Assessment of hurricane wind stress estimation using a 4DVAR ocean assimilation system

This contribution considers the problem of using ocean observations to improve the estimation of wind stress in hurricane conditions. We use 4DVAR ocean state estimation to estimate the wind stress that is most consistent with a set of ocean observations. This system consists of a regional high-resolution configuration of the Massachussets Institute of Technology General Circulation Model and its adjoint. The model is forced with an idealized but realistic westward tracking hurricane wind stress and initialized with horizontally uniform stratification.

In this preliminary study, psuedo data were extracted from the reference run with the idealized hurricane forcing. The model was re-run with a different stress field timeseries which produced values which did not match the reference pseudo data. The reference pseudo data were used with the adjoint method to improve the model consistency with the data by adjusting the wind stress fields. Ideally, the fit to the data will produce adjusted wind stress fields that match the reference model run.

Several experiments have been performed to determine the sensitivity of the adjusted wind stress timeseries to assimilation of various subsets of ocean state variables such as SST or surface current. The observations were simulated both as occurring at fixed space points and along drifter-type trajectories, as representative of realistic ocean sampling programs.