A41F-01 INVITED
An 800-Year Tropical Atlantic Sea Surface Temperature Variability Record From the Cariaco Basin, Venezuela
Here we present an eight century tropical Atlantic SST record based on foraminiferal Mg/Ca recovered from Cariaco Basin sediments that have been calibrated to historical instrumental SSTs. Spatial correlations indicate that the proxy record is representative of SSTs over much of the Caribbean and tropical Atlantic. The Mg/Ca-SST record also correlates well with global land and sea surface temperature anomalies, and captures decadal-scale variations in Atlantic tropical storm and hurricane frequency over the late-19th and 20th centuries. The long-term record displays a surprising amount of variability for a tropical location under essentially modern boundary conditions. The tropical North Atlantic does not appear to have experienced a pronounced Medieval Warm Period relative to the complete record. However, strong Little Ice Age cooling of as much as 3 °C occurred between A. D. 1525 and 1625. Spring SSTs gradually rose between A. D. 1650 and 1900 followed by a 2.5 °C warming over the twentieth century. Viewed in the context of the complete record, twentieth century temperatures are not the warmest in the entire record on average, but they do show the largest increase in magnitude and fastest rate of SST change over the last eight hundred years. Spectral analysis of the Mg/Ca-SST data suggests that 2-5 and ~13 year SST variability that is characteristic of tropical Atlantic instrumental records may change through time.
A41F-02
Multi-centennial Climate Variability in the Tropical Andes--a Product of Tropical Atlantic Variability
In the southern tropical Andes, a high-resolution (sub-decadal) record of Holocene precipitation amount has been reconstructed from stable oxygen isotopic ratios determined on carbonate-bearing sediments from Lagos Umayo and Junin (Peru). On orbital timescales, precipitation amount in this region is well known to be influenced by global-scale forcing, not surprising since wet-season insolation varies on precessional timescales by about one-third of the range of the modern annual cycle. On shorter timescales, the reconstructed precipitation amount underwent significant wet/dry alternations typically lasting a few hundred years. The typical amplitude of these alternations is 30% of the annual precipitation amount. Based on the precipitation dependency of modern agricultural yields on the Altiplano, it seems likely that these long-term droughts would have seriously stressed early inhabitants of the region. In the instrumental period, regional precipitation variability on interannual timescales is clearly influenced by Pacific SST variability; e.g. El Nino events produce dry and warm conditions in much of tropical South America including the central Andes. The existence of a Pacific-SST control on Andean lake levels has also been inferred on longer timescales (Bradley et al. 2003). However, as we have previously posited, the reconstructed wet/dry alternations subjectively appear to be in phase with the Holocene Bond cycles, periods of cold SST in the high-latitude North Atlantic and in the tropical north Atlantic. The mechanism by which the tropical Atlantic ocean forces climate variability on the adjacent continent has still not been fully elucidated.
A41F-03
Teleconnection mechanisms to the tropical Atlantic sector from climate changes following an abrupt freshening of the North Atlantic
Distinct ocean dynamical and atmospheric teleconnection pathways have been previously proposed to explain the pronounced response of the tropical North Atlantic climate - in particular, a southward shift of the Intertropical Convergence Zone (ITCZ) - to high latitude North Atlantic climate changes resulting from forced slowdowns of the Atlantic Meridional Overturning Circulation (AMOC). The climate adjustment to an abrupt freshening of the high Northern Atlantic in an ensemble of Community Climate System Model 3 simulations shows both ocean and atmospheric mechanisms at work to cool the midlatitude North Atlantic sea surface temperatures. In the north tropical Atlantic (NTA) however, atmospheric-mediated teleconnection associated with wind-evaporation-SST feedback dominates the cooling, whereas subsurface ocean dynamical response appears to warm the NTA, countering the atmospheric cooling. Thus, atmospheric mechanisms appear to be responsible for the southward ITCZ displacement in AMOC-slowdown scenarios, though our analysis also suggest that the magnitude of this displacement may depend on the relative strengths of the atmospheric and oceanic responses in the NTA. The mechanisms influencing NTA SST in the AMOC slowdown appear distinctly different from the ENSO/NAO-driven trade wind paradigm for interannual SST variations, and may suggest new avenues for understanding tropical Atlantic variability in particular on longer timescales.
A41F-04
Mechanistic studies of the role of Wind-Evaporation-SST feedback in ocean-atmosphere interaction
The Wind-Evaporation-SST (WES) feedback is believed to play an important role in thermodynamic ocean-atmosphere interaction. In the tropical Atlantic, the positive feedback associated with the WES feedback can lead to amplification of the interhemispheric gradient mode variability. It has also been argued that this feedback can communicate high-latitude Atlantic cooling to the tropical Atlantic. In the tropical Pacific, the WES feedback can help explain the propagation of westward propagation of SST anomalies purely through thermodynamic air-sea interaction. However, it is often difficult to conclusively identify the role of the WES feedback using observational analysis, or even by analyzing comprehensive coupled ocean-atmosphere general circulation models (GCMs), because the WES feedback may be masked by other processes. In this study, we report on studies using a modified version of an atmospheric GCM (AGCM), where the WES feedback is deliberately suppressed in the bulk aerodynamic formulation for surface fluxes. By comparing coupled integrations using the modified AGCM to those carried out using the control AGCM, we can focus on the role of the WES feedback. Preliminary results for tropical Atlantic region show that the WES feedback does lead to increased variability, and also helps explain why the ITCZ tends to stay north of the equator.
