The roughly half-dozen global climate models participating in MSMIP seek to answer the question “What aspects of sub-monthly sea surface temperature anomalies (SSTAs) are most important for eastward propagation of the Madden-Julian oscillation (MJO)—their patterns, phasing, or persistence?” In this MIP, sub-monthly daily mean SSTAs generated by an ocean-atmosphere coupled simulation of a given climate model are altered by 1) removal from the low-frequency background state, 2) shuffling daily SSTA patterns in time, and 3) shuffling grid-point SSTAs in both space and time. Each altered SSTA time series is added back to the slowly varying background SST time series, and prescribed to an atmosphere-only simulation of the same model, ensuring identical SST climatology and low-frequency variability in all experiments for that model.

Changes to MJO propagation are evaluated using standard metrics to identify which SSTA characteristics (pattern, phasing, persistence) are most associated with MJO simulation skill. SSTA feedbacks to the MJO are explored in terms of their direct influence on surface fluxes throughout the MJO lifecycle as well as their influence on the low-frequency spatial distribution of tropical water vapor. Models analyzed to date exhibit varying degrees of sensitivity to altered SSTAs, and thus give different impressions of how strongly the MJO is coupled to the ocean.