Rise
Rise
Rise
0
q x; t ( )e
st
dt (9)
Taking the Laplace transform of equation (8) yields
d
2
x; s ( )
dx
2
= s x; s ( ) q x; t = 0 ( ) = s s; x ( ) x: (10)
Figure 1. Specification of the problem analyzed in this work. The water level in the stream starts from
h
L
and is raised uniformly during a time t
r
until reaching h
0
. It is then left there for an infinite time.
27 - 2 BOUFADEL AND PERIDIER: TECHNICAL NOTE
Equation (10) is an ODE in as a function of x, where s is
merely a free parameter. It is subject to the following
boundary conditions:
x = 0; s ( ) = 0 (11a)
and
x = 1; s ( ) =
1
s
1
gs
2
1 e
sg
( ): (11b)
Equation (10) may be solved by elementary methods, and its
solution is
x; s ( ) =
x
s
1 e
gs
( )
gs
2
sin h x
s
_
( )
sin h
s
_ : (12)
It can be easily verified that equation (12) satisfies the
governing equation, equation (10), and the boundary
conditions, equations (11a) and (11b). To obtain the solution
of the stated problem, one takes the inverse Laplace transform
defined by
q x; t ( ) = L
1
x; s ( ) ( ) =
1
2pi
Z
i
i
x; s ( )e
st
ds; (13)
where is an arbitrary positive real number. To calculate this
integral, the path of the line integral is closed at |s| , and
the closed contour is evaluated using calculus-of-residues
techniques [Brown and Churchill, 1996]. The exact solution
of equation (5) subject to equations (6) (8) is
q x; t ( ) =x 1
t
g
x 1 x
2
6g
2
g
X
m=1
sin mpx ( ) 1 ( )
m
e
mp ( )
2
t
mp ( )
3
;
0 _ t _ g (14)
q x; t ( ) =
2
g
X
m=1
sin mpx ( ) 1 ( )
m
e
mp ( )
2
t
e
mp ( )
2
tg ( )
mp ( )
3
;
t _ g _ 0: (15)
[10] It is important to verify that equations (14) and (15)
provide the same value of q at t = g. By setting t = g in
equation (15), one obtains
q x; g ( ) =
2
g
X
m=1
sin mpx ( ) 1 ( )
m
e
mp ( )
2
g
1
mp ( )
3
=
2
g
X
m=1
sin mpx ( ) 1 ( )
m
mp ( )
3
2
g
X
m=1
sin mpx ( ) 1 ( )
m
e
mp ( )
2
g
mp ( )
3
: (16)
The first term of equation (14) is identically zero at t = g,
and it may be shown that the first term in equation (16) is the
Fourier series expansion of x(1 x
2
)/6g. Hence equations
(14) and (15) are identical for t = g.
4. Water Flow
[11] The dimensional flow rate (assuming a unit cross-
sectional area) in the subsurface is given by
q* = K
dh
dx
: (17)
Using the dimensionless formulation (equations (4a) (4c)),
one obtains
q* = K
h
L
h
0
( )
L
@q
@x
: (18)
A dimensionless flow, q, may be defined as
q =
q*
K
hLh0 ( )
L
=
@q
@x
: (19)
The denominator of the second term in equation (19) is the
initial flow rate. From equations (14) and (15), one obtains
the dimensionless flow q at a selected time t and a selected
location x:
For t _ g,
q x; t ( ) = 1
t
g
1 3x
2
6g
2
g
X
m=1
1 ( )
m
cos mpx ( )e
mp ( )
2
t
mp ( )
2
;
(20)
and for t _ g,
q x; t ( ) =
2
g
X
m=1
1 ( )
m
cos mpx ( ) e
mp ( )
2
t
e
mp ( )
2
tg ( )
mp ( )
2
: (21)
Note that q (or q*) is positive in the positive x direction. The
flow rate at the groundwater/stream interface is obtained by
setting x = 1 in equations (20) and (21), resulting in
q x = 1; t ( ) = 1
t
g
1
3g
2
g
X
m=1
e
mp ( )
2
t
mp ( )
2
; t _ g: (22)
q x = 1; t ( ) =
2
g
X
m=1
e
mp ( )
2
t
e
mp ( )
2
tg ( )
(mp)
2
; t _ g: (23)
The (dimensionless) net outflow volume at the ground-
water/stream interface is obtained by integrating the flow q
over time, namely,
V(t) =
Z
t
0
q x = 1; t
/
( )dt
/
; (24)
where t is an integration variable. In other words, V is the
algebraic sum of the water volume that left the aquifer and
the water volume that entered the aquifer up to time t.
