Anna van Yperen, John M. Holbrook, Miquel Poyatos-Moré , Ivar Midtkandal

The backwater effect (i.e. channel flow influence by a body of standing water) is used to predict down-dip changes in fluvial morphodynamics and consequent sediment distribution on delta plains. These changes include downstream fining, decrease in sinuosity, and deepening and narrowing of channel belt deposits. This study reviews existing methods for estimating backwater length in ancient and modern settings and proposes workflows to minimize ambiguity in resultant estimates.
The proposed workflows are tailored to both modern and ancient settings and are prioritized based on practicality, accuracy, smallest uncertainty ranges and allow different types of data as input parameters. In modern river systems, we recommend using direct field measurements of bankfull thalweg channel depth and river water elevation to determine the location where riverbed elevation intersects sea level (i.e. the upstream limit of the backwater zone). Alternatively, the backwater length (Lb) can be estimated indirectly by Lb = h/S, with h is bankfull thalweg channel depth and S is slope. In ancient settings, bankfull thalweg depth and grain size representative of bedload transport are the most reliably measurable parameters, obtained at one or a few locations.
For the first time, the application of multiple methods to obtain backwater length estimates are tested on a single modern and ancient river system. In the modern case study, the riverbed intersection with sea level matches previously documented major changes in sedimentary trends, such as decreasing channel-belt width/thickness ratios, decreasing meander-bend migration rates, and coarsening grain size followed by distinct downstream fining. However, backwater lengths based on h/S plot downstream of this zone characterized by major changes, when input parameters are derived from discharge and grain size. Therefore, we recommend obtaining bankfull thalweg channel depth from a cross-sectional profile if backwater length is estimated based on h/S. In the ancient case study, bankfull thalweg channel depth derived from fully preserved single story channel fill and slope based on Shields’ empirical relation with grain size, match changes in fluvial architectural style interpreted as a result of backwater effects. Although uncertainty management is improved with the proposed workflows, a degree of uncertainty remains in the resulting backwater length estimates, due to inherent scatter in previously established relationships (e.g. Shields stress relation to obtain slope estimates).
This review is a critical step forward in discussing the shortcomings, and listing and acknowledging the uncertainties and ambiguity in obtaining the necessary input parameters to estimate backwater lengths. The proposed workflows facilitate comparability and applicability of future backwater length estimates and their corresponding influence on the hydrodynamic environment and ultimately the stratigraphic record. Potential scaling relationships between the backwater length, sedimentary trends and avulsion nodes makes this of key importance as the latter two also play a crucial role in devastating floods when rivers change course.