Geoscience Reference
In-Depth Information
followed common seismic practice and stacked the
resulting seismoelectric signals, initially separately for
the four locations, and then for all seismoelectric signals
from all locations. This step is justified because spatial
changes in surface topography of, the horizontal stratifi-
cation of the sandstone units within the vadose zone, and
the depth of the water table were all negligible within the
study area. The resulting noise-minimized and stacked
seismoelectric signal (hereafter
which the available volumetric water contents are inte-
grated (Figure 7.5). We also removed the data from the
first 1.25 m below the ground surface since they were
unreliable given the assumption of straight seismic-ray
travel (see preceding text). Comparison of the prepro-
cessed seismoelectric signal (Figure 7.9a) with the volu-
metric water content-depth profile (Figure 7.9b) readily
reveals an inverse relationship. For instance, the seismo-
electric amplitudes increase over the first ~1
“
preprocessed seismo-
4 m as vol-
umetric water content decreases, followed by a series of
predominantly inverse major fluctuations from ~4 to 9
m, and, respectively, the final decrease or increase in
seismoelectric amplitudes and volumetric water contents
between ~9 and 11 m (Figure 7.9).
To quantify these observations, we picked the principal
peak-to-peak amplitudes of the preprocessed seismoelec-
tric signal (Figure 7.8) and the water content profile
(Figure 7.5). A cross-plot of these changes (Figure 7.10)
confirms a statistically significant (
R
2
= 0.81) linear rela-
tionship between these amplitudes. Note that the least-
squares regression line was forced through the origin
assuming that seismoelectric energy cannot be generated
in the absence of water content changes. Allowing for
some natural and statistical variability, we can therefore
infer that seismoelectric energy returns are produced in
close vertical succession within the vadose zone. We
emphasize that this relationship is robust within the com-
bined study areas of West and Truss (2006) and that
reported here, which were colocated within some 50m
-
electric signal
) (Figure 7.8) is characterized by only a
minor energy return from the water table, which we
expect at ~17.5 ms one-way seismic travel time (marked
by an arrow in Figure 7.8). This expectation is based on
the travel time-depth conversion using the VSP data
(Figure 7.7). Indeed, while there was little seismoelectric
energy received from the saturated zone (>17.5 ms in
Figure 7.8), we observe that strong seismoelectric energy
returns are received from the vadose zone. These returns
are characterized by shorter-term fluctuations, typically
several milliseconds long, superimposed on a longer-
term positive background trend (<17.5 ms in Figure 7.8).
”
7.2.3.2 Seismoelectric
water content relationship
Along with the transient pressure disturbances forced by
a downgoing seismic wave, seismoelectric signals are
sensitive to energy loss due to spherical spreading. We
have therefore spherically corrected the preprocessed
seismoelectric signal and smoothed it with a 0.25 m run-
ning mean that corresponds to the depth interval over
-
2
1
0
-1
-2
-3
-4
-5
0
5
10
15
20
25
30
35
40
45
50
Time (ms)
Figure 7.8
Processed seismoelectric data revealing seismoelectric conversions in the vadose zone. The arrow indicates the position
of the groundwater table at ~17.5 ms.
