Steady-State Diagnostic Model of the New York Bight

Abstract

A qualitative evaluation is made of the output from a finite-element, steady-state diagnostic model to observed time-averaged currents. The model uses a vorticity balance equation with linear bottom friction and inputs observations of near-bottom currents on the model boundary, density field and bottom topography. The output is the near-bottom (barotropic) velocity field over the entire modeled region. Velocity profiles are constructed using the thermal wind equation with the observed density field from May 1976 and a turbulent closure scheme model of Mellor and Durbin to reproduce the top and bottom Ekman layers. Transport is computed in layers above and below the pycnocline by integrating the geostrophic velocity profile and adding the Ekman layer transport. Comparisons of the modeled bottom velocities at three moorings interior to the region and modeled vertical profiles of velocity at the interior moorings and the four boundary moorings to the observations at those points, show favorable agreement. Along-isobath flow is modeled more accurately than cross-isobath flow. Along-isobath flows, both in the shelf valley and outside the valley, are well represented, but near-bottom flows in the valley in two of the patterns are strongly ageostrophic and thus are not in agreement with model results. Errors >100% are found for weak flow events and the accuracy of the vertical shear in the velocity deteriorates away from the time of the density observations. The model results are useful for calculating the advective transport of dissolved and suspended constituents in the water. Though the accuracy of the point velocities have a median relative error of ∼50%, the transport calculation is probably more accurate.