Geoscience Reference
In-Depth Information
Figure 7.18
The South Tyne at Featherstone: (a) Master recession curve and (b) example of event separation
(after Beven et al., 2011).
input (runoff coefficients greater than one). For these periods, even a perfect rainfall-runoff model would
necessarily underestimate the observed discharges (unless the type of rainfall multiplier of Figure 7.5a
is invoked). There are also some very unexpectedly low runoff coefficients for a catchment in which the
long-term average discharge is 73% of the observed rainfalls.
Similar issues might be expected in very many catchments. It seems that disinformative periods of
data might be a generic problem in rainfall-runoff modelling. Even if it will remain difficult to be secure
about identifying disinformative periods because they do not result from simple statistical variability but
