Constraining the Surface Boundary-Layer Stratification Budget in the Eastern North Pacific Subtropical Front

John Mickett, University of Washington, Applied Physics Laboratory, Seattle, United States, Eric L Kunze, NorthWest Research Associates, Redmond, WA, United States, James B Girton, Applied Physics Laboratory, University of Washington, Seattle, WA, United States and J. Thomas Farrar, Woods Hole Oceanographic Inst, Department of Physical Oceanography, Woods Hole, United States
Abstract:
We describe details of the depth-dependent stratification budget of the ocean surface boundary layer (SBL) in the eastern North Pacific Subtropical Front (35°N, 139°W) in the presence of significant lateral density gradients using 4-D (x,y,z, and time) upper-ocean observations collected over 7 days during March 2017. Data were collected with ship surveys (vessel-mounted ADCP, meteorological observations, and a towed profiling CTD system), a drifting surface air-sea flux buoy, and an array of 22 synchronized EM-APEX profiling floats. In addition, we use the Price-Weller-Pinkel (PWP) 1-D mixed-layer model to assist in evaluating the importance of various budget terms. Both model and observations indicate vertical processes dominate evolution of the stratification the ~60-70 m thick SBL, with penetrative solar radiation and vertical mixing being the primary terms. As penetrative shortwave decays roughly exponentially, below ~40 m the direct influence of this term is weak. The dense 4-D observations allowed reliable estimates of lateral advection contributions, which are dominated by inertial shear acting on relatively steady lateral density (buoyancy) gradients. Although this term could be important on short timescales (< 2π/f), particularly early in the time-series when inertial motions and shear were strongest, because of the oscillatory nature of the shear, the 7-day cumulative influence was small compared to other terms. Overall, it is concluded that the typical factor-of-two or more underestimates of lower-SBL stratification by the PWP model may be a consequence of missing physics in the model, rather than lateral processes. We evaluate several hypotheses for this model shortcoming.