Explaining climate change in South America

Projections from a CMIP6 multimodel ensemble show weak signal of climate change in annual and seasonal precipitation over most of South America (SA), with low agreement among models as to the sign of the change over most of the continent. Besides, climate change information from different analyses frequently seems confusing for the public and decision makers. Since climate has a crucial influence on important economic sectors in SA, such as hydroelectric power generation and agriculture, and natural disasters associated with extreme events of drought and excessive rainfall have become more frequent and intense, the future climate behavior should be more clearly described, and supported by a dynamical framework able to explain it, so as to better serve decision-makers in planning actions and adopt effective policies for climate adaptation.

Although weak and with low agreement between models, the climate change projected by the CMIP6 multimodel ensemble for SA shows similarity with the seasonal impacts of El Niño (EN) events on precipitation. Since model projections of future SST indicate an El Niño-like warming pattern in the central-east equatorial Pacific, it is reasonable to hypothesize that changes in precipitation over South America would have the patterns of EN impact and would be mainly due to the strengthening of an EN-type SST anomaly pattern in the Pacific Ocean.

Therefore, to clearly determine the future climate changes, it is necessary to select models that not only simulate well the SA climatology, but also the El Niño-Southern Oscillation (ENSO) and its teleconnections with SA, since ENSO is responsible for most of the climate variability in SA. The assessment covered 31 models that provided at least three runs from the present  (1979-2014) to the future climate (2065-2100). Based on relevant and comprehensive criteria, the models were classified according to both assessments (climatology and ENSO), and four best-performing models were selected.

The changes projected by the ensemble of best models indicate a more EN-like future climate, in which the main climate changes projected for SA resemble the observed EN impacts, remarkably including the tendency to spring-summer reversal of precipitation anomalies in Central-East SA, from dryer spring to wetter summer. While the total monsoon precipitation shows little or no change in this region, there is reduction (enhancing) of early (peak) monsoon rainfall, resulting in a delay and shortening of the monsoon season. The spring response in this region is due to the dynamical effect of the EN-like SST changes via teleconnection, and the reversal in summer is triggered by surface-atmosphere interactions. Also coherently with EN impacts, drier conditions prevail in central-northern-eastern Amazon throughout the monsoon season thanks to changes in the Walker circulation, while in southeast SA, precipitation increases due to tropics-extratropics teleconnection.

The changes projected by the all-model ensemble are much weaker and confusing. This clear description of climate change and its dynamical connection with intensified EN effects give coherence to the different changes throughout different seasons, which otherwise seem incomprehensible and can lead to discrepant interpretations if not understood within a correct dynamic context.