ELSEVIER
Multivariate Statistical Classification
of Soil Spectra
Alicia Palacios-Orueta* and Susan L. Ustin*
The purpose of this work was to evaluate whether
A VIRIS (Advanced Visible Infrared Imaging Spectrome-
ter) bands can be used to discriminate between soils
having similar properties, as well as to compare A VIRIS
spectra with those from laboratory measurements. Multi-
variate analysis techniques show that two soils belonging
to the same series and a third soil belonging to a different,
but related series can be discriminated at a high level of
accuracy using reflectance data from A VIRIS or from
laboratory measurements. It is also shown that wave-
lengths important in discriminating soils were highly
correlated between AVIRIS and laboratory data. The
distribution of variance and weighting functions also show
consistent patterns between these data sets.
INTRODUCTION
The monitoring of soils is a major challenge to soil
scientists. Soils exhibit continuous variation in space
and time. Both natural and anthropogenic processes,
such as agriculture or grazing practices, can change soil
properties in ways that produce cumulative environmen-
tal impacts. Studies of spatial and temporal changes are
needed for identifying environmental problems and for
developing mitigation strategies. Monitoring small
changes are difficult because of the high spatial variabil-
ity. It is likely that most soil changes will be small on
decadal time scales relative to the large spatial heteroge-
neity existing in the pedisol even on a local scale.
Nonetheless, the potential significance of biogeochemi-
cal feedbacks onto the climate system make such mea-
surements important. We have examined the application
of high spectral resolution remote sensors as a means
* Department of Land, Air, and Water Resources, University of
California, Davis
Address correspondence to Alicia Palacios, Dept. of Land, Air,
and Water Resources, Univ. of California, Davis, CA 95616.
Received 5 July 1995; revised 13 November 1995.
REMOTE SENS. ENVIRON. 57:108-118 (1996)
©Elsevier Science Inc., 1996
655 Avenue of the Americas, New York, NY 10010
to provide spatial information on the properties and
conditions of soil surfaces and to evaluate its potential
to detect small differences in soil properties. The soils
chosen for this study exhibit a range of variation compa-
rable to potential decadal scale temporal changes that
result from altered surface processes.
Spectrometry has been used for 25 years to differ-
entiate highly divergent soil types. Researchers have
shown that soils having dissimilar properties can be
discriminated using reflectance measurements. Condit
(1972) classified 160 soil spectra into three types based
on their spectral shapes, the first type corresponding to
chenozern soil, the second to pedalfer, and the third
type to a red quartz and calcite sand. In an analysis of
485 soils, representing the major orders of the globe,
Stoner and Baumgardner (1981) distinguished five dis-
tinct reflectance forms, based on curve shapes and ab-
sorption bands. These soils had different physical and
chemical characteristics; however, spectral differences
were mainly due to organic matter and iron oxide con-
tent. Other recent studies have supported these findings
(Csillag et al., 1993; Henderson et al., 1992; Zhang et
al., 1992).
Satellite multispectral data has been used as well
to discriminate between soil types (Lee et aI., 1988;
Lewis et al., 1975; Smith et al., 1990; Suet al., 1989,
1990; Thompson et al., 1981, 1984; Wright and Birnie,
1986). Agbu et al. (1990) used Systeme Pour I' Observa-
tion de la Terre (SPOT) data and found that some
soil properties like organic matter content, particle size
distribution, and color could be used to predict satellite
reflectance. Coleman et al. (1993) used the Thematic
Mapper (TM) to differentiate surface soils with high
accuracy and found significant correlations between ra-
diance data and organic matter, iron content, and parti-
cle size distribution.
The development of imaging spectrometry makes
it possible to obtain observations at a much higher
spectral resolution. De Jong (1992) investigated the use
0034-4257 / 96 ! $15.00
SSDI 0034-4257(95)00250-2