Inferring upper ocean dynamics from horizontal wavenumber spectra: Insights from Drake Passage

Thursday, 18 December 2014: 9:15 AM
Cesar B Rocha1, Teresa K Chereskin1, Sarah T Gille1 and Dimitris Menemenlis2, (1)Scripps Institution of Oceanography, La Jolla, CA, United States, (2)Jet Propulsion Laboratory, Pasadena, CA, United States
The dynamics governing the upper ocean at submeso to meso scales (~1-300 km) are a topic of recent debate. Horizontal wavenumber spectra of kinetic energy (KE) estimated from satellite data and global models rolloff as k-2 which have been interpreted as evidence for surface quasi-geostrophic (SQG) turbulence. Shipboard ADCP observations in the Gulf Stream, however, indicate steeper (k-3) KE spectra in the 20-200 km range and a ratio of across to along-track KE components of ~3, consistent with quasi-geostrophic (QG) turbulence. We use 13 years of shipboard ADCP measurements, altimeter data, and a new internal-tide-resolving and submesoscale-admitting Massachusetts Institute of Technology general circulation model (MITgcm) simulation to study the horizontal scales of variability of the Antarctic Circumpolar Current in Drake Passage. Horizontal wavenumber spectra of KE in the 10-200 km range from Drake Passage ADCP observations rolloff as k-3, reminiscent of QG turbulence. The average ratio of the KE components, however, is <2. Analyses of the MITgcm numerical simulation and stochastically-generated vector velocity fields suggest that the presence of ageostrophic, horizontally-divergent motions significantly alters the QG spectral characteristics, in qualitative agreement with the ADCP KE spectra. Ageostrophic flows, likely dominated by internal tides, appear to project onto scales as large as 150 km but are more dramatic in the 10-40 km subrange where the KE component ratio is ~1. The simplest interpretation consistent with the available data is one of QG turbulence co-existing and interacting with unbalanced motions to generate the observed upper-ocean KE spectrum in the 10-200 km range. We find no signature of SQG turbulence.