Evaluating Wind Speed Variability and Its Climate Modulators in South America Using CMIP6 Simulations (1980–2014)

Abstract

In this study, we applied an Empirical Orthogonal Functions (EOF) analysis on the behaviour of monthly wind speed at 100 m (WS100) in selected regions of South America to detect its modulation by different atmospheric patterns. Reanalysis data from ERA5 and 14 global climate models from CMIP6 were used for the period 1980–2014. We found that the South Atlantic Anticyclone (SAA) is the main WS100 modulator over northeastern Brazil with a direct relationship (correlation coefficients up to 0.55) between the SAA intensity and the 1st principal component (PC1), which explains 59.1% of the WS100 total variance in that region. This means that as SAA intensity increases, WS100 increases as well over northeast Brazil and vice versa. We discovered a pattern of sea level pressure anomalies centred around the Drake Passage, which we named the Patagonia-Antarctic pattern (PAT), and it is the main WS100 modulator over Patagonia. Significant negative correlations (up to −0.68) were found between PAT and PC1, which explains 52.2% of the total variance for that region. The positive phase of the PAT index causes a decrease of WS100 over Patagonia, while WS100 increases during the negative phase of PAT. We also determined that the Antarctic Oscillation (AAO) is the second variability mode that determines the WS100 field over Patagonia, given by high correlation coefficients (up to 0.52) between AAO and PC2, which explains 22.8% of the total variance. The positive phase of AAO causes an increase of WS100 over the southern tip of Patagonia, and a decrease north of 45°S; the opposite occurs during the negative phase of AAO. These results are of great relevance for a better understanding of wind variability causes in South America, which is essential for wind resource evaluation, long term system planning and risk assessment in current and future wind power projects.