Connectivity and directionality in estuarine channel networks

Estuaries host channel networks that can range from meandering single-thread channels to complex channel networks comprising looping, branching, and offshoot structures through which water, sediment, and nutrients are transported in both the flood and ebb directions. In this article, we use graph theory to quantify the structural and dynamical connectivity of multidirectional estuarine channel networks using network analysis techniques rooted in graph theory that have proven useful for quantifying connectivity in river deltas. Networks from several estuaries around the world are extracted from satellite imagery and compared to a set of schematized networks that represent the end-member of the channel network structure and dynamics found in estuaries. Higher levels of structural connectivity are found for larger networks, which points to increased predilection for looping structures in networks with large numbers of channels. The real-world network structures contain signatures of both mutually evasive flood and ebb channels that typify alluvial estuaries, but also contain branching structures as either bifurcating delta-like structures or converging tidal patterns. The level of dynamical connectivity is modulated by flow direction through the network, with flood direction fluxes more broadly distributed throughout the network and fluxes in the ebb direction tending to be more localized. Analysis of dynamical connectivity also reveals a tidal asymmetry indicative of fluxes in flood-ebb dominant channels in estuaries. This study provides implications for understanding the self-organization of estuarine channel networks on which fluxes are partitioned through estuaries in the flood and ebb directions, and establishes a benchmark for analyzing multidirectional channel networks using graph theory.

Automated summary

This article uses graph theory to quantify the connectivity of multidirectional estuarine channel networks, finding higher levels of structural connectivity in larger networks with looping structures. Real-world networks contain signatures of both mutually evasive flood and ebb channels, as well as branching structures, with flow direction influencing dynamical connectivity.