Abstract/Summary

The South Asian monsoon is strongly influenced by synoptic-scale low-pressure systems (LPSs), which deliver the majority of seasonal and extreme rainfall. Here, we provide a robust synthesis of future LPS characteristics by analysing 13 models from the Coupled Model Intercomparison Project Phase 6 (CMIP6). We develop a novel, multi-component skill score (FLIP: Frequency, Location, Intensity, Precipitation) to evaluate and weight model performance against reanalysis benchmarks. Our weighted multi-model projections reveal a consistent narrative for future monsoons: LPSs are projected to become more frequent and dynamically weaker in terms of low-level vorticity as global warming levels increase. Per-LPS precipitation is projected to increase, with a super-Clausius-Clapeyron scaling for the most intense storms, despite a general weakening of their circulation. We attribute this apparent paradox to a projected increase in the background meridional moisture gradient and a structural change whereby boundary layer winds weaken less than those in the free troposphere, maintaining strong frictional moisture convergence. Furthermore, LPSs are projected to penetrate deeper inland, with a significant increase in post-landfall duration between 1.5°C and 2°C of warming. To deconstruct inter-model uncertainty, we employ a storyline approach based on unsupervised clustering of circulation and moisture anomalies. This reveals four physically self-consistent yet distinct future pathways that explain model diversity in LPS frequency, track, and precipitation changes. For example, a storyline with enhanced easterlies over Southeast Asia leads to a 36% increase in LPS frequency by transporting more precursor disturbances into the Bay of Bengal. Averaging over these diverse futures, the consistent outcome is a monsoon dominated by more numerous and wetter LPSs that track deeper into the subcontinent, expanding the risk of extreme rainfall and flooding to western parts of India.