Enhancing Total Water Level Forecasting for the Great Lakes and Lake Champlain using the Next Generation Water Resources Modeling Framework (NextGen)

Accurate forecasting of total water levels (TWL) is crucial for managing the Great Lakes and Lake Champlain, where their fluctuations significantly impact ecosystems, navigation, and infrastructure. With increasing concerns over climate change and its effects on lake water levels, the need for precise and reliable TWL forecasts has become more pressing. NOAA’s Office of Water Prediction (OWP) is extending its TWL forecasting capability to these regions using the Next Generation Water Resources Modeling Framework (NextGen). This framework is an open-source, model-agnostic water resources model interoperability infrastructure that allows simultaneous operation of multiple model formulations interacting with a unified hydrofabric. The Semi-implicit Cross-scale Hydroscience Integrated System Model (SCHISM) and the D-Flow Flexible Mesh (D-Flow FM) model have been selected as NextGen coastal modules to predict TWL. Freshwater inflows are provided by the National Water Model (NWM) Tree-Based Channel Routing (T-Route) scheme, meteorological forcings are sourced from the NextGen Forcings Engine, and lake water levels are provided by the Finite Volume Community Ocean Model (FVCOM). OWP has developed Basic Model Interface (BMI) applications for these models to run in a two-way coupled fashion under the NextGen framework for the Great Lakes region.

This study focuses on optimizing SCHISM performance across five model domains: Lake Erie, Lake Michigan-Huron, Lake Ontario, Lake Superior, and Lake Champlain. To enhance model performance, we refine and adjust the model meshes to better capture localized processes. Additionally, USGS-recorded flow rates at key riverine outlets, including St. Marys River, St. Clair River, Detroit River, St. Lawrence River, and Niagara River, are incorporated. Model performance is evaluated using gauge observations from NOAA, USGS, and Environment Canada. Combining USGS observed flow rates with NWM modeled discharge as freshwater inflow results in improved TWL accuracy near lake outlets. Mesh refinements also contribute to enhanced local TWL accuracy.