From theoretical considerations, it has long been known that the transient adjustment of ocean circulation to changes in wind and thermal forcing at the sea surface is carried out by Rossby waves. The standard theory for freely propagating, linear Rossby waves yields an eigenvalue problem for which there exists an infinite number of normal mode solutions, ordered by decreasing phase speed. The first-baroclinic mode that is of interest here propagates westward with speeds of the order of 10 cm/s or less, requiring months to decades to cross an ocean basin Figure 1. The ~5 cm surface height signatures of these Rossby waves are mirrored as thermocline displacements of the opposite sign with amplitudes of ~50 m. These large variations of upper ocean thermal content have important implications about the role of the ocean in seasonal to interannual climate variations.
The characteristics of low-frequency, large-scale Rossby waves have been investigated from 3 years of TOPEX/POSEIDON data. The results, reported in detail by Chelton and Schlax (1996), are summarized here. Example time-longitude sections of specifically filtered sea level are shown in Figure 2 for the midlatitude North Pacific and in Figure 3 for the tropical Pacific. A prominent feature is the amplification of midlatitude Rossby waves west of major bottom topography such as the Hawaiian Ridge and the Emporer Seamounts. Other features in these figures are discussed in detail by Chelton and Schlax (1996).
The spatial structure of the observed Rossby waves is evident from the global maps of the filtered sea level fields in Figure 4. These example maps are two frames from a three-year animation (478 KB) of the filtered TOPEX/POSEIDON data. A westward-propagating Rossby wave trough centered on the equator and extending to midlatitudes in both hemispheres can be seen in the Pacific Ocean in the April 13, 1993 frame. The refracted shape that is characteristic of Rossby waves is due to the latitudinal variation of phase speed. In the July 31, 1993 frame, this Rossby wave trough has impinged on the western boundary of the Pacific and an equatorial Kelvin wave trough centered at about 140 W has propagated rapidly eastward more than half way across the Pacific, splitting a newly formed Rossby wave crest that has propagated westward from South America. The time evolution of these and other Rossby wave features are evident in all three ocean basins in the animation (478 KB).
The latitudinal variation of the westward phase speeds of the observed Rossby waves estimated from the 3 years of TOPEX/POSEIDON data is shown in Figure 5; solid circles correspond to Pacific estimates and open circles correspond to Atlantic and Indian Ocean estimates. A detailed comparison with the phase speeds predicted from the standard theory for free, linear Rossby waves (solid line in Figure 5) reveals that the theoretical phase speeds are systematically slower than the observations. The observed propagation speeds are more than a factor of 2 greater than the theoretical phase speeds at latitudes higher than 35 deg (see lower panel of Figure 5). The standard theory is thus evidently deficient in predicting the observed propagation speeds. The implication of these results is that the ocean reacts more rapidly than is generally believed; the transient baroclinic adjustment time of the ocean at 35 deg latitude, for example, is only about half as long as that predicted by the standard theory. The transoceanic transit times shown in Figure 1 for the North Pacific must be modified to account for the observed propagation speeds of extra-tropical baroclinic Rossby waves.
The TOPEX/POSEIDON observations of high phase speeds, as well as the apparent effects of bottom topography on midlatitude Rossby waves noted from Figure 2, provide important consistency checks for evaluating the performance of ocean general circulation models for studies of the role of the ocean in interannual and decadal climate variability.
ReferenceChelton, D. B., and M. G. Schlax, 1996: Global observations of oceanic Rossby waves. Science, vol. 272, pp. 234-238.