Geoscience Reference
In-Depth Information
Fig. 9.20 A storm hyetograph showing the
rainfall intensity
P
as a function
of time, with an assumed initial
loss and a subsequent constant
loss rate. The remaining blank
area is considered the excess
precipitation, which produces the
storm runoff.
P
Time,
t
of the infiltrated water participates actively in the runoff generating process, the loss
rate concept is of questionable validity and not soundly based. Nevertheless, as long
as its limitations are kept in mind, the concept can be useful, especially in situations
with limited data and for design purposes involving large flows and floods. There have
been two major ways of applying the loss rate concept; in the first, which is used more
commonly for large basins, it is assumed that the loss is independent of the rainfall, while
in the second one, which is more widely used for smaller areas, the loss is assumed to
be proportional to the rainfall.
Loss rate independent of rainfall
In this approach the loss rate is usually taken as a constant rate throughout the rainstorm
event, and it is subtracted from the actual rainfall intensity to obtain the excess rainfall
intensity. The underlying idea is that this loss rate represents mainly the space- and time-
averaged behavior of the infiltration capacity which is controlled by the properties of the
soil and which is independent of the rainfall intensities, as long as they are large enough.
Several such indices have been proposed in the past, but Horton's (1937) method has
probably been most widely used.
In brief, the loss rate is determined as the constant value that must be subtracted
from the actual rainfall rate so that the resulting excess rainfall volume over the entire
catchment is equal to the actual storm runoff volume; this storm runoff is derived from
the streamflow hydrograph by subtracting the (assumed) baseflow from the total runoff.
When rainfall is observed at several gages, their input must be properly weighted with
their respective areas of influence (see Section 3.3.1) and several trial loss rates are
required to obtain the solution.
This method is often applied in modified form, by the inclusion of an initial loss,
which is a certain amount that is subtracted from the rainfall at the beginning of the
storm event. The principle is illustrated in Figure 9.20. The initial loss can be defined as
the loss that takes place before the onset of storm runoff in the stream, and it is usually
envisioned to consist of interception storage, depression storage and initial high rate
