Decadal variability in Sahel and northeast Brazil rainfall is correlated with a so-called tropical Atlantic SST dipole, which displays large cross-equatorial sea surface temperature (SST) gradient. While simple/intermediate coupled models suggest this dipole as an intrinsic mode of the coupled ocean-atmospheric system, more sophisticated coupled general circulation models (GCMs) give mixed results.

This disagreement among coupled GCMs may be attributed to the different atmospheric response of their atmospheric component to the SST dipole. In particular, atmospheric GCMs disagree on the magnitude and meridional extent of their trade-wind response off the ITCZ, response characteristics key to ocean-atmospheric feedback. Here we use an atmospheric GCM to investigate how the planetary boundary layer (PBL) physics affect the model adjustment to SST changes.

The GCM displays a response that resembles observed SST-wind relation in the tropical Atlantic and is suggestive of their positive feedback through surface evaporation (Okumura et al. 2001, GRL). The largest wind anomalies are located at 950 hPa and decrease rapidly toward the surface at a rate greater than the surface drag effect. Momentum budget analysis indicates that while the anomalous winds are largely driven by sea level pressure anomalies, SST-induced changes in vertical mixing are key to determining the magnitude of anomalous wind at the sea surface. Over the warm lobe of the SST dipole, weakened stability near the surface enhances vertical mixing, brings faster-moving air from aloft and thereby accelerates the easterly trades at the surface. The resultant easterly anomalies act to reduce the SLP-driven westerly anomalies at the surface. In a sensitivity experiment where the vertical viscosity is held independent of SST anomalies, SLP-driven wind anomalies increase at the surface by a factor of 1.5-2.0.

Our results suggest that the disagreement of atmospheric GCMs in surface wind response may result from the differences in their treatment of the PBL and the vertical shear adjustment. In a given model, the relative importance of the SLP and vertical mixing mechanisms determines the sign and magnitude of surface wind response to SST changes.

The paper will also compare the GCM results with observed wind profiles and discuss what determines the relative importance of these two mechanisms in nature.