Relationship of TOPEX/Poseidon altimetric height to steric height and circulation in the North Pacific (original) (raw)

1998, Journal of Geophysical Research

TOPEX/Poseidon altimetric height is compared with 20 transpacific eddy-resolving realizations of steric height. The latter are calculated from temperature (expendable bathythermograph (XBT)) and salinity (expendable conductivity and temperature profiler (XCTD)) profiles along a precisely repeating ship track over a period of 5 years. The overall difference between steric height and altimetric height is 5.2 cm RMS. On long wavelengths (•, <500 km), the 3.5 cm RMS difference is due mainly to altimetric measurement errors but also has a component from steric variability deeper than the 800 m limit of the XBT. The data sets are very coherent in the long-wavelength band, with coherence amplitude of 0.89. This band contains 65% of the total variance in steric height. On short wavelengths (•, >500 km), containing 17% of the steric height variance, the 3.0 cm RMS difference and lowered coherence are due to the sparse distribution of altimeter ground tracks along the XBT section. The 2.4 cm RMS difference in the basin-wide spatial mean appears to be due to fluctuations in bottom pressure. Differences between steric height and altimetric height increase near the western boundary, but data variance increases even more, and so the signal-to-noise ratio is highest in the western quarter of the transect. Basinwide integrals of surface geostrophic transport from steric height and altimetric height are in reasonable agreement. The 1.9 x 104 m 2 s 'i RMS difference is mainly because the interpolated altimetric height lacks spatial resolution across the narrow western boundary current. A linear regression is used to demonstrate the estimation of subsurface temperature from altimetric data. Errors diminish from 0.8øC at 200 m to 0.3øC at 400 m. Geostrophic volume transport, 0-800 m, shows agreement that is similar to surface transport, with 4.8 Sverdrup (Sv) (106m 3 s -1) RMS difference. The combination of altimetric height with subsurface temperature and salinity proffiing is a powerful tool for observing variability in circulation and wansport of the upper ocean. The continuing need for appropriate subsurface data for verification and for statistical estimation is emphasized. This includes salinity measurements, which significantly reduce errors in specific volrune and steric height. 1. Introduction Changes in the height of the sea surface may be caused either by changes in the mass of water at a given location or by changes in the density at constant mass. The latter, known as steric height change, is thought to dominate sea surface variability at low frequency [e.g. Patullo et al., 1955]. A column of warm water stands higher than a column of cold water of equal mass. In the present work, sea surface height measured by the TOPEX/Poseidon (T/P) satellite altimeter is compared with steric height from a data set consisting of repeating temperature (expendable bathythermograph (XBT)) and temperature/salinity (expendable conductivity and temperature profiler (XCTD)) profiles. The XBT/XCFD data set consists of 20 quarterly eddy-resolving XBT transects, together with sparse XCTDs, along a 12,650 km track during Copyright 1998 by the American Geophysical Union. Paper number 98JC01680. 0148-0227/98/98 JC-01680509.00 the T/P era. The comparison, carried out over spatial scales ranging from tens of kilometers to the basin width, will serve to determine (1) the degree to which sea level height i s, indeed, controlled by upper ocean steric change, (2) an upper bound on errors in the T/P measurements as a function of spatial scale, (3)the limits imposed by across-track spatial resolution in the T/P data (i.e., interpolation errors), and (4) the errors in steric height when salinity information is missing or withheld. In places where the difference between steric height and altimetric height exceeds measurement and interpolation errors, deep-ocean steric change or redistribution of mass must have signlficant effects on sea surface height. The high correlations linking altimetric height to steric height and steric height to subsurface temperature variability permit an inversion of altimetric height in order to estimate temperature as a function of position, depth, and time (T( ). A substantial fraction of subsurface temperature variability can be recovered in this way. The relationship between temperature (or density) variability and altimetric height can be exploited to reduce the sampling deficiencies of 27,947