Full text of DOI:10.1046/j.1526-0984.2000.74001.x

K I M E T A L . : G P R H Y D R O C A R B O N D E T E C T I O N
169
Residual Hydrocarbons in a Water-Saturated
Medium: A Detection Strategy Using Ground
Penetrating Radar
CHANGRYOL KIM, JEFFREY J. DANIELS, ERICH D. GUY,
STANLEY J. RADZEVICIUS, and JENNIFER HOLT
Department of Geological Sciences, The Ohio State University, 275 Mendenhall Lab, 125 South Oval Mall, Columbus, OH 43210
sites where the conditions are similar to those modeled through
this experimentation.
ABSTRACT
•
The focus of this article is to demonstrate through physical model
experimentation a potential means for identifying contaminated
areas where a light non-aqueous phase liquid (LNAPL) hydrocar-
bon has been redistributed by a rising water table in a previously
hydrocarbon residual–free vadose zone using ground-penetrating
radar (GPR). Analogies of the experimentation conducted in this
study are situations where a rise of the water table follows leakage
from a tank or pipe at depth or where an LNAPL hydrocarbon
plume has migrated laterally from a surface source along the top
of the saturated zone and a subsequent rise of the water table oc-
curs. Research to date has provided insight into mechanisms that
may offer the potential for LNAPL detection under certain field
conditions; however, no studies have specifically focused on de-
veloping a potential detection strategy for a case in which residual
hydrocarbon is present in a water-saturated medium.
Key Words: contaminant detection, ground-penetrating radar,
hydrocarbon, light non-aqueous phase liquid, vadose zone.
INTRODUCTION
•
To properly remediate contaminated sites of environmen-
tal concern, it is necessary to have knowledge of the actual
extent of contamination in the subsurface. A considerable
amount of research in recent years has focused on the via-
bility of using different geophysical methods to delineate
and monitor areas of subsurface contamination. Ground-
penetrating radar (GPR) is a technique that has been widely
applied for mapping shallow stratigraphy and fractures that
influence contaminant transport (Davis and Annan, 1989;
Benson, 1995; Martinez et al., 1999) and for aiding in the
characterization of contaminated sites by locating buried
features of interest (Daniels et al., 1998; Guy et al., 1999;
Radzevicius et al., 2000). Although GPR has been shown to
have the potential in certain field conditions for delineating
the extent of light non-aqueous phase liquid (LNAPL) hy-
drocarbons (Olhoeft, 1986; Daniels et al., 1995; Grumman
and Daniels, 1995), further experimental research that fo-
cuses on strategies for detection in additional situations is
necessary, as currently the potential usefulness of GPR prior
to a survey is difficult to assess.
A tank model filled with gravel and sand was designed to allow
GPR measurements to be made on the surface before, during,
and after water and gasoline injections and fluctuations within
the tank. Background GPR measurements were made initially
with only water being raised and lowered in the model, and the
water table was then raised and lowered beneath a volume of
219 liters of gasoline that was injected into the bottom of the
tank. Measurements from the initial raising and lowering of the
water with no gasoline present demonstrate the sensitivity of
GPR for monitoring changes in subsurface water content and
minor fluctuations of the water table. Measurements made dur-
ing the raising and lowering of the water table with gasoline in
the model show differences from the measurements made when
only water was raised and lowered, and a comparison of the data
show that reflections in GPR data can be enhanced when resid-
ual gasoline is present in a water-saturated system because there
is less attenuation of the radar signal. Differences in travel times
to subsurface reflections between the two stages of the experi-
ment are also caused by the residual gasoline present in the wa-
ter-saturated medium. Results of this study provide the basis for
a strategy that has the potential for successful detection and de-
lineation of LNAPL hydrocarbon–contaminated areas at field
Analogies can be made between GPR that utilizes electro-
magnetic energy and the mechanical energy–based seismic
reflection technique. However, with GPR, scattering occurs
at impedance contrasts produced by changes in electrical
properties rather than changes in elastic properties or den-
sity. The electrical properties that determine radar velocity,
attenuation, and scattering from discontinuities are electri-
cal conductivity, dielectric permittivity, and magnetic per-
meability. GPR works well in low loss materials (such as
sand and gravel), in which displacement currents dominate
and attenuation associated with conduction is low (Davis
and Annan, 1989), but performs poorly in high conductivity
materials (such as clays with a high cation exchange capac-
© 2000, AAPG/DEG, 1075-9565/00/$15.00/0
Environmental Geosciences, Volume 7, Number 4, December 2000 169–176

E N V I R O N M E N T A L G E O S C I E N C E S
170
ity). In low loss, nonmagnetic mediums, the reflection and
transmission coefficients at normal incidence are governed
primarily by the relative permittivity (␧r) of the materials
through which the energy is propagating. Few earth materi-
als (unless they are ferrous) have a relative magnetic perme-
ability that deviates significantly from unity; thus, at high
frequencies and/or low conductivity, the GPR velocity can
be assumed to vary inversely with the square root of the rel-
ative permittivity. Therefore, higher permittivity materials
result in slower radar velocities.
