Land-use composition per zone — Task

Implementation: Aggregating a categorical raster to zones (Recipe - QGIS).

Land-use composition per zone is the Task of measuring how much of each land-use class lies in each zone — each landscape, municipality or grid cell — from a categorical land-use raster. The result is a table of areas or shares, one row per zone and one column per class, ready for a chart.

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

Task Schema (tool-agnostic)

1. Choose the land-use source and its classes (decision point) — e.g. Basemap04’s aggregated classes (lu_agg_2021), with names from its legend.
2. Choose the zones — e.g. the landscapes from landscape delineation.
3. Count the cells of each class in each zone (Operator: tabulate area) — as area or as percentage.
4. Group zones of the same type (decision point) — add up the areas of polygons of the same landscape type, then compute the shares.
5. Group classes for presentation (decision point) — e.g. all built-up classes as “Built-up”; 30 classes do not fit in one chart.

Output: a table of land-use area and share per zone (or zone type), and a 100 % stacked bar chart.

Epistemic Choices to record

• Classification: Basemap04 combines several sources; its classes are not those of field observation. “Agriculture, extensive” is a land-use class, not a habitat type.
• Cell-centre rule: a cell counts in the zone that contains its centre; small zones get edge effects.
• Shares of what: of all cells, or only of cells with a class? Unmapped cells and water change the denominator.
• Grouping: combine areas, never average shares: the share for a landscape type with two polygons is the summed area of each class divided by the summed total.