Improving Representation of the Coastal Transition Zone for the Next-Generation Great Lakes Operational Forecast System

The NOAA Great Lakes Operational Forecast System (GLOFS) is a set of three-dimensional hydrodynamic models built on the Finite Volume Community Ocean Model (FVCOM) that provide near-realtime forecast guidance on currents, water temperatures, short-term water level fluctuations, and ice cover throughout the Great Lakes. GLOFS supports commercial navigation, flood preparedness and coastal resiliency, and recreation, among other uses. However, these models have traditionally treated the shorelines as static boundaries, whereby many ports, harbors, and estuaries are not resolved in the model domain, and where there is no representation of the floodplain. There is a need for improved forecast guidance in the Great Lakes coastal transition zone, which has become increasingly pressing in recent years, particularly around Lake Ontario where record-high water levels in 2017 and 2019 led to extensive compound flooding along the coasts. With funding from the Bipartisan Infrastructure Law (BIL) Coastal and Inland Flood Inundation Mapping (CIFIM) provision, we are extending the GLOFS domain to include the coastal floodplain, starting with Lake Ontario, and coupling with inflows from the NOAA National Water Model. The upgrade will improve simulation of compound flooding along the Great Lakes coasts by allowing the shoreline to move dynamically in response to forcing conditions, and providing improved representation of very-nearshore processes in the hydrodynamic domain.

While linking the hydrologic and hydrodynamic model domains is often viewed as a data coupling process, the spatial treatment of rivers within the floodplain has direct implications on the representation of coastal processes. In a purely triangular mesh, the horizontal resolution of the hydrodynamic mesh dictates the lower spatial scales at which hydrologic-hydrodynamic coupling within the floodplain can be achieved. The inclusion of inflows with poorly-resolved thalwegs can negatively impact model results, particularly as it relates to simulations of areal inundation extent. The Semi-implicit Cross-scale Hydroscience Integrated System Model (SCHISM) implements a flexible mesh scheme in which quadrangles and skewed triangles can be included in the mesh, making it possible to resolve key features in the domain while maintaining computational efficiency. We present progress on the development of a next-generation flood forecast system for Lake Ontario using the SCHISM modeling platform with a flexible mesh scheme. In particular, considerations for spatial representation of the coastal transition zone are explored and discussed, with results comparing different approaches and model configurations.