Precursor mechanisms and multidecadal modulation of the Atlantic Meridional Mode-Atlantic Zonal Mode connection

Boreal spring and summer tropical Atlantic variability is governed by the Atlantic Meridional Mode (AMM) and Atlantic Zonal Mode (AZM) at interannual timescales. Previous studies have identified a connection between AMM and AZM through ocean wave propagation and local wind forcing (Foltz and McPhaden 2010; Burmeister et al. 2016; Martín-Rey and Lazar 2019; Martín-Rey et 2023). For a positive phase of the AMM, anomalous negative wind curl north of the equator triggers a downwelling Rossby wave that propagates westward and is boundary-reflected into an equatorial Kelvin wave (KW). This dKW travels along the equator during summer months, activating the oceanic processes responsible to warm the sea surface, thus favouring the development of a positive AZM. However, the existence of anomalous local zonal winds could modulate the impact of the dKW, and consequently, the phase of the AZM following the AMM (Martín-Rey and Lazar 2019; Martín-Rey et al. 2023).

Here, we use the Maximum-Covariance based Python statistical tool Spy4Cast to explore the existence of a AMM-AZM connection, as well as, the relative role of each precursor (local wind vs oceanic waves). Spy4cast (Durán-Fonseca and Rodríguez-Fonseca 2025)  allows for identifying coupled modes of variability as well as to produce statistical predictions. In this way the AZM predictability will be assessed together with the stability of the connection. Thus, the non-stationary behaviour of this connection will be evaluated, as well as the favourable background conditions for each type (positive or negative) of AMM-AZM interaction. Observational datasets, and long-simulations from PIcontrol and historical CMIP6 simulations will be used.