19
Advanced Ground-Penetrating Radar for Soil Moisture Retrieval
2.5-ha field within 3 h. The repeatability error was equal to 0.017 m 3 /m 3 and was
mainly attributed to the interpolation uncertainties, as GPR measurements were not
taken exactly at the same locations between the repetitions. The GPR method for soil
moisture sensing appeared highly precise and reproducible owing to the accurate
modeling of the GPR system and the high-quality information that is recorded by
VNA over a large frequency bandwidth.
2.4.2 coMPaRiSon with the diRect gRound-wave Method
The off-ground GPR method was compared with the commonly used ground-wave
method using on-ground GPR (e.g., Galagedara et al. 2003) for soil moisture sensing in field conditions. In a bistatic GPR system (i.e., composed of transmitting and
receiving antennas), the GPR ground wave is the signal traveling directly from a
transmitting to a receiving antenna through the upper centimeters of the soil, and
it is the only wave of which the propagation distance can be known a priori. GPR
ground wave can thus be used for determining soil moisture without knowledge of
soil depth or in the absence of any method that clearly reflects soil interface. Ground
waves can be identified from single trace analysis (STA) acquisitions, where the
transmitting and receiving antennas are separated by a fixed antenna separation
(single-offset GPR). Compared to multioffset GPR methods, the STA approach is
more appropriate for mapping large areas owing to its practicability (Lehmann and
Green 1999).
A 5-ha field near Bastendorf, Luxembourg, was surveyed using the two GPR
methods in September 2010, a few hours after a precipitation event. A pulse radar
combined with a pair of 400-MHz bow-tie antennas was used for ground-wave
acquisition. The dielectric permittivity was derived from the ground-wave velocity using the STA approach. For the off-ground GPR, the dielectric permittivity
was retrieved using inversion of the radar data in the time domain, focusing on the
surface reflection. Dielectric permittivities were then translated in volumetric soil
moisture using Topp’s relationship (Equation 2.5). Volumetric soil moisture was
independently measured by soil core sampling at 27 locations across the field. Figure
2.5 compares the soil moisture maps derived from the off-ground GPR inversion
and ground-wave on-ground GPR method. There was an overall good agreement in
the average soil moisture between the two techniques, but particular soil moisture
patterns appeared different. These discrepancies could be due to the different penetration depths of the two GPR methods. In that respect, the off-ground GPR may
sense the first 5 cm, whereas the ground-wave technique may reflect soil moisture
from larger depths (up to 20 cm). The larger spatial variability of the soil moisture,
which is observed with the off-ground GPR, could be related to its shallow depth of
characterization, as the shallow soil layer is more influenced by varying atmospheric
conditions than the deeper layer. The high soil moisture values that are sensed at
the east of the field by the off-ground method may also originate from the shallower
characterization of the off-ground GPR, as the survey was following a precipitation
event. The soil sampling locations and corresponding volumetric soil moisture values are depicted with circles on the maps.
Advanced Ground-Penetrating Radar for Soil Moisture Retrieval
2.5-ha field within 3 h. The repeatability error was equal to 0.017 m 3 /m 3 and was
mainly attributed to the interpolation uncertainties, as GPR measurements were not
taken exactly at the same locations between the repetitions. The GPR method for soil
moisture sensing appeared highly precise and reproducible owing to the accurate
modeling of the GPR system and the high-quality information that is recorded by
VNA over a large frequency bandwidth.
2.4.2 coMPaRiSon with the diRect gRound-wave Method
The off-ground GPR method was compared with the commonly used ground-wave
method using on-ground GPR (e.g., Galagedara et al. 2003) for soil moisture sensing in field conditions. In a bistatic GPR system (i.e., composed of transmitting and
receiving antennas), the GPR ground wave is the signal traveling directly from a
transmitting to a receiving antenna through the upper centimeters of the soil, and
it is the only wave of which the propagation distance can be known a priori. GPR
ground wave can thus be used for determining soil moisture without knowledge of
soil depth or in the absence of any method that clearly reflects soil interface. Ground
waves can be identified from single trace analysis (STA) acquisitions, where the
transmitting and receiving antennas are separated by a fixed antenna separation
(single-offset GPR). Compared to multioffset GPR methods, the STA approach is
more appropriate for mapping large areas owing to its practicability (Lehmann and
Green 1999).
A 5-ha field near Bastendorf, Luxembourg, was surveyed using the two GPR
methods in September 2010, a few hours after a precipitation event. A pulse radar
combined with a pair of 400-MHz bow-tie antennas was used for ground-wave
acquisition. The dielectric permittivity was derived from the ground-wave velocity using the STA approach. For the off-ground GPR, the dielectric permittivity
was retrieved using inversion of the radar data in the time domain, focusing on the
surface reflection. Dielectric permittivities were then translated in volumetric soil
moisture using Topp’s relationship (Equation 2.5). Volumetric soil moisture was
independently measured by soil core sampling at 27 locations across the field. Figure
2.5 compares the soil moisture maps derived from the off-ground GPR inversion
and ground-wave on-ground GPR method. There was an overall good agreement in
the average soil moisture between the two techniques, but particular soil moisture
patterns appeared different. These discrepancies could be due to the different penetration depths of the two GPR methods. In that respect, the off-ground GPR may
sense the first 5 cm, whereas the ground-wave technique may reflect soil moisture
from larger depths (up to 20 cm). The larger spatial variability of the soil moisture,
which is observed with the off-ground GPR, could be related to its shallow depth of
characterization, as the shallow soil layer is more influenced by varying atmospheric
conditions than the deeper layer. The high soil moisture values that are sensed at
the east of the field by the off-ground method may also originate from the shallower
characterization of the off-ground GPR, as the survey was following a precipitation
event. The soil sampling locations and corresponding volumetric soil moisture values are depicted with circles on the maps.