ε = K ε0
pavement where salt is used in the snow-removal
program). Since salt water is so highly conduc-
where ε0 = 8.85 pF/m.
tive, the presence of salt in any appreciable amount
causes the probe to read just the resistance of the
soil, or worse, the inductance of the nulling in-
Note that K = 1 for air and for a vacuum.
ductor, since the capacitance of the probe is shorted
Since capacitances are measured using this par-
out by the low resistance of the salt water. The
ticular technique, the dielectric constant is the
extent of this problem can be demonstrated by
ratio of the capacitance of the soil divided by the
such a simple procedure as placing the probe in
capacitance of air (or of some other material whose
water, observing the values of Vref, Vdiv, and Vph,
dielectric constant is well known, such as water).
and then adding salt to the water to lower the
Thus
conductivity of the solution and observing the
Csoil = Ksoil ε0 (a constant
changes in Vref, Vdiv, and Vph. If these altered read-
based on probe geometry)
ings are used to carry out an actual calculation for
capacitance, it will be seen that when the salt is
and
added (simulating an electrical short circuit), Vph
Cair = Kair ε0 (a constant based
goes positive, indicating an inductance rather than
on probe geometry)
a capacitance. If the dielectric constant were to be
calculated for these values, it would be negative
so that
(which is, of course, impossible).
The extent to which each of these two prob-
Ksoil = Csoil/Cair, since Kair = 1.
(2)
lems affects the use of this technique for the mea-
surement of soil moisture is discussed in some
Thus, the air capacitance for each probe must be
detail in this report.
found or else an average value for all probes must
Another concern, not necessarily a problem, is
be established and the resultant error from using
the conversion of a dielectric constant measure-
this average value determined. Another approach
ment to a soil moisture measurement. The sim-
would be to use water rather than air as a refer-
plest and most direct method is to measure the
ence material. For this approach, the temperature
dielectric constant at known values of soil mois-
of the water should be determined and used, as
ture as determined by actually weighing the test
shown in Table 1. Note that the capacitance of the
soil sample. Starting with a totally dry soil and
probe in air can be found using water as a refer-
adding known amounts of water generates a curve
ence material by using the relationship
of dielectric constant vs. moisture content (per-
cent of dry weight) that will serve as a calibration
Cair = Cwater/Kwater .
(3)
curve for all similar type soils. Such a procedure
Typically Kwater is 80 at room temperature, but
will not account for the variability of soil types, so
the actual value for any temperature may be found
a new curve would have to be generated for each
in Table 1.
new soil type; even then there is substantial vari-
A second and much more serious error is the
ability within a given soil type since soils are
use of this measurement technique in a highly
nonhomogeneous. The amount of the error within
saline soil (desert salt flats or possibly under a
any soil type can be established using statistical
methods, but the process requires enough data to
be able to make statistically valid statements. An
Table 1. Dielectric constant of water.
alternative approach would be to determine the
dielectric constant as a function of the volumetric
T (C)
T (C)
K
K
moisture content. These data may then be used to
determine the moisture content for soils with dif-
0
88.00
40
73.28
5
86.40
45
71.59
ferent densities. Such a procedure requires addi-
10
84.11
50
69.94
tional testing to determine the dry density and
15
82.22
60
66.74
the specific gravity of the particular soil under
20
80.36
70
63.68
consideration.
25
78.54
80
60.78
Several researchers have used calibration meth-
30
76.75
90
57.98
35
75.00
100
55.33
ods such as these to generate equations that relate
3
Previous Page