ing proposed that explain how an increase in bulk conduc-
tivity can result from microbial degradation of hydrocarbon
plumes with time, which attenuates the radar energy–pro-
ducing areas of reduced signal (Nash et al., 1997; Lucius,
2000).
Physical model and laboratory experiments have provided
information relevant to GPR detection strategies for various
LNAPL contaminants (Kutrubes, 1986; Lucius et al., 1990;
Monier-Williams, 1995), and theoretical models involving
LNAPLs have identified various circumstances under which
GPR is capable of serving as a useful tool to detect and
monitor these types of contaminants (Endres and Redman,
1993; Barber and Morey, 1994; Powers and Olhoeft, 1996).
Previous field studies also have provided important infor-
mation in regards to GPR detection strategies of LNAPL
hydrocarbons that originate from spills at the surface and
subsequently migrate downward through the vadose zone.
However, no work to date has concentrated on GPR experi-
mentation involving an LNAPL plume introduced at depth
(with no residual hydrocarbon initially directly above the
LNAPL plume) or on the GPR response over a controlled
situation in which redistributed hydrocarbon exists in the
pore space of a water-saturated medium. The focus of this
article is to address this type of situation that often exists at
contaminated field sites, through a physical model experi-
ment, and to provide a potential detection strategy that is
shown to be successful under controlled conditions.
The potential for GPR to detect LNAPL hydrocarbons
such as gasoline can exist under certain conditions because
the electrical properties of such contaminants can be quite
different than water. When gasoline and air are present in
the pore space of dry quartz sand, for instance (␧r air ϭ 1, ␧
r
gasoline ϭ 2–3, ␧r dry quartz sand ϭ 5), there is not a sig-
nificant contrast between these fluids or materials to allow
detection of the gasoline. However, in a case where gasoline
exists in a system with water present (␧r water = 81), a sig-
nificant and detectable contrast can exist in areas where gas-
oline displaces water relative to areas where no water has
been displaced. Air, gasoline, and dry quartz sand are rela-
tively high velocity mediums for propagating radar energy
when compared to water, which is a relatively slow me-
dium. A detectable decrease in radar wave attenuation
should result in areas where LNAPL hydrocarbons (electri-
cal insulators) displace water, which is a more conductive
fluid, although this can be time dependent as in some cases
hydrocarbon-contaminated soils have been found to de-
velop increased conductivity with time (Sauck, 1998).
LNAPL hydrocarbon detection strategies using GPR to
date have focused on identifying anomalous responses asso-
ciated with the capillary fringe and the vadose zone above
the capillary fringe. Approaches have looked at changes in
reflection arrival times, reflection strengths, and changes in
signal character associated with areas of contamination. Re-
flection pull-ups have been demonstrated to be possible in
areas where hydrocarbon pools develop or where gasoline
displaces residual soil moisture during its descent through
the vadose zone (Douglas et al., 1992; DeRyck et al., 1993).
