Inter-Annual Long Equatorial Waves in the Tropical Atlantic (1981-2000)
Serena Illig(1) , Boris Dewitte(1) , Ayoub Nadia(1) , Yves du Penhoat(1) , Gilles Reverdin(2) , Pierre De Mey(1) , Fabrice Bonjean(3) , and G.S.E. Lagerloef(3)
14 av. E. Belin,
(2) LODYC, 4, place Jussieu, 75252 Paris, France
(3) ESR, 1910 Fairview Ave. E., Suite 210, Seattle, WA 98102, USA, France
The purpose of this study is to determine to which extent linear theory can explain the inter-annual variability observed in the Tropical Atlantic and interpret the TOPEX/Poseidon data in terms of long equatorial waves. Due to the relative scarcity of observed data in the subsurface, we first investigate the vertical structure variability in the Tropical Atlantic based on the CLIPPER project high-resolution Ocean General Circulation Model (OGCM) simulation for the 1981-2000 period.
The results reveal that the 6 first baroclinic modes are necessary to accurately represent the surface zonal current and sea level inter-annual variability. The second baroclinic mode is the most energetic, with a variability peak in the central basin. The first and the third modes contribute with comparable amplitude but with different spatial distribution in the equatorial wave guide. The first mode exhibits a variability peak in the western part of the basin, where the largest variability in zonal wind stress is observed, whereas the energy of the third baroclinic mode is confined in the eastern region, where the thermocline rises. The summed-up contribution of the high-order baroclinic modes variability (4 to 6) is as energetic as the gravest modes and is the largest in the east.
Kelvin and meridional Rossby components are then derived for each of the gravest baroclinic mode contributions by projecting onto the associated meridional structures. The effect of longitudinal boundaries close to the equator is taken into consideration. Equatorial Kelvin and Rossby waves propagations, with phase speed values close to the theoretical ones, are identified for the first three baroclinic modes. Comparisons with a multi-mode linear simulation (OLM) and TOPEX/Poseidon derived sea level and currents indicate that long equatorial wave in the Atlantic explain a large part of the inter-annual variability and that they should be identified from altimetry. However, the peculiarity of the equatorial Atlantic thermocline, with regards to the baroclinic mode energy distribution, renders this exercise difficult without the information provided by subsurface observations or model outputs. Nevertheless, OLM forced simulations allow for the interpretation of the observed altimetric signal, in particular during the boreal spring 1996, when the equatorial Atlantic went through anomalous warm conditions.