Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Department of Geology and Geophysics, Texas A&M University, College Station, TX 77843-3115, United States
Department of Irrigation and Drainage, College of Water Conservancy and Civil Engineering, China Agricultural University, Beijing, 100083, PR China
ChineseIsraeli International Center for Research in Agriculture, China Agricultural University, Beijing, 100083, PR China
Department of Environmental Sciences, University of Houston-Clear Lake, Houston, TX 77058-1098, United States
a r t i c l e
i n f o
Article history:
Received 6 March 2008
Received in revised form 21 April 2009
Accepted 28 April 2009
Available online 7 May 2009
Keywords:
Aquitard diffusion
Aquitard advection
Mass conservation
Solute transport
Laplace transform
a b s t r a c t
This study deals with two-dimensional solute transport in an aquiferaquitard system by
maintaining rigorous mass conservation at the aquiferaquitard interface. Advection,
longitudinal dispersion, and transverse vertical dispersion are considered in the aquifer.
Vertical advection and diffusion are considered in the aquitards. The rst-type and the thirdtype boundary conditions are considered in the aquifer. This study differs from the commonly
used averaged approximation (AA) method that treats the mass ux between the aquifer and
aquitard as an averaged volumetric source/sink term in the governing equation of transport in
the aquifer. Analytical solutions of concentrations in the aquitards and aquifer and mass
transported between the aquifer and upper or lower aquitard are obtained in the Laplace
domain, and are subsequently inverted numerically to yield results in the real time domain (the
Zhan method). The breakthrough curves (BTCs) and distribution proles in the aquifer
obtained in this study are drastically different from those obtained using the AA method.
Comparison of the numerical simulation using the model MT3DMS and the Zhan method
indicates that the numerical result differs from that of the Zhan method for an asymmetric case
when aquitard advections are at the same direction. The AA method overestimates the mass
transported into the upper aquitard when an upward advection exists in the upper aquitard.
The mass transported between the aquifer and the aquitard is sensitive to the aquitard Peclet
number, but less sensitive to the aquitard diffusion coefcient.
2009 Elsevier B.V. All rights reserved.
1. Introduction
Interaction between aquifers and aquitards is an important
process affecting ow and transport in subsurface ow systems.
Most aquitards consist of silt and clay and are well capable of
storing water and solute, due to their large values of porosity.
Thus when a solute in an aquifer contacts a previously solutefree aquitard, a concentration gradient exists across the aquiferaquitard interface; and molecular diffusion will drive the solute
into the aquitard. Furthermore, leakage often exists across the
aquitard, thus advection in the aquitard will be another
important mechanism for solute transport there.
Corresponding author. Tel.: +1 979 862 7961; fax: +1 979 845 6162.
E-mail address: zhan@geo.tamu.edu (H. Zhan).
0169-7722/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jconhyd.2009.04.010
163
the aquitard simultaneously for a fully penetrating, horizontally innite source without using the averaged approximation employed by Sudicky et al. (1985), Chen (1985), Tang
and Aral (1992a,b), and others. Mass balance requirement is
rigorously maintained. An example of such a two-dimensional transport scenario is the leaking of toxic materials that
are buried in a long trench or a large landll site. There is no
doubt that three-dimensional numerical simulations of ow
and transport in complex aquiferaquitard systems can be
carried out with the present-day's computational power. For
instance, Martin and Frind (1998) have carried out a threedimensional numerical simulation of groundwater ow and
capture zone description for a multiple layer aquiferaquitard
system in the Waterloo Moraine. Bester et al. (2005) have
carried out a numerical simulation of road salt impact on an
urban welleld located in an aquiferaquitard system.
The purpose of this paper is to illustrate the essence of the
aquitard effect on solute transport from an analytical perspective. Such analytical solutions may serve the purpose of
validating the numerical simulations which may suffer from
various types of numerical errors. For instance, the numerical
errors in numerical models tend to be the largest at the
aquiferaquitard interfaces (Martin and Frind, 1998; Bester
et al., 2005). For the sake of simplicity of presentation, we
only discuss a conservative solute.
