2004 LOM Workshop Tuesday 1:40 - 2:00 p.m.
Origins and dynamics of the 90-day and 30-60 day variations in the equatorial Indian Ocean
Weiqing Han
University of Colorado
whan@enso.colorado.edu
ABSTRACT
A thorough investigation of the origins and dynamics for the surface and subsurface zonal currents at 20--90 day period in the equatorial Indian Ocean is conducted using an ocean general circulation model (OGCM). To help understand the wind-driven equatorial wave dynamics that occur in the OGCM, a linear continuously stratified ocean model (LM) is also used. Solutions are found in a realistic tropical Indian Ocean basin, and they are forced by NCEP 3-day and monthly forcing fields, together with CMAP pentad and monthly precipitation, for the period of 1988--2001. At both the surface and subsurface, the OGCM solution forced by the 3-day mean fields shows a dominant spectral peak of zonal currents at 90-day period and secondary peaks at 30--60 days, with the 90-day maxima shifting westward as depth increases. These features also appear in the LM solution, suggesting that the wind-driven equatorial wave dynamics play a deterministic role in causing the intraseasonal currents. The dominant 90-day peak is observed by the TOPEX/POSEIDON altimetry data throughout the equatorial Indian Ocean, and is detected by current measurement near the Indonesian seas. The 90-day TOPEX/POSEIDON sea level anomaly and the OGCM surface current evidently show the equatorial Kelvin and the first meridional mode Rossby wave structure, demonstrating that the 90-day oscillation in the equatorial Indian Ocean results from the equatorial waves in responding to the basin scale, 90-day wind forcing. Although zonal currents peak at 90 days, zonal winds peak at 30-60 day period. The skew of frequency between forcing and response results primarily from two causes, consistent with the author's 4.5-layer model study. First, the lower frequency, larger scale Kelvin and Rossby waves at the 90-day period are more efficiently excited by the large scale winds than the higher frequency, shorter wavelength 30--60 day waves. Second, Rossby waves reflected from the eastern ocean boundary significantly enhance the interior response, because the second baroclinic mode resonates with the 90-day wind. The Maldives Islands in the central equatorial basin weaken the resonance somewhat. Although most energy is surface trapped due to mixing, pycnocline reflection, and critical layer absorbtion, there is a fair amount of energy penetrates through the pycnocline down to the deep ocean. A clear upward phase propagation presents in the 90-day zonal flow, indicating that energy is propagating down. Energy of the 90-day Rossby waves excited by the wind maxima near 70E--80E propagates down along the WKB ray path, which appears to explain the westward migration of the 90-day variance maxima. These maxima are significantly enhanced by the reflected Rossby waves at all depths owing to the resonance effect. At 30--60 day period, zonal currents also result largely from intraseasonal wind forcing. At the surface and in the basin interior, the 30--60 day zonal currents are caused by directly forced Rossby and Kelvin waves, and effects of reflected Rossby waves are negligible. As depth increases, however, reflected Rossby waves become more important. This is because the WKB ray paths for the 30--60 day Rossby waves are two to three times steeper than the 90-day ones, energy associated with the reflected Rossby waves penetrate down to the deep ocean efficiently and thus has little effects at the surface. Although intraseasonal wind forcing is important, oceanic instabilities are not negligible near the western boundary and in the central basin south of Sri Lanka near 5N.
LOM Users' Workshop, February 9-11, 2004