Reflection pull-downs can also result from depression of the
capillary fringe in areas where contaminant pools develop
due to increased travel distance; however, this effect can be
counter-balanced by a relatively faster velocity of the con-
taminant so that no net change in travel time results (Camp-
bell et al., 1996). Bright spots have been observed through
experimental work and have been attributed to LNAPLs dis-
placing water in the vadose zone, which can cause a higher
permittivity contrast at reflecting interfaces than would ex-
ist without the contaminant presence (Redman et al., 1994;
Campbell et al., 1996). Decreases in reflection amplitudes
over areas of vadose zone hydrocarbon contamination have
been reported by several researchers, with mechanisms be-
EXPERIMENTAL METHODOLOGY
•
The physical model used for experimentation consisted of
a polyethylene cylindrical tank (242-cm diameter; 121-cm
height) filled with pea gravel and sand. The tank model con-
figuration is illustrated in Figure 1 as cross-sectional and
plan views. A port in the center of the bottom of the tank
was installed to allow for the introduction and draining of
water and gasoline. Liquids were injected under a low hy-
draulic gradient using a standard water pump. Feeder tanks
adjacent to the tank model were used for filling and also
waste disposal purposes (Figure 2). A monitoring well was
installed inside the tank close to the wall so that liquid lev-
els in the tank could be monitored. Gravel (0.64- to 0.95-cm
grain diameter) filled the bottom of the tank to a thickness
of 16.5 cm and was overlain by 90 cm of quartz sand (0.05-
to 0.1-cm grain diameter). The interface between these two
layers had a slight dip to allow it to be easily distinguishable
from other horizontal events (such as antenna ring or bot-
tom of tank multiples) in the data throughout the course of
experimentation.
Laboratory measurements yielded average porosity val-
ues for the gravel and sand of 36 and 31% respectively, and
a capillary rise of water in the sand was measured to be 11.4
cm. The gravel and sand were dry when placed in the tank,
and both were assumed to be homogeneous and isotropic

K I M E T A L . : G P R H Y D R O C A R B O N D E T E C T I O N
171
FIGURE 1: Configuration of model tank
and GPR survey grid. (A) Cross-sectional
view. (B) Plan view. Gravel layer is 16. 5 cm
thick and sand layer is 90 cm thick.
layers. A GPR grid consisting of 17 parallel lines each 121
cm long and 7.62 cm apart was situated in the center of the
tank over the sand medium, as shown in Figure 1B. The
GPR recording system used was a GSSI SIR-10A (Geo-
physical Survey Systems, Inc., North Salem, NH), and the
antenna was a fixed offset, bistatic antenna with a center
frequency of 500 MHz (in air). The antenna was positioned
at the tank surface using a track so that precise and repeat-
able measurements could be made along the grid lines (Fig-
ure 3). The only post-acquisition processing applied to the
data were band-pass filtering to remove low frequency drift
(“dewow”) and a linear gain. The same processing parame-
ters were used for all data acquired so that relative ampli-
tude information between data acquired at different stages
of the experimentation would be preserved.
These measurements provided a baseline for comparison of
measurements made during the next phase of the experi-
ment when gasoline was introduced into the model. This
initial raising and lowering of the water table also served the
purpose of residually wetting the sand in the model to better
simulate realistic field conditions for when the gasoline was
introduced. The water table was initially raised to a maximum
level of 66 cm from the bottom of the tank in three stages,
and GPR measurements were made at each of these stages
with water levels of 25.4, 35.6, and 66 cm (Figure 4). After
measurements at the highest water level (66 cm) were made,
the water was drained from the tank, leaving residual satura-
tion of water in the sand. The water level was lowered from
its maximum in stages also, and GPR measurements were
made at water levels of 30.5, 13.8, and 0 cm during draining
(Figure 4).
Background GPR measurements were made prior to any
addition of water into the tank and during an initial raising
and lowering of the water table with no gasoline present.
Prior to the addition of gasoline to the tank, there was
ϳ7.6 cm of water at the bottom of the tank within the gravel
FIGURE 3: Tank model and GPR data acquisition setup. Measurements were made
using an antenna positioned at the model surface guided by a track to assure accuracy
along the grid lines.
FIGURE 2: Experimental setup showing tank model (background left) and one of
the feeder tanks (foreground right) used for filling and waste disposal purposes.

E N V I R O N M E N T A L G E O S C I E N C E S
172
FIGURE 4: Two-dimensional (2D) profiles at different water levels showing effects on the GPR response of raising and lowering the model water table (no gasoline present in
model).
layer. This water was the result of pore water exceeding re-
sidual saturation in the sand and gravel that migrated down-
ward to the bottom of the tank during the 16 hr after drain-
ing the tank of water. This water remained at the time the
gasoline was injected so that a spill from a tank or pipe orig-
inating at the water table and vadose zone interface or a lat-
eral migration of contaminant along this interface would be
best simulated. A volume of 219 liters of gasoline was in-
jected into the model. The gasoline moved upward through
the water and spread within the permeable gravel and ini-
tially raised the total liquid level to 22.9 cm. Water then was
introduced beneath the gasoline to simulate a rise of the wa-
ter table, raising the total liquid level in stages up to 73.7
cm. The water and gasoline were then drained from the tank
in stages. GPR measurements were made over the grid prior
to the start of the injection with 7.6 cm of water present dur-
ing each stage of raising of the water table in the tank with
liquid levels of 22.9, 53.3, and 73.7 cm and at each stage of
drainage with liquid levels of 35.6, 7.6, and 0 cm (Figure 5).
