Water content
tions and two in-rill stations per location. The dis-
I monitored soil water redistribution during
tances were also used to construct rill profiles
each FT cycle using six Hydra probes (Fig. 13).
from which the rill hydraulic radius was deter-
The probes measure a soil's high frequency (50
mined. I originally planned to measure all seven
MHz) complex dielectric constant, which is made
profiles (Fig. 17a) with an automatic acoustic pro-
up of a capacitive and a conductive electrical re-
filer. However, the profiler did not work properly
sponse (Vitel, Inc. 1994). The capacitive part is
across the rill's sloping sides, so I had to measure
most indicative of soil water content, while the
the profiles by hand and only did four.
conductive portion reflects predominantly soil
salinity. As a soil gets wetter, the capacitive re-
RESULTS AND DISCUSSION
sponse increases and, with appropriate calibra-
tion, the dielectric constant measurement is
FT Cycle 1
directly related to soil water content. The probe
also measures soil temperature with a calibrated
Frost formation and thaw
thermistor in the probe head. The temperature is
The panel temperature (Tp) was lowered to 20
to 30C at 1400 hours on Julian day 146 (146.14)
used to remove most of the temperature effects
on soil dielectric constant when those data are
for about 11 hours and was held between 8 to
10C (Fig. 18). The first freeze (F1) began at
converted to water content.
147.15 when Ta reached 0C and ended 233 hours
When we installed the probes, we placed soil
around the tines (Fig. 15) to ensure good soiltine
later when the panels were removed at 157.08.
The soil surface at RG C reached 0C 35 hours
contact, as recommended by the manufacturer.
(149.02) after F1 began (Fig. 18). The 0C isotherm
We dug two holes in the surface of each soil layer
and placed one probe horizontally in each hole.
penetrated to a depth of about 6 in. (15 cm) at RG
The Hydra probes read from 3537% volumetric
C during F1 (Fig. 19), an average of 0.04 in./hr
soil water before the first freeze, which compares
(0.10 cm/hr) (Fig. 20a). The isotherm penetrated
well to that determined from the bulk density
to about 5 in. (13 cm) at RG D (Fig. 19), an average
cores (Table 2). The RGs, thermocouples, and
of 0.03 in./hr (0.08 cm/hr) at RG D (Fig. 20b),
Hydra probes were connected to Campbell data-
while the frostline penetration rate averaged 0.03
loggers (Fig. 16) for data acquisition, storage, and
in./hr at both gages (Fig. 20a,b). The first thaw
relay through the FERF computer data collection
(T1) was complete at 158.23 at RG C and at 158.15
system (Knuth 1989) to my computer.
at RG D (Fig. 14), 39 and 31 hours, respectively,
after the panels were removed.
Surface geometry
The liquid water decreased dramatically in the
I used a millimeter stick to measure the vertical
surface soil as most of the water froze, but only a
distances from an aluminum bar to the soil sur-
small decrease occurred at depths of 7 in. (18 cm)
face before and after each freeze and thaw. I
and greater (Fig. 21) as soil water was drawn to
aligned the stick with a vertical mark on the bar
the freezing front. During F1, the liquid water at
before each reading to ensure that I was measur-
these depths never froze. The similar changes in
ing the true vertical distance to the soil surface
the distribution of soil water diagonally across
(Fig. 17). As a check of accuracy, I measured a dis-
the bin (Fig. 13) suggest that they represent the
tance when the stick was 5 out of vertical. That
soil water movement out of the rill throughout
measurement was 1 mm too long, but the 5 was
the bin. Differences in the heave between the out-
obviously not vertical, so it was easy to hold the
of-rill and in-rill stations suggest that soil water
stick less than 5 out of vertical during each mea-
movement under the rill may have been different.
surement. Thus, a conservative estimate of error
for each measurement was +1 mm.
Rill hydraulic geometry
The bar was bolted in place above the soil at
The soil outside the rill heaved at all 16 mea-
locations 1, 3, 5, and 7 (Fig. 17b). The vertical dis-
surement stations an average of 1.5 cm during F1
tances were measured from stations 5 cm apart
(Table 4), and the difference between maximum
horizontally along the bar, except over the rill,
and minimum heave was 1.4 cm. Soil in the rill
where the stations had to be closer together to
bottom heaved at six of eight stations by an aver-
adequately define the rill's cross-sectional shape.
age of 0.4 cm and the difference between maxi-
Frost heave after each freeze was determined
mum and minimum heave was 1.2 cm, which is
from the vertical distance at four out-of-rill sta-
less than the soil outside the rill. This is some-
13
Previous Page
