GITTA-Logo
PDF Version of this document Search Help

Lesson Navigation IconDiscrete Spatial Distributions

Unit Navigation IconIntroduction

Unit Navigation IconSpatial Dependency

Unit Navigation IconSpatial Arrangement

LO Navigation IconIndices Of Arrangement

LO Navigation IconAt the scale of the study area: “the structure”

LO Navigation IconIndices of structure at the level of the objects

LO Navigation IconIndices of structure at the level of the categories / classes of objects

LO Navigation IconIndices of structure at the level of the whole study area

LO Navigation IconAt the scale of the neighborhood: “the texture”

LO Navigation IconPrinciples of the contextual analysis

LO Navigation IconIndices of central tendency

LO Navigation IconIndices of variability

LO Navigation IconIndices of texture of first order

Unit Navigation IconBibliography

Unit Navigation IconMetadata


GITTA/CartouCHe news:


Go to previous page Go to next page

Indices of texture of first order

1.3.10a Indices of texture of first order

These indices of texture are based on the dispersion of the properties in the neighborhood, they are thus dependent on the level of measurement of information. The following indices will be retained:

Diversity (DIV), nominal level:
c: number of different values present in the neighborhood
Interpretation: It is a simple indication of diversity expressing the thematic richness in the neighborhood.
It equivalent to the index of heterogeneity (NDC) previously described.

Interquartile (IQ), ordinal level:
Q1: first quartile
Q3: third quartile
Interpretation: It expresses the number of different values ranging
between the first and the third quartile (the difference of rank),
therefore the dispersion around the median value.
It is noted that only at the ordinal level the index of dispersion IQ
comes to supplement the pallet of the contextual indices previously presented.

1.3.10b Exploitation of the contextual indices

In order to illustrate the potential and the complementarity, but also the redundancy of contextual indices presented in this Unit, let us apply to the description of the spatial arrangement of land cover types in an urban area. Figure 3.6 presents the spatial distribution of 5 categories of land cover (Forest/bush, Meadow, Building, Road, Pavement) in an urban area made up of 4 distinct zones: individual housing, collective habitat, industrial park and green area. The size of the cells is of approximately 15m on side. These thematic properties being categories, they correspond to a nominal level of measurement.

Urban area made up of 5 different zones
Figure 3.6Figure 3.6

The description of the spatial arrangement of thematic properties of these units of observation (cells) can begin with the identification of the most represented land cover category in the neighborhood. We will thus choose to retain the modal value in the moving window. In order to experiment the effect of the neighborhood size on this index of majority (central tendency), we will choose a window size of 3x3 as well as 5x5.

Modal property of the neighborhood in windows of size 3x3 and 5x5
Figure 3.7 Figure 3.7

One can observe on figure 3.7 the prevalence of the meadow category for the 2 zones of habitat, of the forest/bush type for the green area and of the pavement category for the industrial park. When a window of larger size (5x5) is applied, then the spatial smoothing is more marked, thus reducing the presence of vaster regions (objects) such as the buildings of big size in the industrial park and in the collective habitat.

Indices of texture for windows of size 3x3 and 5x5
Figure 3.8 Figure 3.8

Exercises


  • Which is the influence of the values assigned with the properties on the value of each produced index?
    • what can one conclude from it?
Top Go to previous page Go to next page