The impacts of Amazon deforestation on climate change are investigated through the use of twin numerical experiments with an Atmospheric General Circulation Model (AGCM) with prescribed global sea surface temperature and the same AGCM coupled to an ocean GCM over the global-tropics (CGCM).An ensemble approach is adopted, with ten member ensemble-averages of a control simulation compared with perturbed simulations for three scenarios of Amazon deforestation. The latest 20 years of simulation from each experiment are analyzed. Local surface warming and rainfall reduction are simulated by both models over the Amazon basin, with the coupled model presenting rainfall reduction that is nearly 60% larger compared to its control run than those obtained by the AGCM. The results also indicated that both the fraction of the deforested area and the spatial continuity of vegetated area might be important for modulating global climate variability and change. Additionally, significant remote atmospheric responses to Amazon deforestation scenarios are detectable for the coupled simulations, which revealed global ocean and atmosphere circulation changes conducive to enhanced ocean-atmosphere variability over the Pacific Ocean. This, in turn, is interpreted as a manifestation of enhanced El Niño-Southern Oscillation (ENSO) activity over the Pacific and a positive feedback contributing to the extra rainfall reduction over the Amazon on the coupled simulations.

No abstract

Abstract. The performance of the coupled ocean–atmosphere component of the Brazilian Earth System Model version 2.5 (BESM-OA2.5) was evaluated in simulating the historical period 1850–2005. After a climate model validation procedure in which the main atmospheric and oceanic variabilities were evaluated against observed and reanalysis datasets, the evaluation specifically focused on the mean climate state and the most important large-scale climate variability patterns simulated in the historical run, which was forced by the observed greenhouse gas concentration. The most significant upgrades in the model's components are also briefly presented here. BESM-OA2.5 could reproduce the most important large-scale variabilities, particularly over the Atlantic Ocean (e.g., the North Atlantic Oscillation, the Atlantic Meridional Mode, and the Atlantic Meridional Overturning Circulation), and the extratropical modes that occur in both hemispheres. The model's ability to simulate such large-scale variabilities supports its usefulness for seasonal climate prediction and in climate change studies.

The impact of ocean-atmosphere interactions on summer rainfall over the South Atlantic Ocean is explored through the use of coupled ocean-atmosphere models. The Brazilian Center for Weather Forecast and Climate Studies (CPTEC) coupled ocean-atmosphere general circulation model (CGCM) and its atmospheric general circulation model (AGCM) are used to gauge the role of coupled modes of variability of the climate system over the South Atlantic at seasonal time scales. Twenty-six years of summer [DecemberFebruary (DJF)] simulations were done with the CGCM in ensemble mode and the AGCM forced with both observed sea surface temperature (SST) and SST generated by the CGCM forecasts to investigate the dynamics/thermodynamics of the two major convergence zones in the tropical Atlantic: the intertropical convergence zone (ITCZ) and the South Atlantic convergence zone (SACZ). The results present both numerical model and observational evidence supporting the hypothesis that the ITCZ is a thermally direct, SSTdriven atmospheric circulation, while the SACZ is a thermally indirect atmospheric circulation controlling SST variability underneath-a consequence of ocean-atmosphere interactions not captured by the atmospheric model forced by prescribed ocean temperatures. Six CGCM model results of the Ensemble-based Predictions of Climate Changes and their Impacts (ENSEMBLES) project, NCEP-NCAR reanalysis data, and oceanic and atmospheric data from buoys of the Prediction and Research Moored Array in the Tropical Atlantic (PIRATA) Project over the tropical Atlantic are used to validate CPTEC's coupled and uncoupled model simulations.

This paper shows the Atlantic Equatorial Undercurrent (EUC) revealed by a PIRATA ADCP during 2002 at 0°N, 23°W and simulated by an eddy‐resolving ocean general circulation model (MOM) forced by NCEP/NCAR reanalysis wind stresses. The PIRATA data revealed the shallowing of the EUC between January and May, concurrently with the reversal of the easterly trades to westerly, and the deepening of the EUC from May to December. Empiric Orthogonal Function (EOF) analysis shows two main components; the first mode, explaining 63.7% of the total variance, represents the main seasonal variation of the undercurrent during 2002, while the second mode, explaining 19.9% of the variance, represents higher frequency variability occurring in the EUC core, from 60 to 80 m. All model experiments were able to reproduce this variability, and the best one was chosen objectively using EOF and joint EOF analysis techniques.