2. Conceptual and mathematical models
2.1. Conceptual model
The system investigated is an aquifer bounded at the top
and bottom by two aquitards, or an aquifer bounded at the top
by an aquitard and at the bottom by impermeable bedrock.
The aquiferaquitard and the bedrockaquifer boundaries are
assumed to be horizontal. The aquifer is homogeneous and
horizontally isotropic with constant longitudinal and transverse vertical dispersivities. The aquitards are also homogeneous and sufciently thick so that solute diffusion is not
affected by their thicknesses. This assumption may be reasonable because the penetration depths of solute into the
aquitards via diffusion are often limited; meaning that only
those regions of the aquitards that are very close to the
aquifer will be affected by solute transport. However, if one is
dealing with a very thin aquitard, it is possible that solute can
penetrate through the entire thickness of the aquitard into the
adjacent aquifer. If this is the case, one must consider the
affected adjacent aquifer as well. Such a circumstance is of
interest but will not be discussed in this article. The aquifer,
the aquitards, and the bedrock extend horizontally to innity.
Fig. 1AB show the schematic diagrams of two cases that
will be investigated. They are an aquifer bounded by upper
and lower aquitards, and an aquifer bounded by an upper
aquitard and lower impermeable bedrock, respectively. The
big arrows there show the groundwater ow directions. The
hydraulic conductivity of the aquitard is assumed to be a few
orders of magnitude smaller than that of the aquifer, thus
advective ow in the aquitard is nearly perpendicular to the
aquiferaquitard interface. Flow in the aquifer is nearly horizontal. We set up the coordinate system as follows. The x- and
z-axes are along the horizontal and vertical directions, respectively with the origin at the left boundary. We choose the
164
AC
A2 C
A2 C
AC
= Dx 2 + Dz 2 v
;
At
Ax
Ax
Az
C x = 0; z; t = C0 ;
C x = + ; z; t = 0;
C x; z; t = 0 = 0; for x N 0;
C1 x; z = B; t = C x; z = B; t ;
1 v1z C1 x; z = B; t 1 D01
= Dz
AC1 x; z = B; t
Az
AC x; z = B; t
;
Az
C1 x; z = + ; t = 0;
C1 x; z; t = 0 = 0;
10
C2 x; z = B; t = C x; z = B; t ;
11
2 v2z C2 x; z = B; t 2 D02
= Dz
AC2 x; z = B; t
Az
12
AC x; z = B; t
;
Az
C2 x; z = ; t = 0;
13
C2 x; z; t = 0 = 0;
14
15
165
A2 C D
A2 C D
AC
+
Pe D ;
2
AxD
AxD
Az2D
16
C D xD = 0; zD ; p = 1 = p;
17
C D xD = + ; zD ; p = 0;
18
A2 C 1D
AC
Pe1 1D ;
AzD
Az2D
19
C 1D xD ; zD = 1; p = C D xD ; zD = 1; p;
20
AC 1D xD ; zD = 1; p
Pe1 C 1D xD ; zD = 1; p
AzD
AC D xD ; zD = 1; p
;
AzD
1 1
21
C 1D xD ; zD = + ; p = 0;
22
A2 C 2D
AC
Pe2 2D ;
AzD
Az2D
23
C 2D xD ; zD = 1; p = C D xD ; zD = 1; p;
"
AC 2D xD ; zD = 1; p
Pe2 C 2D xD ; zD = 1; p
AzD
2 2
=
24
#
25
AC D xD ; zD = 1; p
;
AzD
C 2D xD ; zD = ; p = 0;
26
, C 2D
are the Laplace transforms of CD, C1D, C2D,
where C D , C 1D
respectively, and p is the Laplace transform parameter in
respective to the dimensionless time. All the associated terms
are explained in Table 1.