RESULTS AND DISCUSSION
•
Although 17 survey lines were recorded on the surface of
the tank for each stage of raising and lowering of the water
table in the experiment, only data from GPR survey line 2
are presented in this article. The results did not vary signifi-
cantly from line to line as data exhibited similar trends;
however, the amount of diffracted energy from tank walls
FIGURE 5: Two-dimensional (2D) GPR profiles at different liquid levels (water and gasoline present) showing effects of injecting gasoline into the model and then raising and
lowering the water table. ns, nanoseconds.

K I M E T A L . : G P R H Y D R O C A R B O N D E T E C T I O N
173
varied somewhat and was more apparent in some of the
other lines that are not presented.
total liquid level to 22.9 cm (Figure 5). The water table was
then raised in stages using a low hydraulic gradient, pushing
the gasoline upwards and leaving some residual gasoline
beneath the water table. It has been reported that the resid-
ual saturation of hydrocarbons in a water-saturated zone un-
der similar circumstances is ϳ14% of the total pore volume
(Wilson and Conrad, 1984). A capillary fringe (water mixed
with gasoline) formed above the water table in the sand be-
neath the gasoline. The water saturation of this capillary
fringe was Ͻ100%, and the gasoline saturation was greater
than was the residual saturation of hydrocarbons beneath
the water table. The reflection from the top of the capillary
fringe beneath the gasoline throughout this stage of the ex-
periment was detected by GPR as the geophysical water ta-
ble in a manner similar to the initial raising and lowering of
the water table with no gasoline in the tank. Also, the polar-
ities of the reflections from the interface between the sand
and gravel show the same pattern as did those during the
initial (water only) stage of the experiment (Figure 4). A no-
ticeable reflection did not occur from the gasoline layer on
top of the capillary fringe because the relative permittivity
difference between the gasoline and the overlying sand
(with only residual water and/or gasoline) did not provide a
detectable contrast.
GPR Response during Initial Water
Table Fluctuation
GPR two-dimensional profiles at each level of the raising
and lowering of the water table are presented in Figure 4.
Reflections from the water saturation zone on the two-dimen-
sional profile sections are from the top of capillary fringe
above the water table. The difference in relative permittivity
between the water table and capillary fringe interface was too
small for GPR to detect. Therefore, the capillary fringe was
imaged as the geophysical water table and not the water ta-
ble itself. These background measurements made during the
raising and lowering of the water table demonstrate the sen-
sitivity of GPR for detecting small changes in the geophysi-
cal water table, although the ability of GPR to detect the wa-
ter table depends on the frequency used and the material
grain size that influences capillary height.
A reflection of interest on the profile sections in Figure 4
is the interface between the gravel and sand layers. When
the water table was raised, this reflection became quite faint
due to attenuation of the GPR signal resulting from conduc-
tion losses associated with the water. The water that was in-
jected into the tank had a measured conductivity of 27 milli-
Siemens/m. The gravel and sand reflection shows the same
polarity as that of incident wave-form when the boundary
between these layers was above the water table (water lev-
els of 0 and 13.8 cm). This is due to the pore spaces being
filled with air, a fast fluid, which results in the relative per-
mittivity of the sand being greater than that of the gravel be-
cause it has a lower porosity, and therefore it is a slower me-
dium. Conversely, the reflection from this interface shows a
polarity reversal when it is submerged beneath the water ta-
ble (water levels of 25.4, 35.6, 66, and 30.5 cm) because in
these cases the relative permittivity of the sand is lower than
that of the gravel because water is a slow fluid.
