The South American Monsoon as an Atlantic-Pacific Bridge: Causal Pathways to ENSO Diversity

Predicting ENSO diversity remains a fundamental challenge in seasonal-to-interannual climate forecasting, with Eastern (EP) and Central Pacific (CP) events arising from complex ocean-atmosphere interactions and remote forcing. This study highlights the significant role of the South American Monsoon System (SAMS) in modulating ENSO diversity.

Analysis of ERA5 reanalysis data (1940-2024) reveals distinct ENSO precursors during the monsoon onset phase in September-November (SON), five seasons before peak ENSO conditions. Enhanced SAMS precipitation, coupled with cold subtropical Southwestern Atlantic SST anomalies, precedes EP events. Conversely, an anomalous upper-level vortex over subtropical South America (VOSA) combined with positive sea-level pressure anomalies across the subequatorial and subtropical South Atlantic is more closely linked to CP events. These precursor patterns generate a stationary Rossby wave that initiates a cascade of processes in the South Eastern Tropical Pacific, including anomalous vertical motion, weaker trade winds, wind-evaporation-SST feedback, and Ekman coastal dynamics, culminating in ENSO development 12-15 months later.

To rigorously establish causation among these ENSO precursors and other known tropical Atlantic-Pacific basin interaction mechanisms, the Peter and Clark Momentary Conditional Independence (PCMCI+) algorithm was applied. The resulting causal networks largely validate the hypothesized physical mechanisms, also identifying key boreal spring mediators, such as SAMS precipitation, VOSA, and the South Pacific Oscillation. Sliding window analyses reveal a post-1980 intensification of the SAMS precipitation pathway, coinciding with shifts in the Atlantic-Pacific background state and satellite-based observational coverage. The PCMCI+ results challenge the causal significance of conventional precursors such as the Atlantic Niño and South Atlantic Subtropical Dipole, while emphasizing the atmospheric bridge connecting the subtropical South Atlantic and the tropical Pacific through monsoonal precipitation over South America.

Parallel analyses conducted on 20 historical-period simulations from the CESM2 Large Ensemble test these relationships under internal climate variability. Each member was subjected separately to the same method sequence applied to ERA5. While some members corroborate key findings—particularly the role of VOSA—the magnitude and exact timing of some causal links differ from reanalysis results. These discrepancies reflect both internal climate variability and model biases in representing tropical climate modes, including insufficient ENSO diversity, misrepresented teleconnections (e.g., an overemphasized role of the North Pacific Oscillation), and low SAMS variability.

These findings demonstrate that ENSO variability originates partly from cross‐basin processes initiated before the spring predictability barrier, highlighting the potential for enhanced early-season forecasts through incorporation of SAMS transition phase intensity. Despite remaining uncertainties regarding connection strength and multidecadal variability, this study establishes land-atmosphere interactions as significant contributors to pantropical climate interactions, warranting broader investigation of monsoon-ENSO pathways. Furthermore, it advocates for continued efforts to ensure that climate models accurately capture observed patterns and their underlying causal relationships.