Geoscience Reference
In-Depth Information
Phenomena interpreted to be the result of this type of mechanism were observed, for
example, in a swampy area by Novakowski and Gillham (1988) and in a grass-covered
low relief basin by Abdul and Gillham (1989), both in Ontario. In these studies, the
rise of the water table was most pronounced in the near-stream areas. The mechanism
has also been inferred to occur in more rugged terrain. During a sprinkler irrigation
experiment on a very steep (43%) forested hillslope near Coos Bay in Oregon, Torres
et al
. (1998) applied a sudden input spike, after the system had been driven to a steady
state flow and the soil water pressures were mostly between 0 and
10 cm. They
supposed that the timing and magnitude of the pore water pressure and of the dis-
charge rate response to this sudden input were much faster than could be expected
from advective water movement, and concluded that the fast response was triggered
by a pressure wave moving undetected through the unsaturated zone; thus a small
amount of rain on a wet soil profile can supposedly result in a rapidrise in the satu-
rated zone, with a relatively slight increase in hydraulic gradient and a large increase
in hydraulic conductivity. They also observed some preferential flow, but they felt that
inthis particular soil, this effect was minor compared to that of the soil water retention
characteristics.
The concept that suction-saturated capillary fringe water can be easily converted into
water below the water table, that is from a negative to a positive pressure, by a relatively
small amount of rain, is undoubtedly realistic. Clearly, only a little additional water is
required to mobilize the soil water, when the soil is already close to saturation. But the
importance of this mechanism should be kept in perspective. For example, it can only
be expected to be effective when the pore water pressure in the top layers of the soil
is arrivedatduring a drainage phase and not during a wetting phase; as illustrated in
Figures 8.14, 8.18 and 8.19, the capillary fringe is usually much smaller in the wetting
cycle. Similarly, in the absence of any macropores or pipes, the water table (i.e. the locus
of atmospheric- or zero-pore water pressure) can only be expected to move rapidly down
a steep slope, if it is already close to the surface. As illustrated inFigure 10.12, the
drainable porosity
n
e
is smaller when the water table is closer to the surface, that is when
(
−
p
w
)
max
is smaller. While not a perfect representation, hydraulic groundwater theory,
as formulated by the Boussinesq equation (10.29) and its linearized form (10.134), is also
fully consistent with this. This can be seen by considering the advectivity in Equation
(10.29) (also (10.136)); rewritten here for convenience
−
α
k
0
sin
c
h
=−
(11.2)
n
e
it describes the speed of propagation of a given water table height
(or of a disturbance
of the water table resulting from rainfall) down the slope. This shows that, in the absence
of preferential flow paths, large values of
c
h
can result only when the drainable porosity
is small. As seen in Equation (10.151), this is equally consistent with the kinematicwave
approximation.
Capillarity induced flow enhancement has also been linked to soil stratification. In sit-
uations where a fine-textured soil layer overlies a more coarse-grained material, the inter-
face between the two layers can develop into a capillary barrier (Ross, 1990; Steenhuis
η
