APPROACHES
TO CLASSIFICATION AT DIFFERENT SCALES
The approach adopted in a particular assessment exercise will depend on
the level in the hierarchy at which the assessment is taking place and
therefore the scale and level of detail required. The larger the scale
the more reliance there is likely to be on computer techniques to help
with data handling and the less the scope for comprehensive fieldwork
or for significant stakeholder involvement.
Classification at the large (national/ regional) scale
Assessments which cover large areas, at the national or regional level,
may be either: 'top- down', in that they consist of a quite broad,
generalised characterisation based on identification of large-scale
patterns of character which may then be characterised in more detail at
the next level in the hierarchy (demonstrated by the Countryside
Character initiative); or 'bottom-up' in that they are based on
amalgamation of, and generalisation from, more detailed character types
or areas identified at a lower level in the hierarchy (demonstrated by
Scotland's national programme of Landscape Character Assessment).
Top
down assessments tend, for practical reasons,
to be predominantly desk-based exercises
relying largely on map information. They are concerned with identifying
broad regional patterns of character in the landscape resulting from
particular combinations of geology, soils, topography and settlement
and enclosure patterns. Maps of these attributes prepared in the desk
study stage need to be overlaid and patterns identified so that areas
of relatively homogeneous character can be mapped. This can be done
manually, with patterns identified by eye, but at this scale maps in
paper form can become unwieldy and manual analysis difficult. In
general it is also simply not practical to carry out comprehensive
fieldwork at this scale although field survey can be used in a targeted
way to add information to that available from maps.
The
use of GIS combined with computer
classification techniques can be particularly
helpful at this scale, because of these difficulties. These tools allow
different sets of map data to be combined, manipulated and correlated
more easily and help to identify the large-scale patterns of landscape
character. Examples of the types of techniques that can be used and of
their practical application can be found in Topic Paper 4. Such
techniques can either be used alone, or in
combination with manual approaches. It is
important to recognise, however, that work relying wholly on computer
classification is rarely entirely satisfactory as it omits the critical
contributions of both fieldwork and stakeholder involvement.
|