Large-Scale Variability of the Main Thermocline Excited by Stochastic Wind Stress Forcing

Abstract

To investigate the role of stochastic forcing in generating decadal variability in the main thermocline circulation, an isopycnic numerical model is forced by the standard monthly climatological forcings plus daily wind stress anomalies. Two statistical models of daily wind stress anomalies are developed. The first one simulates both the temporal and spatial second moment statistics of the observations, whereby it produces realistic daily wind stress anomalies. The second model produces spatially uncorrelated wind stress anomalies that have the observed local variance at each grid point. The results show that a basinwide mode, characterized by a dipole structure in the heat content of the North Pacific, can be excited stochastically. The structure of the mode does not represent an imprint of the atmospheric forcing but is selected by the ocean itself.