ABSTRACT The Indian summer monsoon (ISM) governs the hydrological and agricultural foundation of South Asia, yet its response to anthropogenic warming remains uncertain and regionally diverse. Here we assess projected mean rainfall, variability and categorised monsoon extremes using six skill‐selected CMIP6 models (BCC‐CSM2‐MR, CMCC‐CM2‐SR5, FIO‐ESM‐2‐0, MCM‐UA‐1‐0, MIROC6, TaiESM1) evaluated against the high‐resolution IMD 0.25° dataset. Analyses are performed for the seasonal (JJAS) and monthly (June–September) scales under four Shared Socioeconomic Pathways (SSP1‐2.6, SSP2‐4.5, SSP3‐7.0, SSP5‐8.5). Century long observed records reveal a quasi‐stationary all‐India mean monsoon rainfall, accompanied by a marked increase in interannual variability consistent with progressive amplification of hydroclimatic extremes. The CMIP6 multi‐model ensemble (MME) reproduces this behaviour and projects a further 20%–35% rise in variability by the late 21st century. The coefficient of variation strengthens by 0.04–0.06 under high‐emission scenarios, with statistically significant regime shifts detected around 2040–2060. Enhanced seasonality is evident, with June and September exhibiting pronounced drying (≈−2 σ ) and July–August showing intensified wet excursions (≥ +3 σ ). Categorical analyses indicate a 40%–60% increase in moderate and extreme excess monsoon years and a 20%–30% rise in deficit years, while event intensities strengthen to +4 σ (wet) and –3 σ (dry). These concurrent amplifications of surplus and deficit rainfall define a hydroclimatic paradox: simultaneous amplification of mean rainfall and interannual variability, leading to heightened probabilities of both surplus and deficit extremes. The findings directly tangled with the reduced seasonal predictability and challenges for reservoir management and crop planning, as evidenced by projected increases in extreme excess years and intensified deficits. The warming climate is reorganising the ISM towards a high‐variance regime, challenging water security and adaptation planning across a region that depends critically on its seasonal rainfall predictability.
Kulkarni et al. (Mon,) studied this question.