Precipitation trends in Southern South America in the last centuries: attribution and mechanisms

Southern South America (SSA) is one of the regions of the world where the largest trends in precipitation have been recorded during the last 120 years. While South-Eastern South America (SESA) has been affected by a noticeable increase in austral summer rainfall, a remarkable decrease has been observed in Southern Andes (SAn). Moreover, long-term precipitacion has been registered in subtropical Andes and Altiplano regions, which show wetter periods during the 17th century in the Little Ice Age (LIA) and dryer periods during the current Global Warming Period (GWP). In spite of the large impacts related to these trends, the attribution of them is still an open-question. 

This work will assess the attribution of the observed austral summer rainfall trends in SSA to anthropogenic and natural forcings using models available in World Climate Research Programme (WCRP) Coupled Model Intercomparison Project - Phase 5 (CMIP5) and Phase 6 (CMIP6). Analysed experiments include Historical, Pre-Industrial Control and Last Millennium simulations to study long-term changes, as well as the Detection and Attribution Model Intercomparison Project (DAMIP) to assess the attribution of last-century trends. 

The assessment of the Last Millennium simulations allows to detect the following changes in LIA (GWP): (a) equatorwards (polewards) displacement of the southern branch of the Hadley cell, in turn associated with wetter (drier) conditions in subtropical south America; (b) negative (positive) upper-level zonal wind changes related with positive (negative) December, January and February (DJF) rainfall changes in the Altiplano; and (c) positive (negative) low-level zonal wind changes associated to positive (negative) JJA rainfall changes in the subtropical Andes, being in turn related to hemispheric wind changes resembling a negative (positive) phase of the Southern Annular Mode (SAM). The last century changes in the Altiplano reveal a signal associated with the anthropogenic forcing in upper-level zonal wind trends, but it is weak as compared with the internal climate variability. 

Regarding last century trends, positive (negative) rainfall trends in SESA (SAn) are identified in most historical simulations. For both regions, greenhouse-gases-forcing-only simulations show trends consistent with all-forcing simulations, while natural-forcing-only simulations exhibit negligible values. SESA (SAn) shows negative (negligible) trends associated with aerosol-forcing-only simulations and high uncertainty (negative trends) for stratospheric-ozone-forcing-only simulations. Moreover, SAn rainfall trends could be also connected to consistent trends of opposite sign for the Southern Annular Mode (SAM). Overall, our results provide evidence for anthropogenic influences on SSA rainfall trends.