Experiments with a Global Ocean Model Driven by Observed Atmospheric Forcing

Abstract

A global ocean model with a 5° horizontal grid size, previously used in the coupled climate model of Washington et al. (1980), is used here with observed monthly mean wind stress and atmospheric temperatures for the purposes of varying model parameters to suggest values for future coupled climate model experiments, and examining heat storage and transport in the model ocean. A control run is carried out with the same internal parameters as in the climate simulation of Washington et al. (1980). Observed atmospheric forcing produces an oceanic circulation closer to the observed than in the coupled climate model. Necessarily large horizontal momentum diffusion leads to weak zonal currents in both cases; however, the stronger observed wind stresses in the decoupled model produce more vigorous and more realistic meridional ocean currents. Three additional ocean experiments test parameter variations. Lowering horizontal heat diffusion increases north-south temperature gradients and intensifies ocean circulation. Reducing vertical mixing has little impact on currents but makes the thermocline shallower. Decreasing wind forcing results in slightly weakened geostrophic currents and considerably less Ekman transport. The magnitude and phase of midlatitude heat storage are adequately simulated. Wind forcing and associated upwelling and downwelling dominate tropical heat storage in the model. Semiannual features of tropical heat transport are reproduced by the model ocean as a result of using a full annual cycle of observed wind stress forcing. Meridional cells on either side of the equator fluctuate in intensity twice a year and are out of phase with each other. This produces a semiannual cycle of tropical heat transport similar to that observed, and is associated with the cross-equatorial excursion of the stronger cell. The sign of individual ocean basin heat transport from the model agrees with observed estimates, including northward heat transport in both the North and South Atlantic. Reduced horizontal heat diffusion in the model produces heat transport closer to observed estimates. Thus, results from the parameter experiments and heat transport studies provide insight into the sensitivity of the ocean model to wind stress and heat diffusion, and contribute to understanding the real ocean processes and assessing future ocean-atmosphere coupled climate model simulations.