Radiocarbon bomb-peak signal in tree-rings from the tropical Andes register low latitude atmospheric dynamics in the Southern Hemisphere

South American tropical climate is strongly related to the tropical low-pressure belt associated with the South American monsoon system. Despite its central societal role as a modulating agent of rainfall in tropical South America, its long-term dynamical variability is still poorly understood. Here we combine a new (and world's highest) tree-ring ¹⁴C record from the Altiplano plateau in the central Andes with other ¹⁴C records from the Southern Hemisphere during the second half of the 20th century in order to elucidate the latitudinal gradients associated with the dissemination of the bomb ¹⁴C signal. Our tree-ring ¹⁴C record faithfully captured the bomb signal of the 1960's with an excellent match to atmospheric ¹⁴C measured in New Zealand but with significant differences with a recent record from Southeast Brazil located at almost equal latitude. These results imply that the spreading of the bomb signal throughout the Southern Hemisphere was a complex process that depended on atmospheric dynamics and surface topography generating reversals on the expected north-south gradient in certain years. We applied air-parcel modeling based on climate data to disentangle their different geographical provenances and their preformed (reservoir affected) radiocarbon content. We found that air parcel trajectories arriving at the Altiplano during the bomb period were sourced i) from the boundary layer in contact with the Pacific Ocean (41%), ii) from the upper troposphere (air above the boundary layer, with no contact with oceanic or continental carbon reservoirs) (38%) and iii) from the Amazon basin (21%). Based on these results we estimated the ∆¹⁴C endmember values for the different carbon reservoirs affecting our record which suggest that the Amazon basin biospheric ¹⁴C isoflux could have been reversed from negative to positive as early as the beginning of the 1970's. This would imply a much faster carbon turnover rate in the Amazon than previously modelled.