A Coupled Empirical-Numerical Model for a Buoyant River Plume in Lake Michigan

Abstract

A coupling technique is developed to predict the behavior of a buoyant river plume in a lake. The model incorporates a 3D hydrodynamic model (POMGL) and a 3D particle tracking model (Partic3D) for the far-field transport computations. The source conditions for the particle tracking model are obtained from a near-field model derived from the characteristics of the plume analyzed from extensive field studies on the Grand River plume, Lake Michigan. The empirical near-field model was developed to predict the geometry of the plume, dilution, and centerline trajectory near the river mouth, and to provide the concentration and location of the particles to be released in the far field. The coupled empirical-numerical model shows improved predictions in the near field versus the single numerical model. The present results strongly advocate the use of model combinations in order to improve coastal diffusion and transport processes. The primary application of the technique is in recreational water early-warning and forecasting systems that will estimate the immediate and short-term risk of exceeding pathogen indicator concentration criteria in lakes and coastal areas.