(6)Barotropic boundary conditions and tide forcing in split-explicit high resolution coastal ocean models
PDF.pdf
M. Solanoa,∗ , M. Canalsb,c, S. Leonardid
aNaval Research Laboratory, Stennis Space Center, Mississippi, United States
b UPRM Center for Applied Ocean Science and Engineering, University of Puerto Rico, Mayagüez, Puerto Rico
cCaribbean Coastal Ocean Observing System, Mayagüez, Puerto Rico
d Department of Mechanical Engineering, The University of Texas at Dallas, Texas, United States
Received 1 October 2019; received in revised form 20 December 2019; accepted 21 December 2019
Available online 15 February 2020
Abstract
A persisting problem in ocean modeling around coastal areas has been the adequate imposition of open boundary conditions, especially
in the presence of complex bathymetric features commonly found in shelf seas. Ocean models with horizontal resolutions of O(1km) can
generate submesoscale current variability not captured by coarser regional models which resolve mesoscale flows approximately in geostrophic
and thermal-wind balance. In practice, higher-resolution systems will provide analyses with enhanced spatial detail but may be less skillful
at predicting the evolution of mesoscale eddies. This study aims at assessing the role of open boundary conditions, tidal forcing and tide
filtering techniques in a coastal ocean model using a split-explicit time marching scheme with horizontal resolutions of O(1km) and its effect
in coastal sea level dynamics and ocean currents. Numerical experiments have been performed using the Regional Ocean Modeling System
(ROMS) with different uniform horizontal resolution: 3.6km, 1.2km and 740m. Initial and lateral boundary conditions are derived from
the U.S. Naval Oceanographic Office (NAVOCEANO) operational AmSeas model forecast, a 3.6-km resolution application of the regional
Navy Coastal Ocean Model (NCOM) that encompasses the Gulf of Mexico and Caribbean Sea. Meteorological conditions are interpolated
from the Navy’s Coastal Ocean-Atmosphere Mesoscale Prediction System (COAMPS) model with the exception of surface stresses, which
are computed from a 2-km application of the Weather Research and Forecasting (WRF) model used by National Center for Environmental
Protection’s (NCEP) National Digital Forecast Database (NDFD). Tidal variability is imposed in two ways: 1) by specifying the time evolution
of sea level at the open boundaries and 2) spectrally by specifying the harmonic phases and amplitudes from the TPXO global inverse tide
solutions at the boundaries. Sub-tidal low frequency conditions are imposed by filtering high frequencies out of the regional NCOM solution.
A spectral analysis of the sea surface height and barotropic velocity is performed via Fourier’s transform, showing significant differences
at higher frequencies. Tide signals are then reconstructed and removed from the open boundary condition’s (OBC’s) in 2 ways: 1) using
Rich Pawlowicz’s t_tide package (i.e., classic harmonic analysis) (Pawlowicz, 2002 [1]) and 2) with traditional low-pass filters. The spectral
analysis provides a comprehensive characterization of the response in coastal sea level dynamics, which is difficult to elucidate in the time
domain. Modeled results of sea surface height and ocean currents are validated with NOAA tide gauges and Acoustic Doppler Current
Profilers. The tide filtering approach with the low-pass filter yields significant improvement of water levels in coastal areas, and similar
performance in the modeled currents. The grid sensitivity analysis shows significant reduction of error when increasing the resolution from
3.6km to 1.2km, but no further improvement is found at higher resolution. Power spectra shows that further increasing the resolution yields
better agreement for higher frequencies but no significant reduction in error.
© 2020 Shanghai Jiaotong University. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license. (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Keywords: Ocean modeling; Boundary conditions; Tides; ROMS.