Comparison of Data Prior to and Subsequent to
Gasoline Injection
The reflections in Figure 5 from the gravel and sand
boundary show stronger amplitudes when residual gasoline
is present above this interface than do reflections from the
same interface in Figure 4 when only water is present above
this interface. Figure 6 shows side by side the two-dimen-
sional sections at maximum water levels from Figures 4 and
5 and additionally shows representative one-dimensional traces
(traces at 60 cm distance on two-dimensional sections) from
these sections. The noticeable reflection differences on the
two-dimensional sections and the one-dimensional traces be-
tween the measurements involving gasoline versus those in-
volving only water were calculated using the one-dimen-
sional traces. The reflection amplitude with gasoline present
was approximately three times greater than without gasoline
present, which is equivalent to an increase of 10 dB in re-
flection power. Although there were differences in this value
depending on which trace and line number were used for
calculation, measurements still exhibited similar trends, and
this calculation is representative of all measurements. This
difference in reflection power can be attributed to the pres-
ence of the residual gasoline (gasoline is an electrical insu-
lator) that resulted in less attenuation of the propagated en-
ergy than when the medium was entirely saturated with water.
In addition to the enhanced reflection strength of the
gravel and sand interface, the two-way travel time to this in-
terface is decreased when gasoline is present in the pore
The two-way travel time of the reflection from the geo-
physical water table during drainage of the tank at a water
level of 30.5 cm is somewhat greater than that of the geo-
physical water table reflection during the raising of the wa-
ter table at a water level of 25.4 cm, even though the two-
way travel distance to this reflector is less when the water
level is at 30.5 cm. This anomalous effect is due to the re-
sidual saturation of water that remained after the peak level
of water in the tank, which resulted in a slower velocity
above the capillary fringe and thus a longer two-way travel
time to the geophysical water table at 30.5 cm.
GPR Response during Water Table Fluctuation
with Gasoline Injection
After the initial raising and lowering of the water table in
the tank, 219 liters of gasoline were injected and raised the

E N V I R O N M E N T A L G E O S C I E N C E S
174
FIGURE 6: Comparison of maximum wa-
ter level two-dimensional (2D) sections and
one-dimensional (1D) traces (traces from 60
cm distance on two-dimensional section)
prior to and subsequent to gasoline injection
and water table fluctuation in the model.
spaces of the water-saturated sand. When comparing the ar-
rival times of the gravel and sand reflection in Figure 6 from
measurements made with and without gasoline, it can be
seen that the reflection with gasoline arrives ϳ3 ns faster.
This observed effect makes sense, as the relative permittiv-
ity of the water-saturated zone without gasoline above the
gravel and sand interface is more than that in the case where
residual gasoline (assumed to be ϳ14%) occupies pore
spaces in the water-saturated zone. Because the sand me-
dium has a lower relative permittivity when residual gaso-
line is present in the pore spaces, it has a faster velocity;
therefore, a decreased two-way travel time to reflections be-
neath this medium resulted.
may result when compared to adjacent areas where no resid-
ual hydrocarbon contamination within the water-saturated
zone exists. Accurate detection of hydrocarbon contami-
nants using GPR is influenced by many factors and has to
date often proven to be a tough task in the field. There is
much that remains to be understood regarding LNAPL hy-
drocarbon detection using GPR. Experimental results from
this study provide the basis for a strategy that in theory is
well-founded and has been demonstrated to be capable of
detection under controlled conditions.
ACKNOWLEDGMENTS
•
The authors thank the U.S. Environmental Protection
Agency, Region V for supporting this research. Gratitude is
extended to Mark Vendl of the U.S. Environmental Protec-
tion Agency, Region V for his assistance. Comments pro-
vided by Conrad Allen of ExxonMobil Corporation, Alex
Martinez, and Mike Nash are appreciated.
CONCLUSIONS
•
Results from the physical tank model experiment show
the applicability of GPR for monitoring minor fluctuations of
the water table and capillary fringe. It has been demonstrated
that the reflection strength of the gravel and sand interface
decreased significantly as a result of increased attenuation
losses when the water table was raised above this boundary
in the initial stage of the experiment involving only water.
However, the power of this reflection after injecting gasoline
into the model and raising the water table was much greater
(10 dB) as a result of decreased attenuation through the me-
dium. Residual gasoline left behind in the pore space as a re-
sult of raising the water table also resulted in a velocity pull-
up (3 ns) of the gravel and sand interface reflection when
compared with data acquired prior to the gasoline injection.
