Abstract
Generally, ports in the North American Great Lakes are not supported with navigational guidance (water level, water temperature, currents, ice) by NOAA’s Great Lakes Operational Forecast System (GLOFS). To examine the benefit of extending model coverage to this critical infrastructure, a linked hydrologic-hydrodynamic framework was developed for the Twin Ports of Duluth-Superior in western Lake Superior, and tested over three case studies of flooding due to storm surge and/or river flooding. Streamflow from 22 National Water Model (NWM) simulated and 3 gauged inflows within the domain was injected into a hydrodynamic model built on the Finite Volume Community Ocean Model (FVCOM), with a wet-dry grid that covered the harbor and surrounding floodplain. Model results from flood simulations compared well against time-series of water level and streamflow at gauges within the harbor, with root mean square error (RMSE) and bias values small relative to the typical fluctuations. Inclusion of NWM-simulated tributaries improved the accuracy of modeled water levels in the harbor, and increased simulated current speeds in navigational channels by up to 0.5 ms −1 . Modeled inundation extent in the floodplain closely matched flood extent surveys conducted in response to a record flood event in 2012, demonstrating the capability of the modeling framework to provide flood guidance in the complex coastal setting.