Abstract
High discharges in the Dutch Rhine branches have initiated large explorative studies to suggest alternative ways of river management. These studies need detailed information on the morphological behaviour of rivers: especially information on the sediment transport dynamics at river bifurcations is wanted, for possible changes in the sediment distribution at
... read more
river bifurcations have implications for all parts of the downstream area. The present study focuses on the heavily engineered IJsselkop bifurcation, the location where the Pannerdensch Kanaal distributes into the rivers IJssel and Nederrijn. The IJsselkop is situated in the eastern part of the Netherlands, and is characterised by bimodal bed sediments (modes at 0.7 mm and 7 mm). The aims of this study were to determine the influence of discharge variations, weir management and river bed composition on the amount and size composition of sediment transport, as well as the distribution of sediment transport over the river width and over the downstream branches of the bifurcation. A sub-aim was to determine the characteristics of the dunes that caused the bed load transport.
To reach these aims, sediment transport measurements were carried out during two discharge waves in 2002 and 2004, and during a low-flow period in 2004. The bed load transport was measured directly using a Delft Nile Sampler (DNS), but also indirectly using multibeam echo soundings and a dune tracking technique. The suspended load transport (excluding wash load) was measured using Acoustic Sand Transport Meters (ASTMs). During the DNS and ASTM measurements samples of the transport load were taken, from which the size composition of the sediment transport was determined. Together with the sediment transport measurements also detailed data on the flow velocity, water levels and water discharge were collected.
The transport measurements at the IJsselkop showed that the suspended load transport, which accounted for most of total transport, strongly varied during a discharge wave. It was largest during the rising part of the discharge wave, and lowest thereafter. This suggests a declining source of suspended sediment during the discharge wave. The measurements just upstream of the bifurcation (in the Pannerdensch Kanaal) showed that the suspended load was almost uniformly distributed over the river width. This means that the distribution of suspended load sediment over the downstream branches of the bifurcation almost equalled the water distribution: the percentage of the total sediment load that entered the river IJssel was about 43% during high discharges and about 90% during low discharges. The suspended sediment distribution thus strongly differed between low-flow and high-flow periods, which must be attributed to the weirs in the Nederrijn, which are closed during low-flow periods and divert the water into the IJssel. Not only the sediment distribution was different between low-flow and high-flow periods, also the composition of the suspended load sediment: during high-flow periods the suspended load sediment had a median grain size of 0.35 mm, while the median grain size during low-flow periods varied about 0.14 mm.
The amount of bed load transport at the IJsselkop closely followed the discharge variations, except for the river Nederrijn. At the beginning of the Nederrijn the maximum transport occurred just before the peak discharge, while a few hundred metres downstream the maximum transport occurred two weeks after the peak discharge. This suggests the presence of a sand wave that slowly migrated downstream during the discharge waves. The exact origin of the sand wave remains unclear, but the occurrence of a sand wave clearly illustrates that the sediment transport at the IJsselkop was supply-limited. Namely, if there were no supply-limitation, the maximum sediment transport would have occurred at the same moment everywhere in the Nederrijn.
The grain size of the bed load transport material at the IJsselkop was much smaller than the local bed material at most measurement locations, which implies that the coarsest part of the local bed material was immobile during the discharge waves. The coarseness of the bed sediments at the IJsselkop thus has a suppressing influence on the amount of sediment transport and is the cause of the supply-limited transport conditions. During the discharge wave of 2004, the distribution of bed load transport over the river width of the Pannerdensch Kanaal was highly asymmetric, with the lowest values near the right bank, where the bed sediments are very coarse due to bend sorting processes. This uneven distribution of bed load transport over the river width also leaded to an uneven distribution of bed load transport over the downstream branches of the bifurcation. The IJssel, which bifurcates from the right part of the Pannerdensch Kanaal, only received 11% of the total bed load transport, while the Nederrijn received about 89%, probably resulting in a higher degree of supply-limitation in the IJssel than in the Nederrijn. During the low-flow period in 2004 the bed load distribution over the bifurcation was totally different, with only 2% flowing into the Nederrijn, due to the closure of the weirs.
The dunes that were present on the river bed were relatively small, with width-averaged heights of 0.1-0.4 m. At the end of the discharge waves small dunes developed on top of the existing dunes, especially in areas with coarse bed sediment. In all branches, the dune zone was smaller than the river width, which is in accordance with the idea of supply-limited transport
show less
