Variability of sea surface salinity in stochastically forced systems

Abstract

The influences of horizontal advection and horizontal diffusion on the variability of sea surface salinity in stochastically forced systems are investigated. Basic ideas are developed using a two dimensional box model and then extended to a more realistic three dimensional ocean general circulation model. It is shown that, in the absence of advection and diffusion, the ocean response is essentially that predicted by Taylor's random walk model. Advection becomes important when the advective time scale is less than the response time of the mixed layer to the stochastic forcing. Advection of parcels from regions of upwelling into regions of downwelling limits their exposure time to the stochastic forcing and thus the maximum attainable variance in the system (variance increases linearly with time). Regions of upwelling and downwelling may be introduced through the thermohaline overturning circulation or by the wind driven Ekman transport, depending on the specific model configuration. Horizontal diffusion is found to be important when the diffusive time scale is less than the mixed layer response time. The primary role of diffusion is to reduce the effective stochastic forcing through rapid mixing of uncorrelated surface forcing events. Because sea surface salinity does not have a negative feedback with the atmosphere, it is more strongly influenced by weak horizontal processes than sea surface temperature (SST). Accurate knowledge of the stochastic forcing amplitude, decorrelation time, and length scale and distribution are critical to model the variance of sea surface salinity. Aspects of the ocean model which strongly influence the variability of sea surface salinity include the surface velocity, horizontal diffusivity, and the mixed layer depth. Implications on modeling of the ocean and coupled ocean-atmosphere systems are discussed.