Landscape delineation — Task

Implementation: Landscape delineation (Recipe - QGIS).

Landscape delineation is the Task of dividing a study area into landscapes: areas where landform, soils, water and land use interact in a recognisable way, such as a till plain, a valley system or hummocky moraine. The landscapes become the zones that later analyses are summarised by, so how they are drawn decides what those summaries can say.

Used in examples: Habitat network at Ledreborg (Example), Soil carbon and land use at Ledreborg (Example).

Unlike most Tasks, this one is mainly interpretation: the Operators are simple, and the work is in reading several layers together and deciding where one landscape ends and the next begins.

Task Schema (tool-agnostic)

1. Define the study area (decision point; Operator: digitise) — a rectangle of about 5 × 5 km, large enough to hold at least three landscapes.
2. Assemble the evidence layers (decision point) — the layers that show the landscape’s structure: landform (geomorphology, terrain model and hillshade), soils, historical maps (land use before modern drainage and amalgamation), and a recent orthophoto.
3. Read the boundaries (decision point) — where do landform, soil and land use change together? A boundary that several layers agree on is a strong one.
4. Split the study area (Operator: split polygon along a line) — draw each boundary across the rectangle; every piece becomes a landscape polygon.
5. Name the landscapes (decision point) — a type name per polygon (e.g. till plain). Several polygons may share a type.

Output: a polygon layer that covers the study area without gaps or overlaps, one landscape type per polygon.

Epistemic Choices to record

• Which layers drove which boundary, and where the layers disagreed.
• Scale: boundaries drawn at 1:25,000 are generalised; a boundary is a zone of transition, not a line in the terrain.
• Types or individuals: two separate valley systems are one type but two landscapes. Decide whether later summaries are per polygon or per type.
• Naming consistency: the type names are the categories the analysis groups by, so they must be spelled identically (a misspelt name becomes a separate landscape).