Modeling and Assimilation of Oceanographic Measurements Collected by Tagged Brandt’s Cormorants at the Mouth of Columbia River

Abstract:

Animal biologging tags are typically used to track animal movements and behavior, but marine birds also offer a distributed, autonomous, and scalable platform for collecting oceanographic measurements in remote coastal regions. The Cormorant Oceanography project aims to develop such an observational approach using tagged cormorants (diving birds) and to incorporate heterogeneous data using inverse modelling. Extensive bird tagging experiments have been undertaken at the mouth of Columbia River in which tags attached with pressure and temperature sensors were deployed on cormorants. These birds forage for their food by diving to the seafloor, up to 60-70 m depth. Brandt’s cormorants tend to dive benthically, mapping the bottom contours while they forage. In between dives, the birds’ natural behavior is to drift while resting at the sea surface; GPS data were used during these resting periods to obtain surface velocity estimates. The resulting unique data set provides high spatiotemporal 3D coverage of the study region. Here, we show results of assimilating a subset of surface velocities and depth soundings into a high-resolution numerical model, in order to predict coastal bathymetry with inverse methods. We used methods previously verified in well-mixed conditions, but which have not yet been applied in large high-energy stratified estuaries such as the mouth of the Columbia River. Results will be presented on depth inversions using the method, including tests of a novel iterative inversion scheme. This method can be extended to include other available data sets such as biologging derived temperature and salinity, satellite imagery, or drifters. Preliminary idealized tests will be presented assessing the relative value of these different observation types, as well as data from different tidal stages or locations throughout a coastal/estuarine environment.