Robust networks of rainfall extremes emerge despite fragile ocean monsoon causality under Internal variability

Climate ready multi sectoral risk management relies on understanding not only where extreme rainfall occurs but also how such events synchronize across regions and interact with internal climate modes. The stability of synchronized rainfall networks and remote climate forcings under internal climate variability (ICV), especially when aggregating ensemble simulations for statistical robustness, remains poorly understood. Using a 50-member EC-Earth3 large ensemble, we show the co-existence of robust and weak climate networks. Despite significant variability in regional rainfall intensities, extreme rainfall spatial synchronization patterns remain remarkably robust, driven by intrinsic atmospheric processes such as Rossby wave propagation. In contrast, slower teleconnections, notably from the El Niño Southern Oscillation and the Indian Ocean Dipole, exhibit marked variability under minor initial condition perturbations, often altering or reversing causal relationships. This fundamental asymmetry reveals the critical limitations of relying on deterministic or single-member model outputs. Our findings advocate ensemble-based frameworks to disentangle robust dynamical pathways from fragile ones, offering a more reliable foundation for climate projections, model evaluations, and sectoral resilience planning under uncertain climate futures.