Geography Reference
In-Depth Information
Distance
Large : 2.828
Small : 0.000
0
1
A
B
Figure 4.3
Distances from linear feature: (A) distance, (B) binary map for
distances less than or equal to 1.4 units (value of 1) or greater (0).
values. A buf er can easily be generated from this grid using a simple classii cation
procedure and Figure 4.3B shows a buf er for distances of less than or equal to 1.4
units. With this approach, all cells with distances of less than, or equal to, the specii ed
amount are coded '1' and all cells with distances greater than this value are coded '0.
It is then straightforward to select all cells in a second image which fall within the
buf er dei ned in the i rst image. Section 10.2 shows how this can be done. Note that
a raster proximity map can be generated directly from vector data, as well as from
particular cells in another raster grid.
h e following section deals with a key concept in spatial data analysis, the moving
window.
Moving windows: basic statistics in subregions
4.6
In many cases, spatial variables have dif erent properties at dif erent locations (for
example, values tend to be large in some areas and small in others). In such cases, it is
useful to be able to account for these dif erences and moving windows of er one solu-
tion. h e idea of the moving window is core to spatial data analysis. In simple terms a
moving window represents a region covering part of the entire study area and this
region or 'window' is moved from one location to another. Usually, the window is
circular or square in shape. In many applications, the window moves in regular steps
across the study region, with some operation (e.g. the calculation of the mean average
of values in the window) conducted at each location. Such an approach to computing
Search WWH ::
Custom Search