Geography Reference
In-Depth Information
transport, which is controlled by the vegetation resistance
(the valley land-use) and bedload delivery which in turn
are controlled by parameters changing at a decadal scale,
and (ii) the scale of the aquatic network within the braided
corridor whose architecture is fluctuating at a sub-annual
timescale. Fifty braided reaches have been selected within
the Rh one catchment, so that the entire area of interest is
covered.
In the case of the first question, which focuses on
a longer timescale, a typology of braided reaches was
developed. It was based, as before, on an object-oriented
remote sensing method applied to orthophotos to extract
various landscape features: gravel, water and riparian
vegetation patches. This information was combined with
vectorial data provided from other GIS sources: the active
channel width divided by catchment area, the channel
slope divided by the catchment area, the mean elevation
above sea level, and the percent of naturally recruited
forest area in the floodplain corridor which is not used
by human activities. Seven braided types were established
using a cluster analysis (Figure 11.11).
A regional organisation was then observed (see detail
in Piegay et al. 2009): western braided rivers have specific
characteristics compared to the eastern ones, whereas
both are located at similar elevations and positions in
the catchment. Upstream to downstream position was
also a discriminating factor. The upper reaches (type 4
for example) are characterised by a significantly narrower
riparian corridor and wider active channel than the ones
downstream. This research illustrates that there are several
characteristic types of evolution for braided rivers within
the Rh one basin. Types 4, 6 and 7 are active braided
channels, much wider than the others. Types 4 and 6,
located upstream, are in direct contact with available sed-
iment sources whereas type 7 further downstream drains a
region which is more sensitive to erosion due to lithology
and climate conditions than the one located to the west.
The western braided reaches may have also undergone
narrowing caused by vegetation encroachment earlier
than the eastern ones because change in human pres-
sure occurred sooner. This regional context also implies
that the western braided reaches presently have relatively
unique riparian corridors which may need to be pre-
served. This approach also shows that braided rivers can
be ordered from very active to inactive according to their
active channel width. Their position can change through
time because of flood series which impose inter-annual
width fluctuations, but the primary factor is probably
long term change in sediment delivery/availability which
explains the more significant changes in width.
When considering the networkscape of aquatic water
bodies in the braided corridor, it is expected to be
linked to both discharge (Egozi and Ashmore, 2008)
and groundwater upwelling and downwelling (Tockner
and Malard, 2003). All the waterbodies of each of the 50
braided reaches were detected on the orthophotos using
an object-oriented remote sensing method (see Belletti
et al., 2010). From this raw information, two braided
indexes were calculated. Indices included all the channels
(total braided index) and just the portion containing flow
(only branches connected upstream and downstream).
We then compared these results with the discharge of
each of the reaches during the survey. On Figure 11.11
are plotted the discharge frequency so that the different
braided reaches can be compared to each others (11.11c).
The figure also illustrates the difference between the total
accumulated length for all channels and just those con-
taining flow (11.11d). This shows that a set of reaches
located 900 km east and 1900 km north (e.g., within a
more focused area than that covered by the large map
shown in Figure 11.11a) have a total aquatic channel
length index much higher than the flowing index despite
being at low flow. The results thus identify an interesting
type of braided river with a large set of aquatic habitats
observed at low flow. The type presumably depends on
intense bedload transport and groundwater delivery, as
detected at the regional scale, again providing river man-
agers valuable information for designing a conservation
strategy.
11.5 Limitations and constraints when
enlarging scales of interest
The use of imagery for characterising biophysical charac-
teristics of river corridors in space and time has limitations
because of practical and methodological constraints asso-
ciated with the need to work with multiple images. The
main issues relate to: (i) image quality, (ii) treatment
errors, (iii) heterogeneity of geographical areas, and (iv)
treatment time.
The first limitation deals with the kind of imagery
which is available to cover large networks/reaches. As
shown by Wiederkehr et al. (2008), who did a detailed
analysis of available sources, most such approaches are
based on archived aerial photos (as opposed to archived
satellite imagery) because it provides a resolution suffi-
cient to include even narrow reaches within a regional
context. As a consequence, the spectral resolution is
often limited to one or in best cases three radiometric
Search WWH ::
Custom Search