Geoscience Reference
In-Depth Information
Figure 5.12
Results of modelling runoff at the plot scale in the Walnut Gulch catchment: (a) the shrubland
plot and (b) the grassland plot; the error bars on the predictions indicate the range of 10 randomly chosen
sets of infiltration parameter values (after Parsons et al., 1997, with kind permission of John Wiley and Sons).
velocity on raingauge catch. However, checking the model predictions for different numbers of raingauges
they showed that combinations of four gauges (that is a density of 1 per hectare) gave a variation in
predicted discharges that spanned the observed discharges and had a similar coefficient of variation to
that estimated for the discharge measurements (Faures
et al.
, 1995). Goodrich
et al.
(1994) also looked
at the sensitivity of runoff production to the pattern of initial moisture content at the larger scale of the
WG-11 subcatchment (6.31 km
2
). They suggest that a simple basin average of initial moisture content
will normally prove adequate and that, again, knowledge of the rainfall patterns is far more important.
A water balance model estimate of initial moisture content did as well as the remotely sensed estimates.
Michaud and Soroochian (1994) compared three different models at the scale of the whole catchment
including a lumped SCS curve number model, a simple distributed SCS curve number model and the
more complex distributed KINEROS model. The modelled events were 24 severe thunderstorms (mean
runoff coefficient 11%), with a raingauge density of one per 20 km
2
. Their results suggested that none of
the models could adequately predict peak discharges and runoff volumes, but that the distributed models
did somewhat better in predicting time to runoff initiation and time to peak. The lumped model was, in
this case, the least successful.
