Environmental Engineering Reference
In-Depth Information
Spatial Scale and Scaling
In addition to understanding the type (e.g., trend, patch) and strength (e.g., degree and
distance) of a spatial pattern, it is useful to identify the spatial scale at which such
patterns are present. Because patterns are the result of multiple processes that each
have their own unique scales of spatial structure [
75
], disentangling the relative
contributions of these processes and assigning relative importance to them and the
scales at which they operate are of fundamental importance to ecology and improving
the understanding of complex systems [
11
,
16
]. Furthermore, the identification of the
relative contributions of different processes and scales to observed patterns is neces-
sary for understanding cross-scale interactions [
33
] which is necessary to make
reliable predictions of system dynamics, ostensibly the objective of any spatial
analysis [
76
].
As stated above, spatial pattern describes a “quantifiable attribute of a spatial
context.” Scale-specific analysis identifies the spatial scale at which that attribute is
structured establishes its specific context. With regard to forest disturbance dynamics,
for example, individual forest stands may seem unstable through the processes of
destruction and renewal through disturbances such as fire, but the larger forest land-
scape (i.e., collection of stands) is in fact stable with respect to the proportion and
relative configuration of the different stand types. This is what is meant when distur-
bance-mediated forest systems are described as a shifting mosaic [
77
]. Different
conclusions would be drawn about forest stability and resilience (sensu [
78
]) depending
on the spatial or temporal scale of investigation.
Scale generally describes the spatial extent, grain, and thematic resolution of
a set of data [
6
]. However, scale can also be used to refer to a level within an
organizational hierarchy to which such data pertain, such as a population, or
community, or ecosystem [
79
]. It is important to note that scale in this latter
sense is not directly equivalent to the former; scaling up, that is, increasing from
local to a broader extent, or aggregating data from a fine to a coarse scale may move
the analysis into another level or an organizational hierarchy, but not necessarily
[
80
]. Although scale is best thought of as a feature of the phenomena of interest, it
can also be a feature of sampling scheme imposed by a researcher, or the methods of
analysis applied [
6
]. All of these features can influence a researcher's ability to
identify scales of structure in spatial data and to make meaningful inferences
regarding the underlying processes. It is therefore important to distinguish structure
that is emergent from the data from those related to sampling or analytic scales (i.e.,
arbitrary scales; [
10
]), as such a priori scales may have little to do with the actual
scale of structure in the ecological phenomena of interest [
76
].
The ability to identify meaningful scales of spatial structure depends on the
methods used and the type of data being analyzed [
77
,
81
]. Methods differ in their
ability to identify local vs. global scales of pattern. Global methods of scale-specific
analysis summarize spatial pattern at a single scale and generally assume that the
underlying processes are stationary. Examples of global methods of analysis
include variography [
58
], spectral (i.e., Fourier) analysis [
82
], and global measures
