CoAuthors

Shane Elipot (University of Miami, United States); Jonathan Brasch (University of Michigan, United States); Dimitris Menemenlis (NASA JPL, United States); Aurelien Ponte (IFREMER, France); Jay Shriver (Naval Research Laboratory, United States); Xiaolong Yu (IFREMER, France); Edward Zaron (Oregon State University, United States); Matthew Alford (Scripps Institution of Oceanography, United States); Maarten Buijsman (University of Southern Mississippi, United States); Ryan Abernathey (Columbia University, United States); Daniel Garcia (University of Michigan, United States); Lingxiao Guan (University of Michigan, United States); Paige Martin (University of Michigan, United States); Arin Nelson (University of Michigan, United States)

Surface oceanic kinetic energy is of interest for many reasons, and a greater understanding of the vertical structure of the kinetic energy would aid interpretation of ongoing and proposed remote sensing missions that are focused on velocity.  In this work, kinetic energy (KE) at the sea surface (0 m) and 15 m depth in high-resolution global simulations (HYbrid Coordinate Ocean Model; HYCOM, and Massachusetts Institute of Technology general circulation model; MITgcm) is compared with KE from undrogued and drogued drifters, which respectively represent flows at 0 and 15 m. Global maps and zonal averages are computed for four frequency bands—low-frequency (<0.5 cpd), near-inertial, diurnal, and semi-diurnal. In the near-inertial band, MITgcm KE is too low relative to drifters, while HYCOM KE lies closer to observations probably due to more frequently updated atmospheric forcing. In the semi-diurnal band, MITgcm KE is too high, while HYCOM KE lies closer to the drifters due primarily to the inclusion of a parameterized topographic internal wave drag. In both drifter and model results, semi-diurnal spectra display a weak dependence on background low-frequency KE, consistent with mesoscale eddy-induced internal tide non-stationarity. Drifter semi-diurnal spectra are inherently wider than model semi-diurnal spectra, due to Lagrangian sampling of spatially varying fields. Vertical structure is defined here by the ratio of zonally averaged KE in 0 m/15 m model results and undrogued/drogued drifter results. Over most latitudes and most frequency bands, the model ratios track the drifter ratio to within error bars. All of the frequency bands except for the semi-diurnal band display measurable vertical structure. Latitudinal dependence in the vertical structure is greatest in the diurnal and low-frequency bands.