Geoscience Reference
In-Depth Information
Figure 6.10 Relationship between storm rainfall and runoff coefficient as percentage runoff predicted by the
USDA-SCS method for different curve numbers.
or a way of predicting Hortonian infiltration excess runoff (e.g. Bras, 1990), and a study by Yu (1998)
has attempted to give it a basis in physical theory by showing that partial area infiltration excess runoff
generation on a statistical distribution of soil infiltration characteristics gives similar runoff generation
characteristics to the SCS method (see Box 6.3).
This is of some interest in itself, but the method becomes even more interesting if we return to its
origins as a summary of small catchment rainfall-runoff measurements by Mockus (1949). Mockus
related storm runoff to rainfalls and showed that the ratio of cumulative discharge to cumulative storm
rainfall shows a characteristic form (see Figure 6.10). In the past, the storm runoff may have been widely
interpreted as infiltration excess runoff, but this is not a necessary interpretation. At the small catchment
scale, the measured runoff in some of the original experiments will almost certainly have included some
subsurface-derived water due to displacement, preferential flows or subsurface contributions from close
to the channel. Certainly the method has since been applied to catchments and hydrological response
units that are not dominated by infiltration excess runoff generation. Steenhuis et al. (1995) have already
interpreted the SCS method in terms of a variable saturated contributing area, excluding, in their analysis,
some data from high intensity events that might have produced an infiltration excess runoff (see also Box
6.3). A sufficient view of the method is that it incorporates some empirical knowledge of fast runoff
generation, by whatever method, at the small catchment scale into a simple functional form. It may
be necessary to check whether that form is appropriate in any particular application but it may be an
appropriate method to use at the HRU scale since it encapsulates knowledge gained at similar scales. In
considering the scale dependency of an HRU model, therefore, it might be more appropriate than, say,
any of the point scale infiltration equations described in Box 5.2, even though they would normally be
considered more “physically based”.
Search WWH ::




Custom Search