Rev. Bras. Ciênc. Solo.2016;40:e0160007.
A Classification for a Geostatistical Index of Spatial Dependence
09/dez/2016
DOI: 10.1590/18069657rbcs20160007
ABSTRACT:
In geostatistical studies, spatial dependence can generally be described by means
of the semivariogram or, in complementary form, with a single index followed by
its categorization to classify the degree of such dependence. The objective of
this study was to construct a categorization for the spatial dependence index
(SDI) proposed by Seidel and Oliveira (2014) in order to classify spatial
variability in terms of weak, moderate, and strong dependence. Theoretical
values were constructed from different degrees of spatial dependence, which
served as a basis for calculation of the SDI. In view of the form of
distribution and SDI descriptive measures, we developed a categorization for
posterior classification of spatial dependence, specific to each semivariogram
model. The SDI categorization was based on its median and 3rd quartile, allowing
us to classify spatial dependence as weak, moderate, or strong. We established
that for the spherical semivariogram: SDISpherical (%) ≤ 7 % (weak
spatial dependence), 7 % < SDISpherical (%) ≤ 15 % (moderate
spatial dependence), and SDISpherical (%) > 15 % (strong spatial
dependence); for the exponential semivariogram: SDIExponential (%) ≤
6 % (weak spatial dependence), 6 % < SDIExponential (%) ≤ 13 %
(moderate spatial dependence), SDIExponential (%) > 13 % (strong
spatial dependence); and for the Gaussian semivariogram: SDIGaussian
(%) ≤ 9 % (weak spatial dependence), 9 % < SDIGaussian (%) ≤ 20 %
(moderate spatial dependence), and SDIGaussian (%) > 20 % (strong
spatial dependence). The proposed categorization allows the user to transform
the numerical values calculated for SDI into categories of variability of
spatial dependence, with adequate power for explanation and comparison.
80
