Ingo Richter

Equatorial Atlantic variability and its relation to mean state biases in CMIP5

Coupled general circulation model (GCM) simulations participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5) are analyzed with respect to their performance in the equatorial Atlantic. In terms of the mean state, 29 out of 33 models examined continue to suffer from serious biases including an annual mean zonal equatorial SST gradient whose sign is opposite to observations. Westerly surface wind biases in boreal spring play an important role in the reversed SST gradient by deepening the thermocline in the eastern equatorial Atlantic and thus reducing upwelling efficiency and SST cooling in the following months. Both magnitude and seasonal evolution of the biases are very similar to what was found previously for CMIP3 models, indicating that improvements have only been modest. The weaker than observed equatorial easterlies are also simulated by atmospheric GCMs forced with observed SST. They are related to both continental convection and the latitudinal position of the intertropical convergence zone (ITCZ). Particularly the latter has a strong influence on equatorial zonal winds in both the seasonal cycle and interannual variability. The dependence of equatorial easterlies on ITCZ latitude shows a marked asymmetry. From the equator to 15°N, the equatorial easterlies intensify approximately linearly with ITCZ latitude. When the ITCZ is south of the equator, on the other hand, the equatorial easterlies are uniformly weak. Despite serious mean state biases, several models are able to capture some aspects of the equatorial mode of interannual SST variability, including amplitude, pattern, phase locking to boreal summer, and duration of events. The latitudinal position of the boreal spring ITCZ, through its influence on equatorial surface winds, appears to play an important role in initiating warm events.

Multiple causes of interannual sea surface temperature variability in the equatorial Atlantic Ocean

The tropical Atlantic Ocean shows sea surface temperature variability on interannual timescales. Observational and model data suggest that some of this variability can be attributed to the advection of anomalously warm northern subtropical waters toward the Equator. The eastern equatorial Atlantic Ocean is subject to interannual fluctuations of sea surface temperatures, with climatic impacts on the surrounding continents1,2,3. The dynamic mechanism underlying Atlantic temperature variability is thought to be similar to that of the El Niño/Southern Oscillation (ENSO) in the equatorial Pacific4,5, where air–sea coupling leads to a positive feedback between surface winds in the western basin, sea surface temperature in the eastern basin, and equatorial oceanic heat content. Here we use a suite of observational data, climate reanalysis products, and general circulation model simulations to reassess the factors driving the interannual variability. We show that some of the warm events can not be explained by previously identified equatorial wind stress forcing and ENSO-like dynamics. Instead, these events are driven by a mechanism in which surface wind forcing just north of the equator induces warm ocean temperature anomalies that are subsequently advected toward the equator. We find the surface wind patterns are associated with long-lived subtropical sea surface temperature anomalies and suggest they therefore reflect a link between equatorial and subtropical Atlantic variability.

What controls equatorial Atlantic winds in boreal spring

The factors controlling equatorial Atlantic winds in boreal spring are examined using both observations and general circulation model (GCM) simulations from the coupled model intercomparison phase 5. The results show that the prevailing surface easterlies flow against the attendant pressure gradient and must therefore be maintained by other terms in the momentum budget. An important contribution comes from meridional advection of zonal momentum but the dominant contribution is the vertical transport of zonal momentum from the free troposphere to the surface. This implies that surface winds are strongly influenced by conditions in the free troposphere, chiefly pressure gradients and, to a lesser extent, meridional advection. Both factors are linked to the patterns of deep convection. Applying these findings to GCM errors indicates, that, consistent with the results of previous studies, the persistent westerly surface wind bias found in most GCMs is due mostly to precipitation errors, in particular excessive precipitation south of the equator over the ocean and deficient precipitation over equatorial South America. Free tropospheric influences also dominate the interannual variability of surface winds in boreal spring. GCM experiments with prescribed climatological sea-surface temperatures (SSTs) indicate that the free tropospheric influences are mostly associated with internal atmospheric variability. Since the surface wind anomalies in boreal spring are crucial to the development of warm SST events (Atlantic Niños), the results imply that interannual variability in the region may rely far less on coupled air–sea feedbacks than is the case in the tropical Pacific.

Challenges and Prospects for Reducing Coupled Climate Model SST Biases in the Eastern Tropical Atlantic and Pacific Oceans: The U.S. CLIVAR Eastern Tropical Oceans Synthesis Working Group

Well-known problems trouble coupled general circulation models of the eastern Atlantic and Pacific Ocean basins. Model climates are significantly more symmetric about the equator than is observed. Model sea surface temperatures are biased warm south and southeast of the equator, and the atmosphere is too rainy within a band south of the equator. Near-coastal eastern equatorial SSTs are too warm, producing a zonal SST gradient in the Atlantic opposite in sign to that observed. The U.S. Climate Variability and Predictability Program (CLIVAR) Eastern Tropical Ocean Synthesis Working Group (WG) has pursued an updated assessment of coupled model SST biases, focusing on the surface energy balance components, on regional error sources from clouds, deep convection, winds, and ocean eddies; on the sensitivity to model resolution; and on remote impacts. Motivated by the assessment, the WG makes the following recommendations: 1) encourage identification of the specific parameterizations contributing to the biases in individual models, as these can be model dependent; 2) restrict multimodel intercomparisons to specific processes; 3) encourage development of high-resolution coupled models with a concurrent emphasis on parameterization development of finer-scale ocean and atmosphere features, including low clouds; 4) encourage further availability of all surface flux components from buoys, for longer continuous time periods, in persistently cloudy regions; and 5) focus on the eastern basin coastal oceanic upwelling regions, where further opportunities for observational–modeling synergism exist.

Phase locking of equatorial Atlantic variability through the seasonal migration of the ITCZ

The equatorial Atlantic is marked by significant interannual variability in sea-surface temperature (SST) that is phase-locked to late boreal spring and early summer. The role of the atmosphere in this phase locking is examined using observations, reanalysis data, and model output. The results show that equatorial zonal surface wind anomalies, which are a main driver of warm and cold events, typically start decreasing in June, despite SST and sea-level pressure gradient anomalies being at their peak during this month. This behavior is explained by the seasonal northward migration of the intertropical convergence zone (ITCZ) in early summer. The north-equatorial position of the Atlantic ITCZ contributes to the decay of wind anomalies in three ways: (1) horizontal advection associated with the cross-equatorial winds transports air masses of comparatively low zonal momentum anomalies from the southeast toward the equator. (2) The absence of deep convection leads to changes in vertical momentum transport that reduce the equatorial wind anomalies at the surface, while anomalies aloft remain relatively strong. (3) The cross-equatorial flow is associated with increased total wind speed, which increases surface drag and deposit of momentum into the ocean. Previous studies have shown that convection enhances the surface wind response to SST anomalies. The present study indicates that convection also amplifies the surface zonal wind response to sea-level pressure gradients in the western equatorial Atlantic, where SST anomalies are small. This introduces a new element into coupled air-sea interaction of the tropical Atlantic.