Abstract/Summary

This study aims to evaluate the ability of the new phase Coupled Model Intercomparison Project (CMIP6) models to estimate the characteristics of Easterly Waves Disturbances (EWD) over the Tropical South Atlantic (TSA) that reach Northeast Brazil (NEB). Initially, the precipitation variable output of CMIP, AMIP, hist-1950 and highresSST-present models skill was evaluated using the Global Precipitation Climatology Center (GPCC) dataset to define the best models that reproduce the spatial and temporal precipitation patterns over the study regions. In total, 17 historical CMIP, 16 AMIP, 7 do hist-1950 and 10 do highresSST-present models were assessed. The ensemble's spatial analysis showed the ability of the models in reproducing annual and seasonal climatological precipitation patterns but with CMIP exhibiting the highest variability. Models underestimated the precipitation intensity on NEB's coast and overestimated on TSA and NEB's north. Larger (smaller) uncertainties among models were observed at higher (lower) latitudes. Regarding the annual precipitation cycle analysis, 8 subareas were selected and investigated within the total study area. Models were capable of representing the annual cycle in all subareas, particularly between July and October. However, greater spread was observed in the first half of the year, especially in the north of the Intertropical Convergence Zone (ITCZ) and in CMIP models. Based on those analyses, 3 best models from each ensemble were selected to evaluate the model's ability to represent EWD over the study area: CMIP (AS-RCEC.TaiESM1, CAS.FGOALS-f3-L, and NCC.NorESM2-MM), AMIP (CAS.FGOALS-f3-L, AS-RCEC.TaiESM1, and CMCC.CMCC-CM2-SR5), hist-1950 (HadGEM3-GC31-HH, CMCC-CM2-VHR4 e ECMWF.ECMWF-IFS-HR) and highresSST (MRI.MRI-AGCM3-2-H, CMCC-CM2-VHR4 e ECMWF.ECMWF-IFS-HR). An objective automatic tracking algorithm was utilized in each model. It was observed that the models were able to represent mean values of lifetime (~6 days) and phase speed (~7 m.s-1) close to climatological values and those of ERA5. However, most models failed to capture the EWD's interannual variability or climatological mean frequency. All models successfully identified the two preferential genesis regions of this system, one over the TSA and another near the West African coast. Thus, despite many improvements still being necessary in CMIP6 models, precipitation climatological patterns and EWD's characteristics over NEB and adjacent TSA were represented by the models, especially in atmospheric and high-resolution models (HighResMIP).