Inserting equations (22) and (23) in equation (24) results in
V(t < g) = t 1
t
2g
t
3g
2
g
X
m=1
1 e
mp ( )
2
t
mp ( )
4
; 0 _ t < g;
(25)
BOUFADEL AND PERIDIER: TECHNICAL NOTE 27 - 3
V
r
= V(t = g) =
g
2
1
3
2
g
X
m=1
1 e
mp ( )
2
g
mp ( )
4
; t = g; (26)
V t > g ( ) = V
r
2
g
X
m=1
1 e
mp ( )
2
g
1 e
mp ( )
2
tg ( )
mp ( )
4
;
t > g; (27)
where V
r
is the net exchange volume at the end of the rise
period.
[12] The final net exchange volume (i.e., for t ) is
obtained from equation (27) (after substituting for V
r
from
equation (26)) as V
> 0; (28a)
where more water leaves the aquifer than enters it;
g < 2=3 =V
< 0; (28b)
where less water leaves the aquifer than enters it;
g = 2=3 =V
= 0; (28c)
where the volume that leaves the aquifer is equal to the
volume that enters it. Noting that g represents the rise period,
one may use the relations above to classify the rise as slow
rise (g > 2/3), fast rise (g < 2/3), and neutral rise (g = 2/3).
5. Discussion
[13] Figure 2 shows plots of the dimensionless piezomet-
ric profile, (1 q), as function of space and time for g = 2/3.
Figure 3 shows plots of the volume V (equations (26) (27))
as a function of time for various g values. At earlier times,
groundwater leaves the subsurface (V positive and increas-
ing) until a critical time, t
max
, when V reaches its maximum
value, V
max
, and starts decreasing, which indicates that the
outward flow has stopped and stream water is entering the
subsurface. The time t
max
is obtained by setting the time
derivative of V, Q (x = 1, t), equation (20), to zero. The
resulting solution contains the sum over m in the general
case. However, because V
max
always occurs for t
max
< g,
one may neglect the sum over m for g _ 1. This results in
t
max
~ g 1=3: (29)
The error in estimating t
max
by equation (29) is less than
2% for g _ 1.
[14] V
max
is obtained by setting t = t
max
in equation (25).
Using the approximation given by equation (29), one obtains
V
max
~ Y=2 1=3 1= 9g ( ): (30)
In such a case the amount of water that enters the subsurface
when t is obtained as
V
in
= V
V
max
~ 1= 9g ( ): (31)
V
in
is negative indicating that even at large g values, a
certain volume of stream water will enter the subsurface.
Although this volume is small at large g values (~ 0.05 at
g = 2), the chemicals that it typically carries (e.g., dissolved
oxygen, nutrients) might be essential for biochemical
reactions around the groundwater/stream interface [Lendvay
et al., 1998; Genereux and Slater, 1999; Boufadel et al.,
1999b]. Equation (31) shows that this volume is inversely
proportional to g. The volume that enters the aquifer after
the rise period is given by
V
r
= V
V
r
=
2
g
X
m=1
1 e
mp ( )
2
g
mp ( )
4
: (32)
[15] Figure 4 shows that the magnitude of V
r
decreases
as g increases. The maximum value occurs at the intercept
(i.e., for g = 0) and is obtained by applying lHospital rule
on equation (32):
V
r
[ [
g=0
=
2
p
2
X
m=1
1
m
2
~ 0:3323 ~
1
3
: (33)
6. Notation
K hydraulic conductivity.
h piezometric head.
Figure 2. Variation of the surrogate piezometric profile, 1 q, as a function of the dimensionless space
and time (equations (14) and (15)).
27 - 4 BOUFADEL AND PERIDIER: TECHNICAL NOTE
h
L
Minitial piezometric head at right boundary
(Figure 1).
h
0
piezometric head at left boundary (Figure 1).
L domain length.
q dimensionless flow rate.
q* dimensional flow rate.
S
0
specific storativity.
t time.
t
r
rise period, time taken by h (L, t) to reach the value
h
0
starting from h
L
(Figure 1).
V dimensionless net exchange volume at the ground-
water/stream interface.
V
max
maximum outflow volume.
V
r
dimensionless net exchange volume at the ground-
water/stream interface at time t
r
.
V
r
dimensionless volume that enters the subsurface after
the time t
r
.
V