Using Ensembles to Account for Uncertainty in Atmospheric Forcing in the Mississippi Bight

We compare three different ocean model ensembles in the Mississippi Bight, cases where the ensemble spread is introduced by applying a time shift to all or some of the fields in the atmospheric forcing. The difference between the ocean ensemble models is which one-way, uncoupled atmospheric forcing variables are linearly time-shifted between -12 and +12 hours for any given day, where the time shift is intended to account for the uncertainty in the phase of processes generating an inertial oscillation response in the ocean currents. The different experiment cases apply the time shift spread to (1) all atmospheric forcing variables, (2) wind speed components only, and (3) wind speeds and atmospheric pressure. We do not time shift the heat flux terms in the second and third cases because we expect little uncertainty in the timing of heat fluxes to be dominated by the high confidence in the solar radiation terms. We evaluate how well the members of each ensemble case represent the phase of near-inertial oscillations observed using high frequency (HF) radar data and moored current meters. While the response of ocean currents to the time shift in atmospheric pressure would be small in the open ocean, the nearshore response is amplified by the coastline in the Mississippi Bight.