Geography Reference
In-Depth Information
11.4.1 Retrospective approach on the Ain River:
understanding channel changes and
providing a sediment budget
additional evidence of a sediment deficit progressing from
upstream to downstream with social and ecological conse-
quences, and allow for the development of a management
strategy promoting, (i) the restoration of the reach still
affected by the dam impact, and (ii) the preservation of
the most interesting reaches downstream (e.g. Martinaz),
not yet impacted, by acting on the critical process, the
bedload transport. Bedload reintroduction was therefore
promoted upstream from the section (Rollet, 2007). The
option previously proposed by managers, an increase of
morphological discharge frequency (e.g. frequent floods),
was abandoned because of the risk to accelerate the prop-
agation of sediment downstream without mitigating the
gravel deficit.
Following these previous conclusions, and because the
sediment deficit also impacts channel shifting with cas-
cading ecological and social consequences downstream
(see Piegay et al., 2005), new questions emerge about the
sensitivity to change of the active shifting corridor due
to the forthcoming sediment deficit. A major question
relates to the capability of the river to recharge itself
through bank erosion processes, so that it may slow the
propagation of the bedload deficit. In such a context,
the understanding of the sediment exchange between the
active channel and the floodplain in the reach that is still
not affected by the dam becomes an issue. We addressed
this question by combining information derived from
aerial photographs with an additional data source, the
topography of the study area that results from a LiDAR
survey achieved in 2008 complemented with bathymetry
developed from a series of cross-section measurements.
We quantified the volume of sediment stored in the
point bar attached to each active meander bend due to
the channel migration, and the one eroded from the
floodplain between 1971 and 2008. The GIS overlays of
the channel positions between 1971 and 2008, derived
from the aerial photographs, provided the eroded and
constructed floodplain surface areas. We calculated the
volume of sediments for each of these polygons know-
ing its thickness from the DEM, the long profile of the
channel defining a reference elevation, and subtracting
the volume of fine sediment provided by overbank flows
that we estimated based on field sampling (soil coring).
For the eroded area, we determined an estimate of the
volume by assuming that the bank height and the surface
morphology eroded were similar to what we still presently
observe along the eroding bank. This assumption was ver-
ified by the analysis of the 1971 aerial photos confirming
the eroded surface is not different in textural characters
from the uneroded one used for present field measures.
The lower valley of the Ain River (40 km long), which is
a tributary of the Rh one River located 50 km north-east
of Lyon, France, underwent important adjustments in
terms of morphology and channel dynamics over the last
century (see Marston et al., 1995). A diachronic study
based on the analysis of historical aerial photographs
(eight series for the study period ranging from 1945 to
2000) was performed to understand how dams continue
to impact the current channel adjustment pattern and
sediment dynamics, and then to assess a long term man-
agement plan for the river. The main hypothesis was that
a dam built 10 km upstream from the alluvial reach in
1960 is interrupting bedload transport and is inducing a
winnowing process downstream.
We identified by photo-interpretation on each of the
aerial photograph series the wetted channel and the
unvegetated bars and integrated the layers in vector for-
mat in a GIS environment. Then we disaggregated these
layers into DGO of 10 m long to characterise the wetted
channel and the bars from upstream to downstream of
the study area. The longitudinal pattern of the cumulated
bar surface revealed (Figure 11.7 a): i) a narrowing and
bar surface decrease that propagated through time from
upstream to downstream. A detailed DGO has been cre-
ated to better distinguish the natural bar migration from
the bar reduction due to progressive sediment winnow-
ing. We did not observe bar migration but rather gravel
bar disappearance and local retention of the sediment
by channel-spanning infrastructure (weirs, bridges
...
).
A propagation rate for the sediment deficit downstream
from the dam was calculated by dividing the distance
between the dam and the front of the sediment deficit
identified by the progressive bar disappearance down-
stream, and the age of the dam. We obtained a minimum
estimate of 500 m a year, which is coherent with other
estimations of average annual bedload transport distance
in the region (Liebault et al., 2005). This observation was
then reinforced by conclusions provided from field data.
We photographed the coarsest particle patch of each of
the 109 bars and measured the median grain size from
imagery based on an automated procedure (see tech-
nique explained in Chapter 15). Figure 11.7b shows a
significant change in the median grain size at km 25 that
we interpret as the front of a region of sediment deficit
previously observed from photo analysis. These findings
are critical in terms of river management. They provide
Search WWH ::
Custom Search