Page 48 - Geochemical Anomaly and Mineral Prospectivity Mapping in GIS

44 Chapter 2

find edges of polygons of different lithologic units whereby it returns a value of [1] for
interior of a lithologic unit, [2] along the contacts of two lithologic units, and [3] or
higher values where three or more lithologic units join (cf. Mihalasky and Bonham-
Carter, 2001).
The preceding examples belong to the aggregation type of neighbourhood operations.
Clear introductory discussions of other different types of spatial filters, particularly those
used in raster image analysis, and the functions associated with such filters can be found
in Bonham-Carter (1994, p. 204-212). Other types of neighbourhood operations involve
‘spread’ or ‘seek’ computations. Spread computations are applicable, for example, to
flood inundation studies (e.g., Peter and Stuart, 1999) or environmental pollution studies
(e.g., Haklay, 2007). Seek computations are applicable, for example, to hydrological
studies (e.g., Vieux, 2004).

Map overlay operations

The previously discussed operations on spatial data – spatial query and selection,
classification and re-classification, measurements, transformations, neighbourhood – are
usually applied to analyze spatial patterns of interest in single maps of geoscience spatial
data sets. However, the previously discussed operations could also actually involve at
least a pair of maps. For example, selection of stream sediment samples in zones
underlain by certain lithologic units (Fig. 2-8) involves a map of stream sediment sample
locations and a lithologic map. In addition, mapping of stream sediment sample
catchment basins via a neighbourhood operation involves a map of stream sediment
sample locations, a map of drainage lines and a DEM. These examples show that map
overlay operations are implicitly involved in some of the previously discussed operations
on spatial data. Map overlay operations are perhaps the most important of all GIS
functionalities. There are two important conditions that must be fulfilled in order to
perform overlay operations: (1) maps are georeferenced to the same coordinate system;
(2) maps must overlap and thus pertain to the same study area. The principle in overlay
operations is to integrate maps of certain attributes of every location in order to produce
a map of new attributes for every location.
The three most common overlay operations are clip, overwrite and intersect (Fig. 2-
16). The clip operation, which is also called an impose operation, restricts the spatial
extent of the first map to the spatial extent of the second map (the clip map) (Fig. 2-
16A). The clip operation is useful, for example, to retrieve from a source thematic map
spatial data pertaining to a study area. The clip operation does not result in a new
attribute table; the output map adopts the attribute table of the first map. The overwrite
operation, which is also called a stamp operation, adopts the data from the first map
except where there are data in the second map; data in the second map take priority in
the output. The overwrite operation is useful, for example, to update an existing
lithologic map with recent results of lithologic mapping (Fig. 2-16B). The overwrite
operation results in an attribute table for the output map only if the second map has new
data attributes. Creation of a new attribute, however, is not necessary if the attribute
table of the first map is updated initially so that it can be associated with the second map.