Analysis of the scale dependence of the atmosphere-ocean thermal coupling using cross-spectral methods.

Lucas Laurindo1, Leo Siqueira2, Benjamin P Kirtman3, Igor V Kamenkovich2 and Richard J Small4, (1)RSMAS, Department of Atmospheric Sciences, Miami, FL, United States, (2)RSMAS, Miami, FL, United States, (3)University of Miami - RSMAS, Atmospheric Sciences, Miami, FL, United States, (4)NCAR, Boulder, CO, United States
Abstract:
Although previous studies of the air-sea thermal coupling identified the main differences between the atmosphere-driven regime at large scales (O[102-103 km]) and the ocean-driven at mesoscales (O[101-102 km]), the causes of this scale dependency remain unclear. This work investigates the physical mechanisms behind the cross-scale transition by examining the linear relationship between satellite estimates of sea surface temperature (SST) and turbulent heat flux (THF) and of SST-tendency and THF over scales between 102-104 km and 101-103 days using cross-spectral statistics, interpreted under light of correspondent analytical solutions derived from a simple thermodynamic model for the coupled air-sea system.

The obtained analytical solutions are found to be linearly dependent on the power spectral density of the model’s forcing terms associated with ocean and atmosphere dynamics, and help to distinguish between the atmosphere- and ocean-driven coupling regimes in the cross-spectra computed from satellite-based estimates of SST/SST-tendency and THF. Similar to recent conclusions drawn for SST and 10-m wind speed, our new results show that the transition between regimes occurs at wavelengths near the atmospheric first baroclinic Rossby radius of deformation, and that the dispersion of correlated signals indicative of the ocean-driven regime resembles that of oceanic planetary waves. We further demonstrate that the use of realistic spectral variance distributions for the theoretical model’s oceanic and atmospheric forcing signals enables the model to reproduce the observed cross-scale transition reasonably well. This correspondence suggests that the transition is linked to the distinct spectral power distribution of intrinsic variability in each medium, occurring because (a) the atmospheric SST modulation is significant at large scales but weaker over the ocean mesoscales, mainly because the mesoscale wind variance is itself weaker; concurrent with (b) the strong ocean-driven SST modulation over mesoscale ranges.