NRL conducts research and development on mesoscale ensemble prediction. This area of research is necessary because inherent uncertainties in model initial conditions, lateral boundary conditions, and model physics exist and the effects of these uncertainties can only be measured and accounted for through probabilistic prediction. Our research addresses the following areas of ensemble modeling: (1) initial condition perturbations, (2) lateral boundary condition perturbations, (3) perturbations of model physics, and (4) ensemble post-processing/metrics. This research is performed using the nested, nonhydrostatic atmospheric model in the Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS®). The ensemble forecasts are used to enhance forecast guidance of high-impact weather phenomena, particularly in coastal environments.