OS033-04
The need for high vertical resolution measurements along the equatorial Pacific to improve estimates of ocean mixing
The need for high vertical resolution measurements along the equatorial Pacific to improve estimates of ocean mixing
Friday, 11 December 2020: 16:14
Virtual
Abstract:
Errors in sea surface temperature (SST), such as the cool bias of the cold tongue in the equatorial Pacific and the warm bias off the coast of South America, are persistent features in numerical simulations with ocean-only and coupled atmosphere-ocean models. Many factors are postulated to contribute to these errors, among them the representation of vertical mixing in ocean models. Previous modeling studies have demonstrated that a change in the specification of the vertical diffusivity along the equatorial Pacific modifies not only the local temperature structure but also the temperature structure eastward following the equatorial wave guide and poleward along the eastern boundary following the coastal wave guide. When the change in temperature occurs on density surfaces that outcrop to the surface, it impacts SST. With numerical ocean model experiments conducted at different vertical resolutions, we show that sufficiently high vertical resolution in the upper ocean captures much of the observed fine-scale vertical shear and stratification, associated with inertia-gravity waves and instabilities. These fine vertical scale structures modify the spatial distribution and temporal variability of the vertical diffusivity, as determined by a Richardson number-based parametrization scheme. On a long-term average, the change in the vertical diffusivity produces a warming of the SST in the equatorial cold tongue and a cooling off the South American coast, thus reducing the commonly seen cool and warm biases in the two regions respectively. Our results emphasize the need for high vertical resolution O(1m) measurements throughout the thermocline along the equatorial Pacific to improve estimates of ocean mixing, leading to improved mixing parameterizations in ocean models for climate research.