Zonal operations are also sometimes called region operations or region functions. We commonly use zonal operations as a potential problem exists when we wish to analyse raster data that is not actually the shape of an individual grid cell. For example, neighbourhood operations define their area of interest as a function of individual grid cells. With zonal operations we can analyse spatial definitions of regions by clustering similar or like cell values into homogenous regions.

Zones can be defined at the data capture stage such as when vegetation polygons are digitised. Alternatively, zones can be created using reclassification techniques discussed elsewhere in this module. Processing of zones is formed on a zone-by-zone basis rather than on a cell-by-cell basis (see the neighbourhood operators lesson for a description of cell-by-cell analysis).

Each zone of a map layer is a set of contiguous locations that exhibit the same value, which might be:
land ownership parcels
political units such as counties or nations
lakes or islands
individual patches of the same soil or vegetation type

A zone such as a vegetation polygon is used in a raster analysis as the definition of the area of interest. Therefore in zonal operations we work with two raster data layers; one defining the zone and the other that we are analysing as a function of the zones. An example of this is where we wish to calculate the average elevation for a particular forest type.

Note that not all map layers will have zones, cell contents may vary continuously over the region making every cell's value unique e.g. satellite sensors record a separate value for reflection from each cell.