Agriculture Reference
In-Depth Information
exposed soil profiles to be described. The data recorded are primarily soil depth,
color, stoniness, field texture, and the presence or absence of CaCO
3
. The results
are usually displayed not as a map but as point profile diagrams that are not nec-
essarily linked to any soil classification. Bramley et al. (2010) refer to statistical
techniques whereby this point data can be converted to continuous distributions
of the soil properties observed. However, because of the limitations of a 75
×
75
m grid survey in detecting short-range variation in a soil property, such maps are
not as useful as those derived from high-resolution sensing and calibration tech-
niques, as described in the next section.
Modern Alternatives
High-resolution sensing of soil properties, either at close range (proximally) or
remotely, can now provide soil information at an appropriate scale for intensive
vineyard management. Soil properties are “sensed” in a dense spatial array in
which the sampling points are accurately located by a global positioning system
(GPS). The GPS relies on fixing a point on the surface with reference to orbiting
satellites, a system in common use being the differentially corrected GPS, which
is accurate to less than ±50 cm in Australia, depending on the distance to the
base station. However, a more accurate instrument finding increasing use, espe-
cially in broad-acre agriculture, is the dual-frequency, real-time kinematic GPS,
which is accurate to 2 cm in the horizontal and vertical planes. When used with
a height reference level, the real-time kinematic GPS enables a digital elevation
model of the site to be constructed. This can then be displayed using a geographic
information system (GIS) in a three-dimensional map. When an real-time kine-
matic GPS is used in combination with a real-time sensing instrument such as an
EM38 (see the following), vineyard soil data can be rapidly collected at a density
of 200 to 600 points per ha, depending on the travel speed and row spacing. These
data are interpolated using a special statistical technique called kriging, to give a
high-resolution map (e.g., a 2
×
2 m grid) of a soil property, such as depth, which
can be directly related to the surface topography through the digital elevation
model. Figure 1.15 illustrates the result. More details of this approach are given in
“Starting the Soil Survey,” chapter 2.
Currently, the most widely used soil sensors are based on the electromagnetic
spectrum, including electromagnetic induction (as in the EM38), gamma-ray
spectroscopy measuring gamma radiation emitted by naturally occurring radio-
active elements in the top 30 to 45 cm of soil, ground-penetrating radar (using
long-wavelength radiation), and laser-imaging radar (using ultraviolet, visible,
and near-infrared radiation). Electronic signals from the sensors are streamed to
a logger, where the data are stored in digital format for subsequent downloading
to a computer and processing in a GIS. A GIS is specialized computer software in