Table 1
Dimensionless variables used in this study.
q
1 =
xD = Bx DDxz ; zD = Bz ; tD = DB2z t
CD =
C
C0
M1D =
; C1D =
C0 B2
C1
C0
; C2D =
M
1
p
;
Dx = Dz
C2
C0
M2D =
1 =
C0 B2
D01
Dz
1
; 2 =
; 2 =
D02
Dz
2
M
2
p
; Pe1 =
Pe = pvB
Dx = Dz
Dx Dz
v1z B
Dz ; Pe2
v2z B
Dz
166
CD =
An exp4
q
Pe2 + 4 p + 2n Pe
2
n=0
3
xD 5 cosn zD + n ;
27
1 V zD V 1;
C 1D =
An exp
2
q
Pe1 = 1 2 + 4p = 1 Pe1 = 1
n=0
q
Pe2 + 4 p + 2n Pe
zD 1 cosn + n ;
zD z 1:
C 2D =
q
Pe2 + 4 p + 2n Pe
An exp
2
q
Pe2 = 2 2 + 4p = 2 + Pe2 = 2
n=0
28
xD
29
xD
zD + 1 cosn n ;
zD V 1:
4F1 n ; n
;
pF2 n ; n
30
= 1 = p:
31
xD = 0 +
4F1 n ; n
;
pF2 n ; n Gn
32
with negligible diffusion into the bedrock (Fig. 1B), the concentrations in the aquifer and aquitard can be easily obtained
by modifying the solutions derived in Section 2.2. This can be
done as follows. One can use the x axis or (z = 0 line) as a
symmetric line to make an image aquifer and an image lower
aquitard from z = 0 to . The image aquifer and aquitard
have the same hydrological parameters as their counterparts,
respectively. Furthermore, the advective velocities in the
upper aquitard and the lower image aquitard have the same
magnitude but opposite direction. The image aquifer and the
aquifer are now combined into an equivalent hypothetical
aquifer with thickness 2B. The system generated in such a way
is a special case of Section 2.2 as mentioned after Eq. (33).
Therefore, the problem shown in Fig. 1B can be regarded as
the upper half-plane of a hypothetical aquifer of thickness 2B
with identical upper and lower aquitard parameters and the
same magnitude but opposite aquitard advective velocities.
With such a modication, the solutions derived in Section 2.2
can be directly applied for the case of Fig. 1B.
q
1 + 4 p + 2n = Pe2 + 1 = 2:
33
167
1
x Pet
x + Pet
erfc D p D + expxD erfc D p D ;
2 tD
2 tD
2
34
where erfc ( ) is the complementary error function. Eq. (34) is the well-known one-dimensional solution derived by Ogata and
Banks (1961).
For the third-type boundary condition at xD = 0, A0 = 1/[pG(0)], and An = 0 for n 0, where G(0) is given by Eq. (33) by setting
n = 0. It is not difcult to verify that the inverse Laplace transform of the general solution (A4) under this circumstance will
become:
CD =
"
#
!1
1
x Pet
1 Pe2 tD 2
x PetD 2
1
x + Pet
2
exp D
erfc D p D +
1 + PexD + Pe tD expPexD erfc D p D
2 tD
4tD
2 tD
2
2
4
35
Eq. (35) is the well known one-dimensional solution for the third-type boundary derived by van Genuchten (1981) and several
others. Detailed derivation of the inverse Laplace transform can be found from van Genuchten (1981) or from the author upon
request.
