that is higher than 0.4 V, or even non-negative.
ity and consequently the strength of the electro-
Because the soil's electrical conductivity varies
magnetic field was high.
directly with its unfrozen moisture content,
On 28 March the shallow soil was moist un-
freezing and thawing of the soil will affect sen-
der the combined effects of the melting of a 3-
sor system performance.
cm-deep snow cover over the previous two days,
An example of the sensor system response to
rainfall on this day, and the surface thawing of
a walking person over the course of a winter is
the soil. The near-surface soil warmed to above
20C at midday, and thawing (soil temperatures
given in Figure A-1. On each day shown, the
of 1 to 0C) had penetrated to 22.5 cm depth.
person crossed the zone defined by the cables at
15 to 18 locations along the 141-m lengths of the
Some of the crossings on 28 March were only
cables (Peck 1992). The resultant processor volt-
marginally detected.
ages define the shaded band. In November 1990,
By 29 April the soil was completely thawed
(temperatures of 12 to 6C at depths of 7.5 to 60
the system response to some crossings was un-
acceptably low, either negative or less than 0.4
cm) and there were both marginally detected
V. Soil temperatures during this month were
and undetected crossings. In response to this un-
above freezing, as determined with a vertical ar-
acceptable detection capability, the alarm thresh-
ray of copper-constantan thermocouples in the
old of the system was lowered. The effect of this
soil at depths of 0 to 60 cm. By 2 January 1991,
was to change the equivalency between sensor
when the soil was frozen to a depth of at least
signal and processor-produced voltage, such
22.5 cm (temperature of the 30-cm-deep soil was
that a smaller sensor signal now caused a posi-
0C), all crossings produced a voltage greater
tive voltage. With this change in system sensitiv-
than 0.4 V. The contrast between non- or mar-
ity, acceptable detection capability was restored.
ginal detection in November and reliable detec-
Had the sensor system been operating in soil
tion from 2 January through 13 March is consis-
having a thermal history similar to Figure 8,
tent with differences in the unfrozen moisture
then daily repetitions of the crossings would
content of the soil. The soil froze in January to
have revealed more variable sensor performance
between 37.5- and 52.5-cm depth and in late Feb-
in early winter when the soil above the sensor
ruary to between 52.5 and 60 cm; it remained so
cables was still experiencing short-term freeze
on 13 March. This means that the soil between
thawfreeze episodes.
the sensor cables had a low electrical conductiv-
22
Previous Page
