Applications of EOF Analysis to the Spatial Estimation of Circulation Features in the Ocean Sampled by High-Resolution CTD Soundings
Open Access
- 1 August 1998
- journal article
- research article
- Published by American Meteorological Society in Journal of Atmospheric and Oceanic Technology
- Vol. 15 (4) , 959-978
- https://doi.org/10.1175/1520-0426(1998)015<0959:aoeatt>2.0.co;2
Abstract
The application of empirical orthogonal function analysis to the estimation of geostrophic circulation features in the ocean is studied with particular reference to spatial series of near-synoptic profiles of CTD data retrieved during intensive surveys of the Alboran Sea (western Mediterranean) and the Bransfield Strait (Antarctica). In both regions, horizontal interpolation of principal components of the observed specific volume profiles provides an efficient method of generating a fully three-dimensional representation of the dynamic height field and geostrophic flow structure, which retains the full resolution of the input profile data. Principal components of profiles sampled by “shallow” soundings are estimated using a simple least squares regression method, leading to a more accurate estimate of dynamic height than one based on profile substitution. Spatial correlation and cross-validation analysis yield consistent results in identifying those vertical modes that account for most of the res... Abstract The application of empirical orthogonal function analysis to the estimation of geostrophic circulation features in the ocean is studied with particular reference to spatial series of near-synoptic profiles of CTD data retrieved during intensive surveys of the Alboran Sea (western Mediterranean) and the Bransfield Strait (Antarctica). In both regions, horizontal interpolation of principal components of the observed specific volume profiles provides an efficient method of generating a fully three-dimensional representation of the dynamic height field and geostrophic flow structure, which retains the full resolution of the input profile data. Principal components of profiles sampled by “shallow” soundings are estimated using a simple least squares regression method, leading to a more accurate estimate of dynamic height than one based on profile substitution. Spatial correlation and cross-validation analysis yield consistent results in identifying those vertical modes that account for most of the res...This publication has 1 reference indexed in Scilit:
- Estimating Subpycnocline Density Fluctuations in the California Current Region from Upper Ocean ObservationsJournal of Atmospheric and Oceanic Technology, 1995