3.2. Averaged approximation of aquitard advection and diffusion
In many previous studies, aquitard diffusion is approximated as a volumetric source/sink term in the governing equation of
transport in the aquifer, and aquitard advection is not taken into account (Neretnieks, 1980; Tang et al., 1981; Chen, 1985; Sudicky
et al., 1985; Fujikawa and Fukui, 1990; Tang and Aral, 1992a,b). In the following, we are going to include advection as part of the
mass transport mechanism. The governing Eq. (1) under this approximation is modied as:
AC
A2 C
AC
C
C
= Dx 2 v
1 + 2 ;
2B
2B
At
Ax
Ax
36
where 1 and 2 are the mass uxes across the upper and lower interfaces of the aquifer and they include both advective ux and
diffusive ux:
C1 =
AC
AC
1 v1z C1 1 D01 1
; C2 = 2 v2z C2 2 D02 2
:
Az z = B
Az z = B
37
The governing equations of the aquitards are still the same as Eqs. (5) and (10). C and C1 are continuous at the upper aquifer
aquitard interface, whereas C and C2 are continuous at the lower aquiferaquitard interface. Changing above equations into their
dimensionless forms and applying Laplace transforms, one can obtain the following solutions in the Laplace domain for the rsttype boundary condition as:
2
6
6
1
C D = exp6
6
p
4
s
q
q
Pe21 + 4p1 + Pe1 + 2
Pe22 + 4p2 Pe2 Pe
Pe2 + 4p + 1
2
2
C 1D = C D xD ; zD = 1; p exp4
C 2D
q
Pe1 =1 2 + 4p = 1 Pe1 = 1
2
3
7
7
xD 7
7;
5
38
3
zD 15;
2q
3
Pe2 =2 2 + 4p = 2 + Pe2 = 2
= C D xD ; zD = 1; p exp4
zD + 15:
2
39
40
If ignoring the advective uxes in the aquitards and assuming the upper and lower aquitards have identical parameters, then
the solutions come back to those derived by Tang et al. (1981), Fujikawa and Fukui (1990), and others. The same procedures can be
168
applied for the third-type boundary condition and will not be repeated here. Inverse Laplace transform of Eqs. (38)(40) will yield
the concentrations in the real time domain. We call the results of Eqs. (38)(40) the averaged approximation (AA) method
hereinafter. Now we compare the AA method with the Zhan method derived in Section 2.2.
4. Results analysis
The results of this study are formulated in dimensionless
forms which can be applied to any suitable real parameters. For
the sake of helping readers to understand the range of those
dimensionless variables, a table of realistically possible values
is provided in Table 2 and is briey illustrated. The effective
molecular diffusion coefcients chosen for the aquitard (D01 and
D02) and the aquifer (D0) are equivalent to a dilute NaCl solution
with D01 = D02 = D0 = 1.16 10 9 m2/s. The average pore
velocity v is chosen to be 0.10 m/day or 1.1610 6 m/s. The
same values of effective molecular diffusion coefcient and
average pore velocity were also used in Sudicky et al. (1985). For a
eld-scale dispersion problem, we choose the longitudinal
dispersivity x =2 m. The transverse vertical dispersivity is
expected to be 0.0095 of the longitudinal one as z =0.019 m.
Such a choice is consistent with eld-scale dispersion (Domenico
and Schwartz, 1998; Fetter, 1999) but is greater than the local
dispersivity used in the laboratory experiment of Sudicky et al.
(1985) and the eld experiment of Sudicky (1986), which is
0.001 m. Therefore, the corresponding dispersion coefcients
are Dx =xv +D0 = 2.32 10 6 m2/s, and Dz =zv +D0 =
2.32 10 8 m2/s. The aquifer thickness is 2B = 4 m. The
corresponding Peclet number Pe is 10. Two different leakage velocities in the aquitard are v1z =1.1610 10 m/s, and
1.1610 9 m/s, respectively, corresponding to Pe1 of 0.1, and 1.
The positive sign of v1z (or v2z) indicates that water is moving
upward. The porosity difference among the aquitards and aquifer
is expected to be a secondary effect, thus the identical value of
porosity 0.36 is assigned for all units. The same porosity is used
for the aquitard in Sudicky et al. (1985). The rst-type boundary
condition is used in the following discussion.
Symbol
Default value
2m
D01, D02
1.16 10 9 m2/s
Dx
2.32 10 6 m2/s
Dz
2.32 10 8 m2/s
v
v1z, v2z
1.16 10 6 m/s
1.16 10 10 m/s
1.16 10 9 m/s
0.36
Porosity
= 1 = 2
169
170
Z
dxD
Z
C 1D dzD ;
M2D =
Z
dxD
C 2D dzD
41
X
n=0
M2D =
X
n=0
4An cosn + n
;
q
q
2
2
Pe + 4 p + n Pe
Pe1 = 1 2 + 4p = 1 Pe1 = 1
4An cosn + n
;
q
q
Pe2 + 4 p + 2n Pe
Pe2 = 2 2 + 4p = 2 + Pe2 = 2
42
where An is given by Eq. (30) for the rst-type boundary
condition at x = 0 or Eq. (32) for the third-type boundary
and
condition at x = 0. Inverse Laplace transforms of M 1D
in Eq. (42) will result in the solution in the real time
M 2D
domain for the Zhan method.
