Low-Frequency Reflection from a Nonmeridional Eastern Ocean Boundary and the Use of Coastal Sea Level to Monitor Eastern Pacific Equatorial Kelvin Waves

Abstract

Analytical theory and wind, sea level, and atmospheric pressure data were used to examine low-frequency dynamics near the equatorial eastern Pacific Ocean boundary. The analytical technique involves linking low-frequency solutions near a nonmeridional boundary with the deep ocean equatorial Kelvin wave for both unforced and wind-forced flows using an equatorial wave orthogonality condition. The following results were obtained. Mathematical and physical arguments show that nonmeridional boundaries should be less reflective than meridional ones, and that the poleward coastal Kelvin wave energy flux should be greater the more the boundary tilts from north to south. The eastern Pacific Ocean boundary is more nonmeridional in the Northern Hemisphere and, as a consequence, calculations for the intraseasonal, semiannual, annual, and interannual frequencies an indicated that poleward coastally trapped energy flux is greater in the Northern Hemisphere than in the Southern Hemisphere. For most frequencies the asymmetry is small, however. Although the Pacific eastern boundary is far from being meridional, only for the higher intraseasonal frequencies was it substantially less reflective than a meridional boundary. The eastern ocean boundary is a special place where it is comparatively easy to determine the equatorial Kelvin wave signal. The sea level at the nonmeridional boundary is partly due to incoming equatorial Kelvin waves and partly due to an alongshore tilt associated with the local wind forcing. A simple formula enables the extraction of the equatorial Kelvin wave sea level signal from the boundary sea level. When there is no forcing, this formula indicates that the boundary sea level is 21/2 times the equatorial Kelvin wave signal at the intersection of the equator and an appropriate average boundary meridian. In agreement with previous analysis, the theory indicates that the intraseasonal sea level signal in the eastern tropical Pacific is due to remotely forced equatorial Kelvin waves, while the annual sea level signal is due to the local alongshore winds. In contrast with earlier work, remote and local semiannual sea level contributions are comparable. The interannual signal is primarily due to remotely forced equatorial Kelvin waves, although on average during El Niño the sea level at Balboa (9°N) is reduced by an anomalous southward wind associated, it seems, with an anomalous southward displacement of the intertropical convergence zone (ITCZ) near the eastern Pacific boundary. Using the interannual sea levels at the eastern boundary it is possible to estimate the interannual equatorial Kelvin wave signal at the equator and eastern boundary since October 1908. This interannual equatorial Kelvin wave signal can be regarded as an El Niño index. According to this index the Largest El Niños since 1908 occurred in 1982?83 with other major El Niños in 1957 and 1941. Numerically, one obtains the interannual equatorial Kelvin wave index by multiplying the interannual La Libertad signal by 0.57 or the interannual Balboa signal by 0.85.