Sea Surface Topography and Geostrophic Flow Estimation from Multi-Satellite Altimetry and Shipborne Gravity Data Using Low-pass Filtering and Wavelet Decomposition

• Main Authors: G. S. Vergos, R. S. Grebenitcharsky, M. G. Sideris
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Subjects: low-pass filtering; sea surface; Newfoundland
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.499.5665; http://olimpia.topo.auth.gr/gg2002/session4/vergg2002_sst.pdf

Abstract. The possibility of improving the estimation of the quasi-stationary sea surface topography (QSST) and the determination of geostrophic velocities is investigated in an area close to the island of Newfoundland, bounded by 40o ≤ φ ≤ 50o and 310o ≤ λ ≤ 320o. Multi-satellite (ERS1, GEOSAT) geodetic mission altimetry and shipborne gravity data are used in an attempt to improve the estimation of the QSST and determine the general circulation pattern in the area under investigation. Newly estimated local bathymetry models are implemented in the predictions, aiming at providing as smooth residuals as possible before the QSST estimation takes place. The EGM96 geopotential and DOT models are used to model the low-frequency part of the gravity field signal and correct the altimetry data for the QSST respectively. The estimation of the QSST using altimetry and shipborne gravity anomalies aims to detect remaining signal at wavelengths shorter than about 2000 km, contrary to what global DOT models imply. Additionally, ocean current velocities, based on the theory of geostrophic flow, are derived to detect the general circulation pattern of the areas under study. Blunders in the shipborne data together with high-frequency oceanic phenomena contaminating altimetry sea surface heights can cause the SST estimates to have unreasonably high or small values. To detect and smooth such discrepancies, wavelet decomposition and low-pass filtering are used. The localization property of wavelets is used to detect the distribution of such irregularities, which can be smoothed using a wavelet transform as multiscale differential operator. Smoother SST and geostrophic velocity estimates are derived, showing very good agreement with the circulation pattern of the area.