Floodplain avulsion channels: understanding their distribution and how they reconnect to the parent channel

Abstract

This study is concerned with new river channels that develop on floodplains. These channels can develop gradually, or they can develop more rapidly (avulsions). This study concentrates on the relatively more rapid channel changes known as avulsions. An avulsion specifies the gradual abandonment of an existing river-channel and in response to this, the processes of development of a new channel on a floodplain nearby.

The study addresses two specific knowledge gaps: 1) how common are river floodplain avulsions globally, and 2) what are the detailed processes that occur at the up and downstream points where avulsions connect to the main channel?

Using random sampling from a global spatial layer I discovered that developing avulsions are extremely common on alluvial floodplains globally, wherever the floodplain is wider than several channel widths. Avulsions are most common on single thread meandering floodplain types, but a review of avulsion literature shows that research is biased to relatively less common floodplain types. Avulsions increase the rate of valley widening, particularly in narrow floodplains. There is a relationship between floodplain width and the number of avulsion channels. The rest of the thesis is focussed in the major process knowledge gap which is how avulsion channels connect into the main channel at the up and downstream ends.

The focus of the process component of this study is the broad Murray river floodplain from Yarrawonga to Echuca, SE Australia. I mapped and classified developing channels on the floodplain and found that the avulsion connection point here develops in an unusual way, involving the development and coalescence of low points (depressions) on the levee (this mechanism is very different from normal crevasse splay development). The chain of low points on the alluvial levee coalesce to form a levee channel. Rather than forming by erosion as expected from the literature, form progressively by locally reduced vertical accretion. This identifies a new process by which topography is developed on floodplains. Initially the levee channels are not connected with the Murray main stream and slope away from the river. The connection occurs by lateral migration of the river bank into the levee channel, but also by progressive upslope (river ward) migration of the deepest part of the levee channel towards the river, narrowing the gap between the river and the levee channel. Following connection, the levee channel captures flow from the river, and hydraulic modelling shows that shear stress is sufficient to erode the upstream end of the levee channel. As the channel erodes the shear stress declines, but the proportion of back-flow from the flood recession increases. The result is that the slope of the levee channel reverses to slope towards the river. This is a new mechanism, and it is critical in the sequence of avulsion development. The final stage of the development of avulsion is when a knickzone moves up the levee channel joining another levee channel that is leaving the river upvalley. This is new mechanism of avulsion likely to operate in low energy river systems dominated by fine-grained sediments.