If using Eqs. (39) and (40) for the AA method, one can
as:
and M 2D
obtain M 1D
4= p
M 1D = s
q
q
Pe2 + 4p + 1
+ 2
Pe
1
q
;
Pe1 = 1 2 + 4p = 1 Pe1 = 1
4= p
M 2D = s
q
q
2
2
Pe1 + 4p1 + Pe1 + 2
Pe22 + 4p2 Pe2 Pe
Pe + 4p + 1
1
q
:
Pe2 = 2 2 + 4p = 2 + Pe2 = 2
43
in Eq. (43)
Inverse Laplace transforms of M 1D and M 2D
will result in the solution in the real time domain for the AA
method. As an example, a comparison of the M1D values
calculated from Eq. (42) (the solid line) and Eq. (43) (the
dashed line) is shown in Fig. 5 for 1 = 2 = 0.05, 1 = 2 = 1,
Pe = 10, Pe1 = Pe2 = 0.1 and 1. Several observations can be
made from Fig. 5.
171
Fig. 5. The dimensionless mass transported to the aquitard per unit width
(M1D) calculated from the Zhan method and that from the AA method versus
the dimensionless time (tD) under the conditions of 1 = 2 = 0.05, 1 =
2 = 1, Pe = 10, Pe1 = Pe2 = 0.1 and 1.
Fig. 6. The dimensionless mass transported to the aquitard per unit width
(M1D) calculated from the Zhan method and that from the AA method versus
the dimensionless time (tD) under the conditions of 1 = 2 = 0.1, 1 =
2 = 1, Pe = 10, Pe1 = Pe2 = 0.1 and 1.
only 0.4% from Figs. 5 and 6. This nding implies that the mass
transported between the aquifer and the aquitard is relatively
insensitive to the aquitard diffusion when the aquitard Peclet
number is greater than unity.
Fig. 7 is similar to Figs. 5 and 6 except that the upper and
lower aquitards have different diffusion coefcients. In this
case, the upper aquitard diffusion coefcient is one half of the
lower one. One can tell that by reducing the upper aquitard
diffusion, the M1D value is slightly reduced. But in general, the
M1D value is less sensitive to the difference of the aquitard
diffusion coefcients. A great discrepancy has been observed
between the Zhan method and the AA method when
Pe1 = Pe2 = 1.
Fig. 8 shows the results for different aquitard Peclet
numbers. In this example, the Peclet number of the upper
aquitard is 1/10 of that of the lower aquitard. As can be seen
obviously, the M1D value is sensitive to the difference of the
Fig. 7. The dimensionless mass transported to the aquitard per unit width
(M1D) calculated from the Zhan method and that from the AA method versus
the dimensionless time (tD) under the conditions of 1 = 0.05, 2 = 0.1,
1 = 2 = 1, Pe = 10, Pe1 = Pe2 = 0.1 and 1.
172
Acknowledgements
This study is partially supported by National Science Foundation of China (#50428907), the CONACYT-TAMU Collaborative Research Program, and the Advanced Research Program
(ARP) from the Texas Higher Education Coordination Board,
USA. We thank two anonymous reviewers for their constructive comments. In particular, we thank the Editor-inChief E. O. Frind for his critical and detailed suggestions of this
manuscript. Dr. Frind's comments have helped us improve this
manuscript substantially.
Fig. 8. The dimensionless mass transported to the aquitard per unit width
(M1D) calculated from the Zhan method and that from the AA method versus
the dimensionless time (tD) under the conditions of 1 = 2 = 0.05, and
1 = 2 = 0.1, 1 = 2 = 1, Pe = 10, Pe1 = 0.1 and Pe2 = 1.
