Environmental Engineering Reference
In-Depth Information
hydrological system (e.g. Engelen, 2000) or in larger
systems of which hydrology is a small, but significant, part
(Engelen, Chapter 21). Though it is generally policy mak-
ers who use these tools to support the decision-making
process, there are even efforts to make catchment-scale
model results available direct to the public in real time over
the Internet for flood-warning purposes (see Chapter 25
and also de Roo
et al
., 2000; Al-Sabhan
et al
., 2002; Cloke
and Pappenberger, 2009).
Q
Q
Time
Time
(a)
(b)
11.2.8 Connectinghillslopeswithchannels
One of the major issues facing the study of catchment
hydrology is the way reductionist approaches to science
have led to the division of catchment elements into
different areas of study. Thus, researchers working on
rivers usually consider hillslopes as irrelevant areas 'up
there', while those working on hillslopes consider the
rivers 'down there' to be irrelevant. Wainwright
et al
.
(2011) have suggested that this approach shows a lack of
disciplinary connectivity. Church (2002) demonstrated
that the coupling of hillslope and channel elements pro-
duce dynamic interactions that affect the movement of
water and sediment through the system, with important
feedbacks to the spatial distribution of sediment and
associated habitats for plants and animals. Figure 11.7
shows how some of the different ways in which slopes are
coupled to channels will affect the corresponding flow in
the channel. Subsequently, there are feedbacks between
channel flows and hillslope form, as demonstrated by
Armstrong (1987). As the slope form changes (producing
different convexities and concavities in plan and profile),
the concentration of flow reaching the channel will also
change, thus providing a feedback. Indeed, in many steep,
upland catchments, channel undercutting of slopes can
produce landslides that subsequently divert the channel
to the opposite side of the valley, producing undercut-
ting there (an example from southern Spain can be seen
in Figure 27.1). Thus, understanding hillslope-channel
coupling is fundamental for understanding catchment
hydrology on a variety of timescales.
At the scale of an individual flood event, Michaelides
and Wainwright (2008) used a laboratory flume to eval-
uate some of the critical feedbacks between hillslope and
channel flow. The laboratory experiment enabled control
over parameters such as hillslope angle, channel angle,
flow rates from the hillslope and from the upstream chan-
nel, and presence or absence of floodplains. These param-
eters were chosen as a result of earlier model sensitivity
analyses (Michaelides and Wainwright, 2002). However,
Q
Q
Time
Time
(c)
(d)
Figure 11.7
Effects of different hillslope-channel coupling
configurations on the water and sediment reaching the channel
system: (a) directly coupled hillslope and channel; (b) hillslope
decoupled from channel by floodplain; (c) decoupled case with
more complex pathways; (d) hillslope completely decoupled
from channel with fan deposition on floodplain (from
Wainwright, 2006). Church (2002) suggests that there will be a
spatial pattern with (a) more characteristic of headwater
catchment and (b)-(d) more frequently found in lowland
catchments, although this pattern will not necessarily always be
the case (Reproduced with permission from Wainwright, J.
(2006) Degrees of separation: hillslope-channel coupling and
the limits of palaeohydrological reconstruction,
Catena
, 66,
93-106).
the approach is not straightforward, as the scaling
relationships required in the laboratory (in this case
using a 1:100 scale model) mean that different flow char-
acteristics (discharge, depth, velocity and time) all vary in
different ways relative to the scale of the model. The results
of the laboratory analysis suggested a strong spatial auto-
correlation in the effect of flows coming off hillslopes into
the channel, so that simulating flow patterns is strongly
dependent on getting spatial interactions right. They also
suggested that there are interactions between slope and
channel variability, but it depends on the ways in which
one measures model error as to how well these interactions
can be interpreted (see also discussion in Chapter 2).
Over multiple flow events, feedbacks between pro-
cess and form become more important. For example,
McGuire and McDonnell (2010) defined three timescales
of feedback on flows in a temperate catchment in the US,
depending on the relative speed of different surface and
subsurface flow pathways. In their comprehensive review
of hydrological connectivity, Bracken and Croke (2007)
Search WWH ::
Custom Search