Geoscience Reference
In-Depth Information
particularly if it has been possible to calibrate the model parameters by a comparison with observed
discharges; most model structures have a sufficient number of parameters that can be varied to allow
reasonable fits to the data. The problem is more often that there are many different combinations of
model structure and sets of parameter values that will give reasonable fits to the discharge data. Thus, in
terms of discharge prediction alone, it may be difficult to differentiate between different feasible models
and therefore to validate any individual model. This will be addressed in more detail in Chapter 7 in
the context of assessing uncertainty in model predictions and testing models as hypotheses about how a
catchment responds to rainfall.
On the other hand, the discharge predictions, together with any predictions of the internal responses
of the catchment, may also be evaluated relative to the original perceptual model of the catchment of
interest. Here, it is usually much more difficult to find a model that is totally acceptable. The differences
may lead to a revision of the parameter values used; to a reassessment of the conceptual model; or even,
in some cases, to a revision of the perceptual model of the catchment as understanding is gained from
the attempt to model the hydrological processes.
The remainder of this chapter will be concerned with the different stages in the modelling processes.
An example of a perceptual model of catchment responses to rainfall is outlined in Section 1.4; the
additional information that might be gained from considering geochemical information in Section 1.5;
the functional requirements of runoff production and runoff routing in Section 1.6; the definition of a
conceptual model in Section 1.7; and model calibration and validation issues in Section 1.8.
1.4 Perceptual Models of Catchment Hydrology
There are many outlines of the processes of catchment response available in the literature. Most general
hydrological texts deal, in greater or lesser detail, with the processes of catchment response. The volumes
edited by Kirkby (1978), Anderson and Burt (1990), and Beven (2006d) are of particular interest in
that the different chapters reflect the experience of a number of different hydrologists. Hydrological
systems are sufficiently complex that each hydrologist will have his or her own impression or perceptual
model of what is most important in the rainfall-runoff process so that different hydrologists might not
necessarily agree about what are the most important processes or the best way of describing them. There
are sure to be general themes in common, as reflected in hydrological texts, but our understanding of
hydrological responses is still evolving and the details will depend on experience, in particular, on the type
of hydrological environments that a hydrologist has experienced. Different processes may be dominant
in different environments and in catchments with different characteristics of topography, soil, vegetation
and bedrock. Part 10 in Volume 3 of the Encyclopaedia of Hydrological Sciences (Anderson, 2005) also
gives a review of different types of runoff processes with contributions from different hydrologists; a
review of runoff processes in semi-arid areas, for example, is provided by Beven (2002c) and in tropical
areas by Bonell (2004).
One of the problems involved in having a complete understanding of hydrological systems is that
most of the water flows take place underground in the soil or bedrock. Our ability to measure and assess
subsurface flow processes is generally very limited. Most of the measurement techniques available reflect
conditions only in the immediate area of the measurement probe. When the characteristics of the flow do-
main vary rapidly in space (and sometimes in time), the small-scale nature of such measurements can give
only a very partial picture of the nature of the flow. Thus, there is much that remains unknown about the
nature of subsurface flow processes and is, indeed, unknowable given the limitations of current measure-
ment techniques. It is necessary to make inferences about the processes from the available measurements.
Such inferences add information to the perceptual model of hydrological response, but they are inferences.
One way of gaining further understanding is to examine a part of the system in much greater detail.
Many studies have been made of the flow processes on particular hillslopes or plots, or columns of
Search WWH ::




Custom Search