The role of relative floodplain width in forming anabranching rivers

Friday, 18 December 2015: 14:25
2005 (Moscone West)
Douglas A Edmonds1, Sara Eugenia Moron Polanco2 and Kathryn Amos2, (1)Indiana University, Geological Sciences, Bloomington, IN, United States, (2)Adelaide University, Australian School of Petroleum, Adelaide, Australia
Anabranching rivers (including anastomosing) are a relatively common channel pattern, especially among the world’s biggest rivers, and are defined as a system of multiple channels separated by immobile alluvial islands. The origin of anabranching remains poorly understood and is an important topic of research. Previous studies on Australian rivers and a recent empirical compilation show that floodplain width (relative to the size of the channel) might play an important role in the formation of anabranching rivers. To test this idea further we carried out two sets of morphodynamic simulations using Delft3D. In the first set we create a generic channel-floodplain complex with uniform floodplain and channel width, slope, and grain size and allow the system to adjust to passing floodwaves. In successive runs we hold all variables constant, except we increase floodplain width. Results of these simulations show a transition from single channel to braided to anabranching as floodplain width increases. Anabranching arises because as floodplain width increases, alluvial bar growth occurs on the floodplain. The emergence of bars causes flow bifurcation, and subsequent bifurcation instability leads to reduction of channels and the emergence of multiple anabranches. Transition to a stable anabranching pattern is achieved because as anabranches increase their cross-sectional area, Shields stresses on the intervening bars are reduced until they bars stop migrating. To test the idea that alluvial bar growth can be a precursor to anabranching we carried out a second experiment set using boundary conditions from four different field scale anabranching rivers. Results from these simulations show that anabranching can initiate from alluvial bar growth. Compared to field measurements of anabranching rivers our simulations accurately predict number of channels, supporting the idea that relatively wide floodplains might be an important attribute of anabranching rivers.