Is there a relationship between the intermodel spread of biases and historical simulation of Indian Summer Monsoon Rainfall in CMIP6 coupled models ?

Robust projections of the Indian Summer Monsoon Rainfall (ISMR) are critical as it provides 80 % of the annual precipitation to more than one billion people who are very vulnerable to changes. However, even over the historical period, CMIP (Coupled Model Intercomparison Project) coupled models have difficulties to reproduce the observed ISMR trends and are affected by a large inter-model spread, which question the reliability of the ISMR projections. When studying climate response, three main sources of uncertainties exist : scenario uncertainties, internal variability and models bias. We study the impact of the latter on the historical response of ISMR of 34 models from CMIP6. First we show that model local biases over India do not impact significantly how they simulate the response of ISMR over the recent period. However, when we enlarge the analysis to the whole tropics and study the impact of regional and remote rainfall and SST biases on ISMR historical response by using a Maximum Covariance Analysis (MCA), we do find statistical significant relationships, which may provide observational constraints on future ISMR projections. Our results highlight the key-role of the temperature gradient errors between the arid regions surrounding India and the Arabian Sea on one hand, and of Pacific rainfall and SST biases on the other hand, as an important source of inter-model spread in the ISMR response. The physical mechanisms underlying these statistical relationships between ISMR response and the inter-model spread are finally explored.