These experimental results have demonstrated and ex-
plained a mechanism that can potentially allow for GPR to
serve as a useful tool for detecting areas of contamination in
situations where zones of residual gasoline are present in a
water-saturated medium. Less attenuation and faster two-
way travel times to reflections beneath the saturated zone
REFERENCES
•
Barber, W. B., and Morey, R. (1994). Radar detection of thin lay-
ers of hydrocarbon contamination. In Proceedings of the Inter-
national Conference on Ground Penetrating Radar (pp. 1215–
1228). Kitchener, Ontario, Canada.
Benson, A. K. (1995). Applications of ground penetrating radar in
assessing some geological hazards: Examples of groundwater
contamination, faults, cavities. J Appl Geophys, 33, 177–193.
Campbell, D. L., Lucius, J. E., Ellefsen, K. J., and Deszcz-Pan, M.
(1996). Monitoring of a controlled LNAPL spill using ground-
penetrating radar. In Proceedings of the Symposium on the Ap-
plication of Geophysics to Environmental and Engineering
Problems (pp. 511–517). Keystone, CO.
Daniels, J. J., Roberts, R. L., and Vendl, M. A. (1995). Ground
penetrating radar for detection of liquid contaminants. J Appl
Geophys, 32, 195–207.

K I M E T A L . : G P R H Y D R O C A R B O N D E T E C T I O N
175
Daniels, J. J., Brower, J., and Baumgartner, F. (1998). High resolu-
tion GPR at Brookhaven National Laboratory to delineate com-
plex subsurface targets. J Environ Eng Geophys, 3, 1–5.
Davis, J. L., and Annan, A. P. (1989). Ground-penetrating radar
for high-resolution mapping of soil and rock stratigraphy. Geo-
phys Prospect, 37, 531–552.
posit. In Society of Exploration Geophysicists Annual Meeting
Expanded Abstracts (pp. 582–585). Society of Exploration Geo-
physicists, Tulsa, OK.
Monier-Williams, M. (1995). Properties of light non-aqueous
phase liquids and detection using commonly applied shallow
sensing geophysical techniques. In Proceedings of the Sympo-
sium on the Application of Geophysics to Environmental and
Engineering Problems (pp. 1–13). Orlando, FL.
DeRyck, S. M., Redman, J. D., and Annan, A. P. (1993). Geophys-
ical monitoring of a controlled kerosene spill. In Proceedings of
the Symposium on the Application of Geophysics to Environ-
mental and Engineering Problems (pp. 5–19). San Diego, CA.
Douglas, D. G., Burns, A. A., Rino, C. L., and Maresca, J. W.
(1992). A study to determine the feasibility of using a ground
penetrating radar for more effective remediation of subsurface
contamination. Cincinnati, OH: U. S. Environmental Protection
Agency Risk Reduction Engineering Laboratory Office of Re-
search and Development.
Nash, M. S., Atekwana, E., and Sauck, W. A. (1997). Geophysical
investigation of anomalous conductivity at a contaminated site.
In Proceedings of the Symposium on the Application of Geo-
physics to Environmental and Engineering Problems (pp. 675–
683). Reno, NV.
Olhoeft, G. R. (1986). Direct detection of hydrocarbon and organic
chemicals with ground penetrating radar and complex resistiv-
ity. In Proceedings of the NWWA/API Conference on Petroleum
Hydrocarbons and Organic Chemicals in Ground Water—Pre-
vention, Detection and Restoration (pp. 284–305). Dublin, OH.
Powers, M. H., and Olhoeft, G. R. (1996). Modeling the GPR re-
sponse of leaking buried pipes. In Proceedings of the Sympo-
sium on the Application of Geophysics to Environmental and
Engineering Problems (pp. 525–534). Keystone, CO.
Radzevicius, S. J., Daniels, J. J., Guy, E. D., and Vendl, M. A.
(2000). Significance of crossed-dipole antennas for high noise
environments. In Proceedings of the Symposium on the Applica-
tion of Geophysics to Environmental and Engineering Problems
(pp. 407–413). Washington, DC.
Endres, A. L., and Redman, D. J. (1993). Modeling the electrical
properties of porous rocks and soils containing immiscible con-
taminants. In Proceedings of the Symposium on the Application
of Geophysics to Environmental and Engineering Problems (pp.
21–38). San Diego, CA.
