Reconstruction of the 3-D Dynamics From Surface Variables in a High-Resolution Simulation of North Atlantic

Type Article
Date 2018-03
Language English
Author(s) Fresnay S.1, Ponte AurelienORCID2, Le Gentil Sylvie2, Le Sommer J.3
Affiliation(s) 1 : OceanDataLab, Brest, France.
2 : Univ Brest, CNRS, IFREMER, Lab Oceanog Phys & Spatiale,IUEM,IRD, Brest, France.
3 : Univ Grenoble Alpes, CNRS, IRD, IGE, Grenoble, France.
Source Journal Of Geophysical Research-oceans (2169-9275) (Amer Geophysical Union), 2018-03 , Vol. 123 , N. 3 , P. 1612-1630
DOI 10.1002/2017JC013400
WOS© Times Cited 18
Keyword(s) ocean dynamics, quasi-geostrophy, potential vorticity, sea surface height, correlation analysis, spectral analysis

Several methods that reconstruct the three-dimensional ocean dynamics from sea level are presented and evaluated in the Gulf Stream region with a 1/60° realistic numerical simulation. The use of sea level is motivated by its better correlation with interior pressure or quasigeostrophic potential vorticity (PV) compared to sea surface temperature and sea surface salinity, and, by its observability via satellite altimetry. The simplest method of reconstruction relies on a linear estimation of pressure at depth from sea level. Another method consists in linearly estimating PV from sea level first and then performing a PV inversion. The last method considered, labeled SQG for surface quasigeostrophy, relies on a PV inversion but assumes no PV anomalies. The first two methods show comparable skill at levels above -800∼m. They moderately outperform SQG which emphasizes the difficulty of estimating interior PV from surface variables. Over the 250-1000∼m depth range, the three methods skillfully reconstruct pressure at wavelengths between 500 and 200 km whereas they exhibit a rapid loss of skill between 200 and 100 km wavelengths. Applicability to a real case scenario and leads for improvements are discussed.

Full Text
File Pages Size Access
Publisher's official version 24 11 MB Open access
Supplementary Information SO1 17 MB Open access
Supplementary Information FigSO1 5 7 MB Open access
Top of the page