Geoscience Reference
In-Depth Information
their applicability in this respect since they take no account of the effects of macropores in the
soil on infiltration and redistribution (though see the work of Beven and Clarke (1986) for one
attempt at an infiltration model that takes account of a population of macropores of limited
depth).
B5.2.8 Derivation of Soil Moisture Characteristics from Infiltration Measurements
Measuring the rate at which water infiltrates into the soil surface is one of the simplest ways of
trying to assess the hydrological characteristics of soils. A number of techniques are used, in-
cluding ponding water within single and double ring infiltrometers, porous plate infiltrometers
with pressure controls that can be used to exclude infiltration into macropores, and plot scale
sprinkler experiments. There is a long history of using analytical solutions to the infiltration
equation (including three dimensional effects beneath an infiltration ring of limited dimension)
to derive soil moisture characteristics from infiltration rates. A recent example is provided by
the BEST technique of Lassabatere
et al.
(2006) that is based on scaled forms of infiltration
equation presented by Braud
et al.
(2005).
B5.2.9 Estimating Infiltration at Larger Scales
All of the equations above are for the prediction of infiltration at the point scale. They have
often been applied wrongly at hillslope or catchment scales, as if the same point scale equation
applied for the case of heterogeneous soil properties. However, because of the nonlinearities
inherent in any of the infiltration equations, this is not the case. It is therefore necessary to
predict a distribution of point scale infiltration rates before averaging up to larger scales. Similar
media theory (see Box 5.4) can be useful in this case, if it can be assumed that the soil can
be treated as a distribution of parallel non-interacting heterogeneous columns. Clapp
et al.
(1983), for example, have based a field scale infiltration model on this type of representation.
Philip (1991) has attempted to extend the range of analytical solutions to the case of infiltration
over a sloping hillslope, albeit a hillslope of homogeneous soil characteristics.
There have been many hypothetical studies of the effects of random variability of soil prop-
erties and initial conditions on infiltration at the hillslope and catchment scales, some taking
account of the infiltration of surface flow from upslope, others not (see, for example, the study
of Corradini
et al.
, 1998). These have mostly been based on purely stochastic heterogeneity
but it is worth noting that some variation may be systematic (see, for example, the depen-
dence of infiltration on depth of overland flow and vegetation cover included in the model of
Dunne
et al.
, 1991). To apply such models to a practical application will require considerable
information on the stochastic variation in soil properties and, even then, will not guarantee
accurate predictions of runoff generation (see the discussion of the R-5 catchment case study in
Section 5.7).
Storage-based approaches need not be considered only as point infiltration models. A func-
tion such as that of Equation (B5.2.6) may equally be considered as a conceptual representation
of basin wide infiltration, with the storage as a mean basin storage variable. Other storage ap-
proaches have taken a more explicit representation of a distribution of storage capacities in a
catchment in predicting storm runoff. These include the Stanford Watershed model and vari-
able infiltration capacity models (see Section 2.4 and Box 2.2), and the probability distributed
model (see Section 6.2).
A further widely used method for predicting runoff generation at larger scales, often inter-
preted as an infiltration model, is the USDA Soil Conservation Service (SCS) curve number
approach. This is discussed in more detail in Chapter 6 and Box 6.3, where it is shown that
it can be interpreted in terms of both spatially heterogeneous infiltration rates and dynamic
contributing areas generating saturation excess surface (and possibly subsurface) runoff.
