Soil and Hillslope (Eco)Hydrology - Environmental Modelling: Finding Simplicity in Complexity - page 158

10

Soil and Hillslope
(Eco)Hydrology

Andrew J. Baird
School of Geography, University of Leeds, UK

systems (CAS) (defined below). I argue that, if we are to
find the 'rules' that lie behind the hydrological functioning
of hillslopes, we need to look first at biophysical patterns
both on and in hillslopes and need to recognize that our
rules necessarily will incorporate linkages between ecolog-
ical and hydrological processes. Once we have found these
rules, we will be much better placed to produce useful
( sensu Box, 1979) hydrological models of hillslopes. This
search for biophysical structures, how they vary accord-
ing to, for example, vegetation type and climate, and how
they affect and are affected by water flow comprises what
may be termed an ecohydrological approach.
The world around us is far from uniform. Hydrolo-
gists have known this for a long time, and some have
appreciated its implications. There are many studies that
show that key parameters such as soil hydraulic conduc-
tivity ( K ) display considerable variability (coefficients of
variation [CV] often > 100%) and that it is necessary
to address such variability in our hydrological models.
Notable among these is the study of Binley et al . (1989a,
b). Using a 3-D finite-element model of subsurface hill-
slope water flow based on the Richards equation and
a simple overland flow routine, Binley et al . (1989a)
showed that hydrological functioning (as represented by
the hydrograph of water discharging from the base of the
hillslope in response to rainfall events) depends not just
on the variability of a parameter like hydraulic conduc-
tivity ( K ) but also on the spatial arrangement or pattern
of that parameter. Thus, for the same probability den-
sity function (PDF) of K , we might expect differences
in hillslope hydrological behaviour, depending on the

10.1 Hillslope e-c-o-hydrology?

Hillslope hydrology has been studied by many hundreds
of researchers for many decades, yet papers published
since the mid-1980s suggest we have failed to build models
that adequately capture the hydrological 'behaviour' of
hillslopes (e.g. Beven, 2001a; James et al ., 2010). What is
wrong with our hillslope hydrological models? Is there an
underlying simplicity to hillslope hydrological behaviour
that we're missing, or is the problem of hillslope hydrology
one of a system that is essentially unknowable because it
is too complicated?
The latter argument was, in part, made in the first edi-
tion of this chapter (Baird, 2003). Here I attempt to move
the argument on and suggest that hydrologists have been
somewhat myopic in the way they have chosen to con-
ceptualize hillslopes. I adopt a deliberately provocative
style in the hope that it will stimulate debate on how we
should frame the problem of modelling hillslope hydro-
logical behaviour in future. Those expecting a detailed
presentation of existing hillslope hydrology models will
be disappointed; my purpose here is to think at the higher
level of modelling approach and to address the question
of the processes and feedbacks that should be included in
new models of hillslopes. My argument centres on a gen-
eral (albeit by no means universal) failure of hydrologists
to recognize that hillslopes are not static and/or solely
physical entities, but are complex and often dynamic bio-
physical structures (even over periods of a few years or
decades) that show evidence of being complex adaptive