FRONTIER Newsletter No.9 Jan.2000

Inter-Hemisphere Decadal Variations in SST, surface wind, heat flux, low-level cloud over the Atlantic basin

Youichi Tanimoto (Climate Variations Research Program), Shangping Xie (IPRC)
1. Introduction
In the tropical Atlantic, we can find a strong air- sea coupled variability as we also can realize that the air- sea coupled ENSO dominates in the tropical Pacific. The configuration of the Atlantic coupled variability, however, is very much different from that in the Pacific.An equatorial symmetric cold tongue expands from the eastern boundary into the central basin in the Pacific.Whereas, in the Atlantic, an equatorial anti- symmetric pattern is seen in the SST and SLP fields. As many studies showed before, the tropical dipole mode is associated with a migration of the Atlantic ITCZ.
However, the dipole mode is not the only mode of SST variability in the tropical Atlantic.A scatter plot of zonal mean SST and SLP anomalies in 10 - 20 degree latitudinal bands presents an apparent inter-hemispheric decorrelation in both SST and SLP fields. In the present study, we further examine this inter- hemispheric relation in the SST variability from different perspective.

2. Data
For the present study, we construct 4 degree longitude by 4 degree latitude monthly averaged marine meteorological data sets from quality controlled ship and buoy observations,compiled in Long Marine Reports of COADS.We will examine SST,SLP,vector and scalar surface wind,sensible and latent heat fluxes.A higher resolution data set of cloudiness is also included here.We also complement the COADS with the NCEP reanalysis because COADS observations in the South Atlantic may suffer large sampling errors due to few ship observations.For example,the NCEP provides dynamically- consistent SLP and wind velocity data.In contrast,grid point values of COADS are calculated from independent observations of SLP and wind velocity.We would like to take both advantages into our study.

3. Results
Several spectral analyses showed that large variance appears in a decadal frequency domain in the tropical Atlantic. The cross-equatorial gradient index of the tropical Atlantic varied on decadal time scales. Before we perform an EOF analysis, SST anomalies are averaged for boreal winter, December through March and then filtered through the 8 - 16 year decadal band. These leading EOFs are derived from statistically independent fields so that there is no a priori reason to be highly correlated. Figure 1 shows a regression map of the leading EOF for the North Atlantic and the South Atlantic.This figure also displays the Pan- Atlantic pattern with centers of action lining up meridionally from south of Greenland all the way to the South Atlantic. We perform the same analysis to the boreal winter SLP anomalies in the two subdomains. The four accompanying PCs of the SST and SLP fields correlate to each other. This means that these spatially coherent patterns actually fluctuate almost in phase on the decadal time scales. We made composite maps of several marine meteorological variables to examine features common to a distinct Pan- Atlantic decadal oscillation. Surface forcing by the latent heat flux to the ocean induces the warming of SST.This means that the wind-evaporation- SST feedback works in this air-sea coupled decadal variability. In the atmospheric dipole mode, the amplitude in either side of equator is not comparable although the observed SST pattern indicates a rough symmetry in amplitude. This indicates another physical process involving an SST symmetry.

Figure 1 : The first SST EOFs for the North and the South Atlantic of decadal band-passed (8-16 years) boreal winter SST anomalies. Regressed SST anomalies onto the PCs for the North (South) Atlantic are shown in the upper (lower) panel. Contour interval is 0.1 .

4. Discussion
In the near- equatorial region, Figure 2 shows the cloudiness and convergence increases over the warmed SST, and a decrease of cloudiness and convergence south of the equator, which is associated with a migration of the equatorial high cloud, the Atlantic ITCZ.These near-equatorial cloudiness change act as negative feedback to diminish the cross- equatorial gradient. We can find additional changes in off- equatorial region that have not been noted to our knowledge. These cloudiness anomalies correspond to changes in low-level stratiform clouds because cloudiness anomalies correlate with local SST anomalies, but not associated with significant changes in surface convergence / divergence.
Negative SST anomalies increase the stratification across the top of the boundary layer if temperatures in the free atmosphere do not change. Enhanced capping of the boundary layer leads to an increase in stratiform cloud trapped over the top of the boundary layer. Increased cloud will in turn cause a further cooling in SST.These processes complete a positive feedback loop between local SST and startus clouds.

Figure 2 : (a) Boreal spring (Feb.-Apr.) composite of cloud cover anomalies associated with the Pan-Atlantic decadal oscilation (left panel; blue color < -3.0% & orange color > 3.0%).
(b) Zonal mean anomalies of SST (upper right) and cloud cover (%; orange line) along with surface wind divergence (10 -6 s -1 in blue line; lower right panel).