Geoscience Reference
In-Depth Information
flows as a uniform sheet flow; treating pollution transport as a simple advection-dispersion problem, etc.)
should also be considered misleading. What this should teach us is that we need to be careful about
evaluating process descriptions and constraining principles as hypotheses using appropriate but uncertain
data. The issue in doing so is how far the information is commensurate with model predictions and whether
uncertainties might mean that some periods of data might be disinformative in inferring a good model of
the system, especially when the input data are poorly specified (Section 7.17; Beven
et al.
, 2008; Kuczera
et al.
, 2010a; Beven and Westerberg, 2011).
9.9 Key Points from Chapter 9
•
A new generation of rainfall-runoff models is needed to replace the concepts of the Freeze and Harlan
(1969) blueprint for physically based models.
•
It is suggested that a suitable framework for a novel approach already exists in the representative
elementary watershed (REW) concepts of Reggiani and others, which is based on equations of mass,
energy and momentum balance in discrete elements of the landscape.
•
These equations, however, give rise to a
closure problem
, because they involve fluxes of mass, energy
and momentum across the boundaries of the elements. Since it is difficult to show by measurement
that these balance equations hold for real elements of the landscape, there will necessarily be some
uncertainty inherent in defining closure schemes.
•
Implementations of the REW concepts to date are not satisfactory in that they do not generally recognise
the hysteretic and scale dependence required of a closure scheme.
•
Recent work that might form the basis for better closure schemes is reviewed, including the multiple
interacting pathways (MIPs) modelling strategy that attempts to predict both flow and residence times
of water in the system within the same framework.
•
It might be possible to impose on the representations other constraints than the balance equations.
Recent work has suggested that a local optimality principle for vegetation to maximise net carbon
production given available energy and water might be useful in constraining the prediction of actual
evapotranspiration.
•
In general, it is useful to view catchments as complex adaptive systems subject to different external
forcings that have variable
relaxation times
.
