Spatial Distribution of Diapycnal Diffusion of Heat and Buoyancy in the Indian Ocean

Friday, 19 December 2014: 11:20 AM
Tonia astrid Capuano1, Lynne D Talley2, Julie McClean3, Alison M Macdonald4 and Caitlin Whalen3, (1)University of Western Brittany, Brest, France, (2)University of California San Diego, La Jolla, CA, United States, (3)Scripps Institution of Oceanography, La Jolla, CA, United States, (4)WHOI, Woods Hole, MA, United States
The deep Indian Ocean’s role in the global overturning circulation is to upwell abyssal waters to deep and thermocline waters. This requires diapycnal diffusion of buoyancy downwards. The required diapycnal diffusivity ("kappa") is on the order of the Munk value 10ˆ-4 mˆ2/sec, but the spatial variability of kappa is enormous, ranging from greater than 10^-3 to less than 10^-6 mˆ2/sec. Moreover, large diapycnal mixing does not necessarily result from large diffusivity. We use the finescale parameterization for diapycnal diffusivity with WOCE CTD/LADCP profiles to map dissipation, diffusivity, and diapycnal mixing of heat, salt and buoyancy relative to topography, isoneutral surfaces that span the water column, and monsoon phase (northern Indian). Largest diapycnal mixing is found in the most energetic regions: downstream of Kerguelen Plateau, in the Agulhas, in the complex of western boundary currents along Madagascar, Somalia and the Arabian Sea, and along the eastern boundary (throughflow, Australian coast and Andaman Sea). Broad mid-basin diapycnal mixing occurs in the Madagascar and Mascarene Basins. Regions of high deep diffusivity but low diapycnal mixing include the Central Indian and West Australia Basins where the deep waters have low stratification. Regions of lowest mixing and lowest diffusivity, top to bottom, are the Bay of Bengal and central Arabian Sea. Upper ocean diffusivities are much lower than in the deep ocean, and match published estimates from Argo profiles (Whalen et al., GRL 2012). Diapycnal diffusivities and mixing from the high resolution POP general circulation model, which includes some but not all physical processes responsible for diapycnal mixing, are compared with these observations, providing insight into the multiple mechanisms controlling diapycnal mixing.