An C xD xD ; p; n cosn zD + n ;
A1
n=0
p + n C xD = 0:
2
dxD
dxD
A2
Pe2 + 4 p + 2n Pe
X
4
@
AxD 5 cosn zD + ;
CD =
An exp
n
2
n=0
V zD V 1
A4
q
Pe1 = 1 2 + 4p = 1 Pe1 = 1
2
zD z 1:
3
zD 15;
A5
An cosn zD + n = 1 = p:
C 2D
zD V 1:
A6
A7
2
n
1 2
q
Pe21 = 4 + p1 + Pe1 = 2
q
Pe22 = 4 + p2 Pe2 = 2
= n
A8
tan2
n
q
1 Pe21 = 4 + p1 + 1 Pe1 = 2
q
+ 2 Pe22 = 4 + p2 2 Pe2 = 2 ;
n + 1 2
q
Pe21 = 4 + p1 + Pe1 = 2
q
Pe22 = 4 + p2 Pe2 = 2
= n
An =
4
F ;
1 n n ;
p
F2 n ; n
A11
where
F1 n ; n = sinn cos n ;
A12
F2 n ; n = 2n + sin2n cos2 n :
, and C 2D
are derived after substituting
Therefore, C D , C 1D
Eq. (A11) into Eqs. (A4), (A5), and (A6), respectively. For the
special case in which the upper and lower aquitards have
identical parameters, n = 0.
References
A10
n=0
Similarly,
2q
3
Pe2 = 2 2 + 4p = 2 + Pe2 = 2
= C D xD ; zD = 1; pexp4
zD + 15;
2
173
A9
tan2
n
q
1 Pe21 = 4 + p1 + 1 Pe1 = 2
q
2 Pe22 = 4 + p2 2 Pe2 = 2 ;
n is determined from Eq. (A8). After that, n is calculated
from Eq. (A9). For the numerical calculation, accurate determination of n and n are important since these parameters have been
used in solutions (see Eqs. (27)(33)). As can be seen from Eq.
(A8), n is the root of a function tan(2n)=n/(n2 b), where
a and b are two dummy variables associated with all the other
2
parameters of Eq. (A8).
p Notice that function n/(n b)has a
singularity at n = b, and the function of tan(2n) has innite
numbers of singularities at n =/4+n/2, where n is an
integer varying from negative innity to positive innity. The
Ball, W.P., Liu, C., Xia, G., Young, D.F., 1997a. A diffusion-based interpretation
of tetrachloroethene and trichloroethene concentration proles in a
groundwater aquitard. Water Resour. Res. 33 (12), 27412757.
Ball, W.P., Xia, G., Durfee, D.P., Wilson, R.D., Brown, M.J., Mackay, D.M., 1997b.
Hot methanol extraction for the analysis of volatile organic chemicals in
subsurface core samples from Dover Air Force Base, Delaware. Ground
Water Monit. Remediat. 17 (1), 104121.
Batu, V., 1996. A generalized three-dimensional analytical solute transport
model for multiple rectangular rst-type sources. J. Hydrol. 174, 5782.
Bear, J., 1972. Dynamics of Fluids in Porous Media. Elsevier, New York, USA.
Bester, M., Frind, E.O., Molson, J.W., Rudolph, D.L., 2005. Numerical investigation of road salt impact on an urban well eld. Ground Water 44 (2),
165175.
Chen, C.S., 1985. Analytical and approximate solutions to radial dispersion
from an injection well to a geological unit with simultaneous diffusion
into adjacent strata. Water Resour. Res. 21 (8), 10691076.
de Hoog, F.R., Knight, J.H., Stokes, A.N.,1982. An improved method for numerical
inversion of Laplace transforms. SIAM J. Sci. Statist. Comput. 3 (3), 357366.
Domenico, P.A., Schwartz, F.W., 1998. Physical and Chemical Hydrogeology,
2nd ed. Wiley, New York, USA.
Fetter, C.W., 1999. Contaminant Hydrogeology, 2nd edition. Prentice-Hall,
Upper Saddle River, NJ, USA.
Fujikawa, Y., Fukui, M., 1990. Adsorptive solute transport in fractured rock:
analytical solutions for delta-type source conditions. J. Contam. Hydrol. 6,
85102.
Gillham, R.W., Sudicky, E.A., Cherry, J.A., Frind, E.O., 1984. An advective
diffusive concept for solute transport in heterogeneous unconsolidated
geological deposits. Water Resour. Res. 20 (3), 369378.
