Geoscience Reference
In-Depth Information
and tracers, although the number of reports is dominated by results from humid temperate catchments
and small to moderate rainfall events (Sklash, 1990). The study of Sklash and Farvolden (1979) is
particularly interesting in showing that samples from surface runoff at a point were sometimes dominated
by “old” water and sometimes by “event” water. The technique can be extended, using other environmental
tracers, to three-component mixing to differentiate the rainfall contribution from “soil water” and “deep
groundwater” components, where these components can be differentiated geochemically (see Bazemore
et al.
, 1994). Again, a major component of pre-event water is often found even, in some cases, for very
rapidly responding processes such as pipe flows in wet soils (Sklash
et al.
, 1996).
The pre-event water is displaced from storage by the effects of the incoming rainfall. This must
therefore necessarily involve subsurface flow processes. The fact that the rising limb of the hydro-
graph is often dominated by the pre-event water component reveals that this displacement can take
place rapidly, despite the fact that subsurface flow velocities are generally assumed to be much slower
than surface flow velocities. This perception is, in fact, one of the reasons for the continuing use
of the Hortonian conceptualisation of runoff production, even now. If subsurface velocities are so
slow, how can subsurface flow and pre-event waters make a major contribution to the hydrograph
(Kirchner, 2003)?
The answer lies, at least in part, in the physics of the flow processes; in particular, in the saturated
zone. It can be shown that there is a difference between the flow velocity of water and the velocity with
which a disturbance to the saturated zone is propagated as a pressure wave, which is called the
wave
speed
or
celerity
. The type of disturbance of interest here is the addition of recharge due to rainfall during
an event. The theory suggests that an infinitely small disturbance at the water table will be propagated
infinitely quickly. Larger disturbances will have a much smaller wave velocity, the magnitude of which
is a function of the inverse of the
effective storage capacity
in the soil (the difference between the current
soil moisture in the soil immediately above the water table and saturation). In a wet soil or close to a
water table, the effective storage capacity may be very small so that the wave velocity may be very much
faster than the actual flow velocity of the water (see Section 5.5.3). The increase in discharge to the
stream during an event will then depend more on the response of the hydraulic potentials in the system,
which will be controlled by the local wave velocities, than the actual flow velocities of the water. Thus
if discharge starts to increase before the recharging water has had time to flow towards the channel, it
will be water stored in the profile close to the stream that flows into the channel first. This water will be
predominantly pre-event water, displaced by the effects of the rainfall. There may also be local exchanges
between event water and pre-event waters that cause displacements into local surface runoff with higher
velocities (Iorgulescu
et al.
, 2007).
Similar effects may take place in unsaturated soil, but here the picture or perception is made more
complicated by the relative mobility of water stored in different parts of the pore space and by the effects
of preferential flows within the structural voids of the soil. The important message to take from this section
is that in many catchments, particularly in humid environments, an important part of the hydrograph may
be made up of “old” water and may not be rainfall flowing directly to the stream. Certainly, it should
not be assumed that fast runoff is always the result of overland flow or surface runoff on the hillslopes
of a catchment. A more extensive discussion of the identification of runoff sources and modelling of
residence times in catchments is to be found in Chapter 11.
1.6 Runoff Generation and Runoff Routing
The evidence discussed in the previous two sections has been primarily concerned with the processes
of runoff generation, both surface and subsurface. Runoff generation controls how much water gets into
the stream and flows towards the catchment outlet within the time frame of the storm and the period
immediately following. There is also, however, a further component to consider, which is the routing of
the runoff from the source areas to the outlet. The boundary between runoff generation and runoff routing