Grumman, D. J., and Daniels, J. J. (1995). Experiments on the de-
tection of organic contaminants in the vadose zone. J Environ
Eng Geophys, 0, 31–38.
Guy, E. D., Daniels, J. J., Radzevicius, S. J., and Vendl, M. A.
(1999). Demonstration of using crossed dipole GPR antennae
for site characterization. Geophys Res Lett, 26, 3421–3424.
Kutrubes, D. (1986). Dielectric permittivity measurements of soils
saturated with hazardous fluids. Master’s thesis, Colorado
School of Mines. Golden, CO.
Redman, J. D., DeRyck, S. M., and Annan, A. P. (1994). Detection
of LNAPL pools with GPR: Theoretical modelling and surveys
of a contolled spill. In Proceedings of the International Confer-
ence on Ground-Penetrating Radar (pp. 1283–1294). Kitch-
ener, Ontario, Canada.
Lucius, J. E. (2000). Detectability of crude oil in the subsurface
near Bemidji, Minnesota, using ground penetrating radar. In
Proceedings of the Symposium on the Application of Geophysics
to Environmental and Engineering Problems (pp. 311–319).
Washington, DC.
Sauck, W. A. (1998). A conceptual model for the geoelectrical re-
sponse of LNAPL plumes in granular sediments. In Proceedings
of the Symposium on the Application of Geophysics to Environ-
mental and Engineering Problems (pp. 805–817).
Lucius, J. E., Olhoeft, G. R., Hill, P. L., and Duke, S. K. (1990).
Properties and hazards of 108 selected substances. Location: U. S.
Geological Survey Open File Report 90–408.
Wilson, J. L., and Conrad, S. H. (1984). Is physical displacement
of residual hydrocarbons a realistic possibility in aquifer restora-
tion? In Proceedings of the NWWA/API Conference on Petroleum
Hydrocarbons and Organic Chemicals in Ground Water–Pre-
vention, Detection and Restoration (pp. 274–298). Dublin, OH.
Martinez, A., Carr, T., Beaty, D., Byrnes, A. and Stiles, J. (1999).
Abstract: Comparison of ground-penetrating radar response and
rock properties in a sandstone-dominated incised valley-fill de-

E N V I R O N M E N T A L G E O S C I E N C E S
176
ABOUT THE AUTHORS
Changryol Kim
•
Erich D. Guy
Changryol Kim received a B.S. (1986)
in geology from Yonsei University (Seoul),
and an M.S. (1997) in geology from The
Ohio State University and is currently
working on his Ph.D. at The Ohio State
University. His research interests include
LNAPL detection, the effects of seasonal
variations on GPR data, and GPR data pro-
cessing and modeling. He is a member of
EEGS, SEG, and AGU.
Erich Guy received his B.S. (1996) in
geology from Mount Union College and an
M.S. (1998) in geology from Bowling
Green State University, and is currently
working towards his Ph.D. at The Ohio
State University. His current research inter-
ests focus on various aspects of surface and
borehole GPR as well as seismic data pro-
cessing and interpretation for environmen-
tal, geotechnical, and geologic applications. He is a member of
AAPG, AEG, AGU, DEG, EEGS, GSA, and SEG.
Jeffrey J. Daniels
Jeffrey Daniels received his B.S. (1969)
Stanley J. Radzevicius
and an M.S. (1970) in geology from Michi-
gan State University and a Ph.D. (1974) in
geophysical engineering from the Colo-
rado School of Mines. He worked for 11
years for the U. S. Geological Survey in
Denver with primary interests in borehole
geophysics. In 1985, he joined the faculty
of the Department of Geological Sciences
at The Ohio State University. His current areas of research interests
involve using GPR to image the shallow subsurface for engineering
and environmental applications. He is a member of AGU, EEGS,
SEG, and SPWLA.
Stanley Radzevicius received a B.A.
(1992) in earth science, a B.A. (1992) in
mathematics from Saint Cloud State Uni-
versity, and an M.S. (1995) in geophysics
from Indiana University, and he is cur-
rently working on his Ph.D. at The Ohio
State University. His research interests in-
clude LNAPL and DNAPL detection using
geophysics, electromagnetic modeling, the
use of GPR polarization and antenna pattern information to improve
subsurface imaging and interpretation, and various aspects of seis-
mology. He is a member of AAPG, AGU, DEG, EEGS, IWRA,
OGS, and SEG.