Tuba Özkan-Haller and PhD student Greg Wilson are working on a model to predict river bathymetry. Wilson had investigated the relation between surf zone currents and the underlying bathymetry and developed a method to predict what the bottom was like using measurements of the overlying currents.
An Office of Naval Research program manager became interested in this bathymetric inversion method and asked Özkan-Haller if it could be used in rivers.
Özkan-Haller and Wilson are testing the Regional Ocean Model System (ROMS) to predict river bathymetry. For a modern dataset, they partnered with researchers in the COHSTREX (Coherent Structures Experiment) project who obtained a detailed bathymetry of the Snohomish River in Washington in order to study small-scale turbulence. Özkan-Haller is interested in using the larger-scale flow through river bends, and such data were also collected; COHSTREX researchers will, in turn, receive the model that Özkan-Haller produces.
Calculating bathymetry in rivers must take into account both the total flow rate and curvature. If rivers were straight, the fastest flow would always occur over the deepest part of the channel. However, at river bends, the fastest flow is not at the outer edge where the water is often deepest. Curvature effects move the fast flow towards the inner bank of the river. If observations of the flow across the channel width are available, this tug-of-war between the deep channel and the river bend can be exploited to get at information about the channel bathymetry.
Another potential user of such a inversion tool is the USGS. Curvature is easily observed (think Google Earth) and current velocity can also be measured remotely (a visible feature moving a certain distance over time). If the inversion method provides channel cross-section, the total flow rate in the river can be determined. This method would be a much cheaper alternative to current river gaging methods that often involve in situ observations.
Özkan-Haller is excited about the possibility of the community model ROMS being applied more broadly. "ROMS was developed as a shelf scale model. Then a lot of effort got put into adding a surf zone to it, to be able to do breaking waves and related circulation. If one model could go all the way from the freshwater river through an inlet, through the surf-zone breaking wave region, to the outer shelf, how cool would that be?"