Hendry, M.J., Ranville, J.R., Boldt-Leppin, B.E.J., Wassenaar, L.I., 2003. Geochemical and transport properties of dissolved organic carbon in a clay-rich
aquitard. Water Resour. Res. 39 (7), 1194. doi:10.1029/2002WR001943.
Hunkeler, D., Chollet, N., Pittet, X., Aravena, R., Cherry, J.A., Parker, B.L., 2004.
Effect of source variability and transport processes on carbon isotope
ratios of TCE and PCE in two sandy aquifers. J. Contam. Hydrol. 74,
265282.
Huyakorn, P.S., Ungs, M.J., Mulkey, L.A., Sudicky, E.A., 1987. A threedimensional analytical method for predicting leachate migration.
Ground Water 25 (5), 588598.
Johns, R.A., Roberts, P.V., 1991. A solute transport model for channelized ow
in a fracture. Water Resour. Res. 27 (8), 17971808.
174
Johnson, R.L., Cherry, J.A., Pankow, J.F., 1989. Diffusive contaminant transport
in natural clay: a eld example and implications for clay-lined waste
disposal sites. Environ. Sci. Technol. 23, 340349.
Kreft, A., Zuber, A., 1978. On the physical meaning of the dispersion equation
and its solutions for different initial and boundary conditions. Chem. Eng.
Sci. 33, 14711480.
Kreft, A., Zuber, A., 1979. On the use of the dispersion model of uid ow. Int.
J. Appl. Radiat. Isot. 30, 705708.
Leij, F.J., Skaggs, T.H., van Genuchten, M.Th., 1991. Analytical solutions for
solute transport in three-dimensional semi-innite porous media. Water
Resour. Res. 27 (10), 27192733.
Liu, C.X., Ball, W.P., 1999. Application of inverse methods to contaminant
source identication from aquitard diffusion proles at Dover AFB,
Delaware. Water Resour. Res. 35 (7), 19751985.
Liu, C.X., Ball, W.P., 2002. Back diffusion of chlorinated solvent contaminants
from a natural aquitard to a remediated aquifer under well-controlled
eld conditions: predictions and measurements. Ground Water 40 (2),
175184.
Liu, H.H., Bodvarsson, G.S., Zhang, G., 2004. The scale-dependency of the
effective matrix diffusion coefcient. Vadose Zone J. 3, 312315.
Martin, P.J., Frind, E.O., 1998. Modeling a complex multi-aquifer system: the
Waterloo moraine. Ground Water 36 (4), 679690.
Moench, A.F., 1991. Convergent radial dispersion: a note on evaluation of the
Laplace transform solution. Water Resour. Res. 27 (12), 32613264.
Moench, A.F., Ogata, A., 1981. A numerical inversion of the Laplace transform
solution to radial dispersion in a porous medium. Water Resour. Res. 17 (1),
250252.
Moreno, L., Neretnieks, I., Eriksen, T., 1985. Analysis of some laboratory tracer
runs in natural ssures. Water Resour. Res. 21 (7), 951958.
Neretnieks, I., 1980. Diffusion in the rock matrix: an important factor in
radionuclide retardation? J. Geophys. Res. 85 (B8), 43794397.
Neretnieks, I., Eriksen, T., Thtinen, P., 1982. Tracer movement in a single
ssure in granitic rock: some experimental results and their interpretation. Water Resour. Res. 18 (4), 849858.
Ogata, A., Banks, R.B., 1961. A solution of differential equation of longitudinal
dispersion in porous media. U.S. Geol. Surv. Prof. Pap. 411, A1A7.
Parker, B.L., Cherry, J.A., Chapman, S.W., 2004. Field study of TCE diffusion
proles below DNAPL to assess aquitard integrity. J. Contam. Hydrol. 74,
197230.
Rasmuson, A., Neretnieks, I., 1981. Migration of radionuclides in ssured
rock: the inuence of micropore diffusion and longitudinal dispersion.
J. Geophys. Res. 86 (B5), 37493758.