A Nowcast/Forecast System for Coastal Ocean Circulation Using Simple Nudging Data Assimilation

Abstract

This study describes the establishment of a Nowcast/Forecast System for Coastal Ocean Circulation (NFS-COC), which was run operationally on a daily basis to provide users ocean surface currents and sea levels that vary with synoptic winds, and seasonal and mesoscale variability intrinsic to the Florida Current. Based on the requirements of users, information about possible oil spills, trajectories, etc., is also provided by NFS-COC. NFS-COC consists of two parts: a 3D ocean nowcast/forecast circulation model, Princeton Ocean Model (POM), and a 2D trajectory model. POM is automatically run to forecast ocean variables for up to 2 days under forcing of the Florida Current inflow/outflow and the predicted surface winds, which are automatically transferred (by ftp) from a file server at the National Meteorological Center (now known as the National Centers for Environmental Prediction). The winds from the mesoscale Eta Model are called Eta winds. Then the trajectory model is run to predict the path due to 1) the POM-predicted ocean surface currents, 2) wind drift due to the predicted Eta winds, and 3) turbulent dispersion based on a random flight (Markov process) model. The predicted surface trajectories can be used to estimate the physical transport of oil spills (and other drifting or floating objects) in the Straits of Florida and many other coastal seas. A simple data assimilation scheme (nudging to the volume transport) is designed into the NFS-COC, although some powerful data assimilation methods exist for assimilating other physical variables.