "features": [{**feature, "id": "bodemtypes.3"}],
+        },
+    ]
+    session = FakeSession(pages)
+
+    result = list(
+        module.iter_wfs_pages(
+            session,
+            (196000.0, 205000.0, 211000.0, 224000.0),
+            page_limit=2,
+            timeout=30,
+        )
+    )
+
+    assert len(result) == 2
+    assert [call["params"]["startIndex"] for call in session.calls] == ["0", "2"]
+    assert all(call["params"]["typeNames"] == "bodemkaart:bodemtypes" for call in session.calls)
+    assert all(call["params"]["bbox"].endswith("EPSG:31370") for call in session.calls)
+    assert all(call["params"]["sortBy"] == "gid" for call in session.calls)
+
+
+def test_soil_feature_is_exactly_clipped_and_keeps_governed_properties() -> None:
+    module = load_operator()
+    boundary_wgs84 = box(5.005, 51.005, 5.015, 51.015)
+    boundary_lambert72 = transform_geometry(module.TO_LAMBERT72.transform, boundary_wgs84)
+
+    normalized, was_clipped = module.normalize_feature(soil_feature(module), boundary_lambert72)
+
+    assert normalized is not None and was_clipped is True
+    persisted_geometry = shape(normalized["geometry"])
+    assert persisted_geometry.within(boundary_wgs84.buffer(1e-7))
+    properties = normalized["properties"]
+    assert properties["source_name"] == "dov_soil_map"
+    assert properties["soil_texture_class"] == "zand"
+    assert properties["soil_drainage_class"] == "nat"
+    assert properties["survey_period"] == "1949-1971"
+    assert properties["clipped_area_ha"] > 0
+    assert "may differ today" in properties["historical_drainage_limitation"]
+
+
+def test_soil_map_uses_existing_semantic_selection_architecture() -> None:
+    dataset = Dataset(
+        id=uuid4(),
+        project_id=uuid4(),
+        name="dov_soil_map_mol.geojson",
+        dataset_type="vector",
+        source="operator_official_import",
+        source_name="dov_soil_map",
+        reference_layer_name="soil",
+        source_metadata={
+            "theme": "soil",
+            "selection_aggregation": {
+                "method": "intersection_area",
+                "label": "Bodemkaartoppervlakte",
+                "unit": "ha",
+            },
+        },
+        status="ready",
+    )
+
+    assert VectorFeatureService._dataset_theme(dataset) == "soil"
+    assert VectorFeatureService.supports_selection_summary(dataset) is True
+    assert VectorFeatureService.can_use_full_area_fast_path(dataset, None) is False
+
+
+def test_soil_operator_contract_has_no_direct_persistence_and_is_packaged() -> None:
+    operator = (ROOT / "scripts" / "provision_mol_soil_map.py").read_text(encoding="utf-8")
+    dockerfile = (ROOT / "deploy" / "unraid" / "Dockerfile.all-in-one").read_text(encoding="utf-8")
+    readiness = (ROOT / "scripts" / "run_readiness_check.sh").read_text(encoding="utf-8")
+    map_workspace = (ROOT / "frontend" / "src" / "components" / "map" / "MapWorkspace.tsx").read_text(encoding="utf-8")
+
+    assert "/datasets/upload" in operator
+    assert "vector_features" in operator
+    assert "does not write directly" in " ".join(operator.split())
+    assert "SessionLocal" not in operator and "INSERT INTO" not in operator
+    assert "COPY scripts/provision_mol_soil_map.py" in dockerfile
+    assert "py_compile scripts/provision_mol_soil_map.py" in readiness
+    assert "id: 'soil'" in map_workspace
+    assert "dataset.source_name === 'dov_soil_map'" in map_workspace
+
+
+def test_incomplete_wfs_pagination_fails_closed() -> None:
+    module = load_operator()
+    session = FakeSession(
+        [
+            {
+                "type": "FeatureCollection",
+                "numberMatched": 2,
+                "numberReturned": 0,
+                "features": [],
+            }
+        ]
+    )
+
+    with pytest.raises(RuntimeError, match="returned 0 of 2"):
+        list(module.iter_wfs_pages(session, (0.0, 0.0, 1.0, 1.0), page_limit=100, timeout=30))