Geoscience Reference
In-Depth Information
Figure 7.11
Prediction bounds for four recording boreholes in the Saeternbekken catchment for the 1989
simulation period, showing prior bounds after conditioning on the 1987 simulation period and posterior bounds
after additional updating with the 1989 period data (after Lamb
et al.
, 1998b, with kind permission of Elsevier).
Blazkova
et al.
, 2002a, 2002b; Gallart
et al.
, 2007). It is, perhaps, representative in showing that we
might not expect a rainfall-runoff model to reproduce all of the observations all of the time, even when
the predictions are associated with some uncertainty bounds. It is also perhaps representative in revealing
that there may be difficulties associated with using information from internal state measurements in model
calibration or conditioning. Firstly, the use of such data may require adding additional local parameter
values; secondly, such local data may not have great value in conditioning the prediction bounds for
catchment discharges.
7.12 Case Study: Application of GLUE Limits of Acceptability
Approach to Evaluation in Modelling the Brue Catchment,
Somerset, England
This recent application of the GLUE methodology makes use of the limits of acceptability approach
to model evaluation in an application of Dynamic TOPMODEL to the Brue catchment (135 km
2
)in
