Key points are not available for this paper at this time.
Recent idealized studies have shown that both explicit horizontal diffusion and the implicit diffusion associated with the advection scheme in high-resolution z-coordinate models may drive unrealistically high rates of diapycnal mixing. The aim here is to see whether the diapycnal mixing associated with the advection scheme in a global eddy-permitting (¼° by ¼°) z-level model is sufficiently strong to corrupt the thermohaline circulation. This paper diagnoses the diapycnal fluxes by using the ideas of water mass transformation. In the Southern Ocean, the model deep and bottom waters drift rapidly away from the Levitus climatology, with dense isopycnals moving downward at rates of up to 35 m yr−1. The strong upward flux (up to 50 Sv) through the dense isopycnals cannot be explained by the incorrect surface forcing (as a result of poor surface fluxes and no ice model) as most of the anomalous diapycnal fluxes are occurring in the deep ocean far from surface forcing. Hence, the excessive diapycnal flux is driven by diffusion in the model, both explicit and implicit. The "effective" diapycnic diffusivity driven by the numerical diffusion (associated with the horizontal advection scheme) is found to be the same order, 1–10 cm2 s−1, as that driven by the explicit horizontal diffusion. For strong vertical velocities (∼20 m day−1) as in models forced by high frequency winds, the vertical advection scheme also gives similar effective diffusivities. These effective diffusivities are considerably greater than suggested by observations. To alleviate these problems, it is suggested that eddy-resolving z-level climate models will require 1) less diffusive horizontal advection schemes and 2) better vertical resolution throughout much of the water column.
Lee et al. (Wed,) studied this question.