A41F-05 INVITED
Dust impacts on precipitation in Sahel and North Atlantic
The role of direct radiative forcing of desert dust aerosol in the change from wet to dry climate observed in the African Sahel region in the last half of the 20th century is investigated using simulations with an atmospheric general circulation model. The model simulations are conducted either forced by the observed sea surface temperature (SST) or coupled with the interactive SST using the Slab Ocean Model (SOM). Our model uses dust that is less absorbing in the shortwave band and larger particle sizes than other simulation studies. As a result, our simulations show less shortwave absorption within the atmosphere and larger longwave radiative forcing by dust. Simulations using SOM show reduced precipitation over the intertropical convergence zone (ITCZ) including the Sahel region and increased precipitation south of the ITCZ when dust radiative forcing is included. In SST- forced simulations, on the other hand, significant precipitation changes are restricted to over North Africa. These changes are considered to be due to the cooling of global tropical oceans as well as the cooling of the troposphere over North Africa in response to dust radiative forcing. The model simulation of dust cannot reproduce the observed change in desert dust when allowing dust to respond to changes in simulated climate, even including changes in vegetation, similar to previous studies. If the model is forced to capture observed changes in desert dust, the direct radiative forcing by the increase of North African dust can explain up to 30% of the observed precipitation reduction in the Sahel between wet and dry periods. A large part of this effect comes through atmospheric forcing of dust, and dust forcing on the Atlantic Ocean SST appears to have a smaller impact. The changes in the North and South Atlantic SSTs may account for up to 50% of the Sahel precipitation reduction, but this effect could be due to factors other than desert dust aerosols (e.g., biomass burning aerosols and other low-frequency processes). Vegetation loss in the Sahel region may explain about 10% of the observed drying, but this effect is statistically insignificant due to small number of years in the simulation. Greenhouse gas warming seems to have an impact to increase Sahel precipitation that is opposite to the observed change. Although the estimated values of impacts are likely to be model dependent, our analyses suggest the importance of direct radiative forcing of dust and feedbacks in modulating Sahel precipitation.
A41F-06
Barrier layers and tropical Atlantic SST biases in coupled GCM's
In the ocean the vertical stratification in salinity is sometimes much stronger than in temperature. In that case a barrier layer develops, which is defined as the layer in between the base of the isothermal surface layer and the base of the mixed layer. This layer forms a barrier to entrainment of cold subsurface water into the mixed layer. The dynamics of barrier layers is complex and consequently difficult to model. Biases in rainfall/evaporation, river discharge, ocean currents, etc., affect the salinity stratification, and hence the formation of barrier layers in models. This may contribute to biases in SST. We have explored the contribution from biases in barrier layers to tropical Atlantic SST biases in coupled GCM's. In reality, the strong barrier and the associated temperature inversion below the mixed layer tend to form in the western tropical Atlantic during boreal fall and winter, contributing to the maintenance of warm sea surface temperatures. Many coupled models underestimate this contribution, which may be an important factor in the cold SST bias in the region. In contrast, all models investigated show an erroneous barrier layer in the eastern equatorial Atlantic in boreal summer, where a strong warm SST bias exists. The warm SST bias prevents the ITCZ from migrating towards the north, causing too much rainfall and surface freshening over the eastern equatorial Atlantic, which in turn tends to maintain the erroneous presence of barrier layers. An important implication of our study is that model SST biases in the tropical Atlantic can be significantly improved by reducing the biases in salinity stratification.
A41F-07
Precipitation from African Easterly Waves in a Coupled Model of the Tropical Atlantic Ocean
A regional coupled climate model is configured for the tropical Atlantic to explore the role of synoptic-scale (2-6 day) African Easterly Waves (AEWs) on the simulation of mean precipitation in the marine Inter-Tropical Convergence Zone (ITCZ). Sensitivity tests with varying atmospheric resolution in the coupled model show that these easterly waves are well represented on both fine and coarse grids of the atmospheric model, with the mean variance of the waves on finer grid being roughly 20% larger. The resultant wind shear associated with the strong phase of the AEWs for both atmospheric grids is comparable. Significant differences in the model simulations are found in the precipitation fields, where extreme rainfall events occur in the strong shear of the easterly waves only on the higher resolution grid. This is indicative of the strong coupling between the easterly waves and rainfall. This is because the low-level convergence due to the waves is much larger and more realistic in the fine-resolution simulation, which enables strong precipitation events. The variability in rainfall on these time scales accounts for a significant fraction of the total variability. As a result, the simulation of mean rainfall in the ITCZ and its seasonal migration become more realistic in the higher-resolution case. This suggests that capturing these transient waves and the resultant low-level convergence is a key to improving the simulation of precipitation in global coupled climate models, with obvious implications for regional climate prediction.