Environmental Engineering Reference
In-Depth Information
drylands and temperate hillslopes will be likely to require
both field and modelling investigations and will almost
certainly lead to the conclusion that different atomic
scales are required for different classes of hillslope. Cross-
scale dependency is one of the defining features of CAS.
As noted in Section 10.2 in the discussion of drylands,
local transfers of resources can lead to positive feed-
backs whereby patches develop and these form patterns
at larger scales which then affect resource flow through
the system. Linkages between fast and slow processes are
also defining features of CAS. Fast processes including
the redistribution of water across a hillslope during and
after rainfall affect slower-processes such as changes in
vegetation composition and patch migration.
It is useful to revisit the uniqueness problem because
this seems to be identified by several leading groups of
researchers. Beven and Pappenberger (2003: 333), for
example, note: 'Uniqueness of place is going to force a
change in the way that modelling is done.'
noted earlier regarding drylands and peatlands), but it is
difficult to envisage a situation where a hillslope showing
characteristic patterning in one part of a catchment has
hydrological behaviour that is much different from one
with similar patterning in another part of the same catch-
ment. If such similarity proves true, considerable effort
will be saved in characterizing catchment behaviour: there
is no need to treat each hillslope separately.
To conclude: the central theme of this chapter is pat-
tern. As shown in the foregoing, there is ample evidence
from drylands, peatlands and from the ecological and
biological literature on soil fauna such as earthworms and
ants, that hillslopes contain sometimes striking biophys-
ical patterns that may be important in affecting water
storage and movement. Despite such evidence, very few
hydrologists have attempted to look for such patterns
and even fewer have incorporated them in hillslope mod-
els. Until pattern is properly investigated, it will not be
possible to rank its importance alongside that of other
factors such as bedrock permeability (see Section 10.1;
Hopp and McDonnell, 2009). To build better hillslope
hydrological models, we need to (i) investigate surface
and subsurface pattern in the field, (ii) consider the
hydrological implications of that pattern - for example,
how do certain structures enhance or reduce hydro-
logical connectivity?, (iii) recognize and understand the
factors that promote changes or stability in pattern, and
(iv) understand how biophysical structures vary between
hillslopes and between ecosystems. A CAS approach pro-
vides a framework for this endeavour. There is, of course,
a possibility that pattern is a will-o'-the-wisp, and that
uniqueness is indeed a real problem, but there is suffi-
cient evidence to suggest that the CAS and ecohydrological
approaches to conceptualizing and investigating hillslope
hydrological behaviour will prove fruitful. Hydrologists
have ignored the 'eco' for too long.
There will be no single representation of the flow domain
that can be considered to be uniquely consistent with the
observations available (the equifinality [original empha-
sis] problem). The optimum characterization for one
event, will not necessarily be the optimum for another
event (especially in a different season or of different
magnitude). And even if we knew the equations of the
perfect model or the model 'whose representation is as
complete as possible' ... this problem does not go away
if the parameters of that model have to be identified on
the basis of limited data. (p. 332)
Whether uniqueness is a problem depends on what
we want from our hydrological models and how we view
the hillslope hydrological system. It is unreasonable to
expect hillslope hydrological models to provide good pre-
dictions of the discharge of water from the base of the
hillslope at all times; the level of detail at which the proper-
ties of the hillslope (including levels of soil water storage)
would need to be known would never justify the cost of
measurement. In any case, given that most measurements
interfere with or damage the hillslope, such perfect mea-
surement knowledge would result in substantial change to
the system being modelled - 'We murder to dissect' ( The
Tables Turned ,in Lyrical Ballads , William Wordsworth,
1798) - so that its behaviour might be quite different from
that before the measurements. But, more importantly and
as shown earlier, many hillslopes may be far from unique
because of their characteristic biophysical structures. It
is, of course, possible that hillslopes that possess simi-
lar patterns display different hydrological behaviours (as
10.6 Acknowledgements
Many of the ideas in this chapter have developed from
wide-ranging and thought-provoking discussions on
'matters ecohydrological' with Lisa Belyea, John Wain-
wright, Paul Morris, Sarah Dunn, Mike Waddington,
Nick Kettridge, and Nikki Dodd. I am also grateful
to Nikki Dodd, Mike Kirkby, Alice Milner, and two
anonymous reviewers for suggesting how an earlier draft
of this chapter could be improved. As ever, the usual
disclaimer applies: any inconsistent or loose arguments,
or indeed any other errors, are mine and mine alone.
Search WWH ::




Custom Search