Intra-seasonal oscillations of South Asian summer monsoon in coupled climate model cohort CMIP6

Understanding of intra-seasonal oscillations (ISOs) acts as a crucial bridge in deciding the fate of the seasonal total rainfall during monsoon season in densely populated South Asian monsoon (SAM) domain. Based on daily precipitation data sets from state-of-the-art coupled climate model cohort that participated in Coupled Climate Model Inter-comparison Project Phase 6 (CMIP6), this study brings out the efficacy of these models in resolving the intra-seasonal signatures of monsoon season rainfall. Out of 27 CMIP6 models considered in the present study, only 21 are able to simulate the annual cycle well within the acceptance bound derived from the observations based rainfall. CMIP6 models Can-ESM5, HadGEM3-GC31, MRI-ESM2, UKESM1, ACCESS-CM2 and ACCESS-ESM1-5 are failed to resolve the onset phase of the monsoon in the month of June. Subsequent analysis revealed reasonable skills of most of the 27 CMIP6 models in demarcating 2 dominant intra-seasonal oscillations (ISOs) of monsoon viz 10–20 and 30–60 days. It is noted from the present analysis that CMIP6 models resolve the 10–20 days ISOs signal significantly but unexpectedly show conspicuous shift in 30–60 days ISOs with respect to the observational data sets over the central Indian region which results in extended dry spell in the beginning of the monsoon season. Westward propagating 10–20 days and Northward propagating 30–60 days ISOs characteristics are well simulated by 50% of CMIP6 models considered in the present study. Multi model mean of 27 CMIP6 models seems to preserve the shape of the distribution of 10–20 days ISOs in good agreement with the observational datasets as compared to low frequency 30–60 days oscillations, nevertheless, considering huge inter-model variations in ISOs particularly for 30–60 days ISOs and inability of CMIP6 models in detecting climatological wet and dry spells of ISOs, caution must be taken while delineating the intra-seasonal variability of the South Asian monsoon using CMIP6 models. Inter-model spread amongst the CMIP6 models considered in the present study may be attributed to the pathways of the interaction of model components, variants, model physics and representation of feedback mechanism.