sion location and at an undisturbed (control) lo-
ries obtained from BC1 simulations highlights
cation indicate that at 7.5-cm depth, the soil tem-
the difficulty of reliably knowing either the fro-
perature is slightly colder, by 0 to 0.3C, when it
zenunfrozen status of soil at depth or when rel-
overlies the center of the 16-cm-wide sand inclu-
ative changes in unfrozen moisture content of
sion, and that soil at 37.5-cm depth is slightly
frozen soil occur, unless soil temperature is mon-
warmer, 0 to 0.2C, when it underlies the center
itored directly by making real-time measure-
of the same sand inclusion. The measurements of
ments or indirectly from numerical simulations.
soil temperatures at the inclusion location were
Otherwise, the occurrence of changes in electro-
made weekly. The magnitudes of the differences
magnetic sensor system effectiveness that result
in soil temperature are small, but the differences
from short-term variations in temperature-
can be quite large (1530%) relative to soil tem-
dependent unfrozen soil moisture may be unrec-
peratures when no inclusion is present.
ognized.
The differences in soil temperature and in
frost depth in response to a change in soil mois-
SUMMARY AND CONCLUSIONS
ture content or a change in boundary condition
For a silty soil of variable wetness (10, 17, 25%
(Warmest, Warm, Cold) demonstrate the poten-
moisture content by weight) subjected to freezing
tial for variability in sensor system performance
conditions ranging from mild (BC-Warmest) to
from year to year. All other things being equal,
severe (BC-Cold), numerical simulations have
an anomalous moisture content at the beginning
quantified the disruption in frost and thaw pene-
of winter, or winter weather of greater or lesser
tration caused by the presence of a sand inclu-
severity, can result in such different soil temper-
sion in the soil. For the narrowest inclusion con-
atures and frost depths that the likelihood and
sidered, 16.5 cm, two-dimensional simulations
duration of improved sensor system perform-
show that frost penetration beneath a surface
ance, related to freezing of the soil, may be
point centered on the inclusion is insignificantly
affected.
different from the case of no inclusion. Heat
The presence of a sand inclusion potentially
effectively flows around the sides of an inclusion
improves electromagnetic sensor system perfor-
this narrow, resulting in small temperature dif-
mance in certain situations. If frost depth is less
ferences relative to the no-inclusion case. When
than cable depth (BC-Warmest simulations),
the inclusion is wider, 91 cm, frost penetration
then the thickness of the layer of unfrozen, lossy
below the inclusion center is effectively that asso-
soil above the cables is reduced by an amount
ciated with an infinitely wide inclusion (one-
equal to the half-thickness of the sand inclusion.
dimensional simulations). For the cases of the 91-
If frost depth exceeds cable depth, but the frozen
cm-wide and infinite-width inclusions, both the
soil overlying the cables normally (no inclusion)
rate and depth of frost penetration are disrupted
is not very cold (BC-Warm simulations), then the
by the presence of the sand inclusion. The low
fact that the soil is colder when a sand inclusion
latent heat content of the dry (3% moisture con-
is present may mean that the soil's unfrozen
tent) sand causes it to freeze more rapidly than
moisture content is appreciably less. However, if
the surrounding soil. Thereafter, however, the
normally the soil overlying the cables is much
low conductivity sand acts as insulation, causing
colder (BC-Cold simulations), the difference in
the soil above it to be colder and the soil below it
soil temperature that results from the presence of
to be warmer.
the sand inclusion will not significantly affect
The dependence of frost penetration on soil
sensor system performance. Finally, if the soil
surface temperature reaches or crosses 0C dur-
moisture content demonstrates the potential for
inconsistent performance by an electromagnetic
ing the winter (other than at initial freezeup and
sensor system within its detection zone. If the
final thaw), such as late in the BC-Cold winter or
moisture content of the soil surrounding a buried
throughout the BC1 winter, the sand inclusion
sensor system varies along its cable length, frost
does not prevent 7.5-cm-deep soil from warming
to 0C or higher. Its presence, however, does re-
penetration will not be uniform within the detec-
tion zone. There will be significant location-
sult in complete thaw of the soil/sand/soil sec-
dependent differences in sensor performance
tion at the end of winter occurring earlier than
that persist until the overlying soil throughout
does complete thaw of just the soil section (no
the entire detection zone is frozen.
sand inclusion); this is jointly because the soil
The complexity of the soil temperature histo-
underlying the sand inclusion is already warmer
19
Previous Page
