Geoscience Reference
In-Depth Information
observed differences to surface scattering, lateral variations in the depth dependence of the water
content, or other spatial variability in the water content. This was demonstrated with numerical
modeling that showed the GPR measured water content can be strongly influenced by stratification
of the water content distribution.
24.5 SCAtteRInG fRoM SURfACe RoUGhneSS
As discussed, surface scattering has also been assumed to be a contributor to the observed dif-
ferences between water content measured with TDR and the GPR surface reflectivity method.
Numerical modeling was performed to demonstrate the importance of this effect. Modeling soft-
ware (GPRMAX2D) developed by Giannopoulos (1997) was used to compute the model response.
This software uses a finite difference time domain (FDTD) two-dimensional (2D) algorithm. The
standard Ricker wavelet was used to describe the source pulse shape. In this 2D model, all struc-
tures extend an infinite length perpendicular to the model plane. An example model (Figure 24.5a)
was used to simulate the response for the data example presented in Figure 24.3, the profile from
the grass surface onto the corn field. In the model, the grass is perfectly flat and the corn rows are
modeled as half cylinders (diameter 0.3 m) spaced every 76 cm, the spacing of the corn rows. The
modeled media has a water content of 40 percent corresponding to a relative dielectric permittiv-
ity of 25. The modeled GPR system is elevated 1.3 m above the surface, and the pulse has a center
frequency of 450 MHz.
The results (Figure 24.5b) for this simple model demonstrate that the scattering effects in the
corn field are substantial and consistent with the variability observed in the water content measure-
ments in the field example (Figure 24.3). This example presents one of the worst-case effects of
scattering. The results of field measurements for a grass field (Figure 24.4) are more typical of the
type of variation that can be attributed to scattering.
Tx
Rx
Soil W. C. = 40% Kr = 25
(a)
50
40
30
20
10
0
(b)
-10
0
1
2
3
Position (m)
4
5
6
fIGURe 24.5
Two-dimensional model (a) for the case of a grass field abutting a cornfield and modeled
water content measurements (b). The water content of the modeled media is 40 percent, giving a relative
dielectric permittivity of 25. The water content at 2 m is not shown because the high reflection coefficient from
scattering at this position in the model gives unrealistic values.
