Obtaining synthetic riverbed topography of meandering rivers from satellite imagery: a case study of the Tallahatchie River, Mississippi

Riverbed topography is critical for the study of sediment transport, river planform evolution, channel regulation, hydrodynamic simulation, and flood risk evaluation. However, both present and historical riverbed topography information are not always available. In this work, we provide a framework to obtain synthetic riverbed topography of meandering rivers from satellite imagery through two python-based packages, geeRiverCL and pyRiverBed. geeRiverCL is developed for river centerline extraction from spectral images with traditional spectral bands (i.e., red, green, blue, and near infrared). Compared with other existing tools, geeRiverCL allows one to identify river masks for centerline extraction with less spectral bands or lower spatial resolution. The vectorized one-pixel-wide centerline obtained from geeRiverCL is one of the model inputs of pyRiverBed. pyRiverBed is designed to compute the riverbed topography at meander curvatures under a morphologically stable state using the Beck's (1988) equations, which require three river channel characteristics, mean water depth, mean channel width, and channel curvature.

In order to calibrate the model and estimate the performance, we implemented this framework to obtain the synthetic bed topography of the Tallahatchie River, a meandering river with a near-constant width in the Yazoo-Mississippi Delta, and compared the results with measured datasets from multibeam sonar. The results show that the centerline extracted from the remote sensing imagery matches well with the centerline from measured data and the difference between the synthetic and measured riverbed elevation is in a reasonable range.