High resolution hydrodynamic models are computationally expensive to run – especially if ensemble forecasts are desired. This can be problematic within coastal estuaries which are not well resolved by today's operational meteorological forecast models. As an alternative, this paper evaluates the wind forcing for three setup parameterizations (based on the Zuiderzee, modified Zuiderzee, and long wave equations) using a combination of observed setup from in-situ water level gauges and local wind observations. In addition, three methods are explored for developing hourly time series of wind forcing from 5-min observations: top of the hour, hourly mean, and wind run approach. The wind forcings, which are weighted by the length of two lagoon-oriented axes, are used to drive the setup parameterizations. The observed setup is used to tune each of the parameterizations via a least squares approach. The observation spread, linear model residuals, coefficient of determination (R2), and root mean squared error (RMSE) indicate that the wind run out performs the other two methods. In terms of the three parameterizations, the modified Zuiderzee had consistently higher R2 values, lower RMSE, and narrower 95% confidence intervals than the two other methods. This optimized parameterization is currently being used operationally to generate ensemble setup forecasts for the Indian River Lagoon, a restricted estuary on Florida's east-central coast. These simple ensemble forecasts are designed to guide the National Weather Service (NWS) in identifying potentially significant setup events that warrant high resolution hydrodynamic simulations.