um hexafluorophosphate solut~ons, using freshly dstilled tetrahydrofuran or methyene chloride. A soutons were prepared under an atmosphere of nitrogen and degassed completely before injecton into the SEC cell. Blank reference solut~ons contalnng 0.1 M tetra-n-butyl ammonlum hexafluorophosphate were used for the Fourer transform R solvent subtractons. A Princeton Appl~ed Research (PAR) model 175 Unversal Programmer w~th a PAR model 176 Current Follower were used to effect and monitor thn-ayer bulk eectroyses. The IR spectra were acquired wth a Mattson Research Seres Fourier transform R equipped w~th a MCT (mercurycadmium-telluride) detector. 23 R. E. W t t r g and C P. Kubak, J. Electroanal. Chem. 393. 75 (1995). 24 F -W. Grevels et a / . Angew. Chem int. Ed. Engi. 26.885 (1 987) 25. H. L. Strauss, J Am. Chem. Soc. 11 4, 905 (1 9921, 26 J J. Turner. C:'M. Gordon. S. M. Howde, J. Phys. Chem. 99. 17532 (1 995).

No abstract

Equatorial wave dynamics are essential in most oceanic models for reproducing aspects of the El Niño‐Southern Oscillation (ENSO) phenomenon. In this paper, observational evidence of first baroclinic equatorial Kelvin and first symmetric meridional Rossby waves is found in Geosat‐derived sea level anomalies (SLA) and surface zonal current anomalies (ZCA) in the equatorial Pacific ocean during the 1986–1987 El Niño and the ensuing 1988–1989 La Niña. This was made possible after extensive quality control and specific processing of the recently improved Geosat geophysical data records pertaining to the 17‐day Exact Repeat Mission. In particular, the processing was made so that the Geosat‐derived ZCA best fit near‐surface zonal currents from three equatorial moorings at 165°E, 140°W, and 110°W. The Geosatderived SLA and ZCA are decomposed into first baroclinic equatorial Kelvin and gravest Rossby modes. The emphasis is then put on the chronology of the ZCA at the equator where the currents are the most energetic and where Kelvin and first symmetric Rossby waves explain most of the variance in ZCA, in similar proportion. The 1986–1987 El Niño is mostly characterized by a strong downwelling Kelvin wave in December 1986, a series of downwelling Kelvin waves in March–October 1987, and a strong upwelling Rossby wave in March–September 1987. These waves are consistent with wind forcing, and all give rise to notable eastward ZCA for almost an entire year. During the El Niño‐La Niña transition period there is the occurrence of two downwelling Rossby waves originating from the eastern Pacific at times of favorable wind forcing. These Rossby waves switch the basin‐wide ZCA from eastward to westward, terminating the warm event. Then, the 1988–1989 La Niña shows the generation of a series of upwelling Kelvin waves which are clearly identified in March/April and September/October 1988. These upwelling Kelvin waves are apparently related to the enhancement of the trade wind in the western half of the basin. These waves maintain the basin‐wide westward ZCA for more than a year, although the ZCA is slightly reduced by a front of upwelling Rossby waves in phase with the normal seasonal cycle. The Kelvin wave forcing and sign of Kelvin wave contribution versus Rossby wave contribution to ZCA in the western Pacific are discussed, and we conclude that neither pleads in favor of the delayed action oscillator mechanism.

Abstract. In the equatorial Pacific, zonal displacements of the eastern edge of the warm pool represent an intrinsic manifestation of E1 Nifio Southern Oscillation (ENSO) events, with numerous dynamical and biogeochemical consequences. Following a previous work dedicated to the 1986-1989 Geosat period, we focus on the 1992-1998 zonal displacements of the warm pool using mainly TOPEX/Poseidon data. We also used a simple linear model forced by monthly ERS winds to help in the interpretation of the results. We found that the 1992-1998 zonal displacements of the warm pool resulted mainly from horizontal advection by zonal current anomalies, through a combination of interannual equatorial

Profiling current meter (PCM) measurements under adrifting buoy are compared with concurrent shipboard acoustic Doppler current profiler (ADCP) measurements camed out in the western equatorial Pacific in March 1991, from 10"s to 7ON along the 165"E meridian. The mean (ADCP minus PCM) zk rms differences between zonal and meridional velocity components are 5.7 t 11.2 cm s-' and 0.0 t 8.8 cm s-'. respectively, when PCM measurements are relative to 600 m. The mean f rms differences decrease to 2.3 t 7.8 cm s-' and 0.0 k 6.3 cm s-' when the PCM and ADCP data are both referenced to the same layer (on a mean, 16-240 m). As compared with ADCP, it is found that PCM underestimates velocities of less than 20 cm s-' by about 25%.