[1] Using numerical experiments, we investigate how river-dominated delta channel networks are likely to respond to changes in river discharge predicted to occur over the next century as a result of environmental change. Our results show for a change in discharge up to 60% of the initial value, a decrease results in distributary abandonment in the delta, whereas an increase does not significantly affect the network. However, an increase in discharge beyond a threshold of 60% results in channel creation and an increase in the density of the distributary network. This behavior is predicted by an analysis of an individual bifurcation subject to asymmetric water surface slopes in the bifurcate arms. Given that discharge in most river basins will change by less than 50% in the next century, our results suggest that deltas in areas of increased drought will be more likely to experience significant rearrangement of the delta channel network.
Motivation[2] An immediate impact of climate change and human modifications to river catchments is that the long-term average discharge of most rivers will change appreciably over the next century [Nohara et al., 2006;Palmer et al., 2008]. On rivers that terminate in marine or lacustrine basins, this hydrological change also will impact their deltas, and yet no framework exists for predicting the adjustments deltas may undergo as discharge changes. Relative sea-level rise is already threatening the world's deltas [Blum and Roberts, 2009;Syvitski et al., 2009], and that problem could be further compounded if permanent changes in discharge also dramatically alter delta landscapes, which provide wetlands, biodiversity,, and homes to a significant part of the world's population [Coleman, 1988;Day et al., 2007]. Studies to date have focused on identifying which deltas are at risk [Day et al., 1995;Ericson et al., 2006] and how they might respond to perturbations such as sea-level rise [Jerolmack, 2009]. But, how deltaic distributary networks will respond to changes in river discharge remains unknown.[3] As the long-term average river discharge (Q) changes, the most dramatic response of a deltaic channel network, apart from a full avulsion caused by delta-lobe switching [Coleman, 1988], is abandonment or initiation of distributary channels. This scenario could be expected because the number of bifurcations in a distributary network correlates positively with the input Q and inversely with the power in the marine environment [Syvitski and Saito, 2007]. However, it is unclear whether an individual delta's response to changes in Q would follow the same trend between Q and channel number observed from the Syvitski and Saito delta compilation.[4] Past work has shown that the splitting of discharges at channel bifurcations is subject to fairly stringent stability conditions. Two distributaries with identical downstream boundary conditions and a given upstream Shields stress (Q) distribute water and sediment asymmetrically because this is a stable equilibrium configuration [Wang et al., 1995;S...