Cloud regimes and associated MJO variability over the Maritime Continent in austral summer

In this study, a new technique to determine distinct cloud regimes and their variation in space and time is proposed, evaluated, and applied to two satellite products over the Maritime Continent (MC). Compared to previous methods, the method presented here allows different types of cloud to co-exist in the same grid at the same time, giving rise to physically explainable and spatially continuous patterns in cloud regimes. Similar results generated by ISCCP – H and Himawari 8 data suggests that the method is robust. The 4 cloud regimes determined using this method are associated with shallow, mid-level, deep convective and high level clouds respectively. The analysis shows that he MJO–induced variation in total cloud fraction is dominated by day-time high–level clouds, while the diurnal MJO variability is mostly demonstrated by low–level cumulus. Spatial and temporal rainfall variability over the MC during austral summer is dominated by high–level clouds, while most local signatures and land–sea differences are attributed to deep convective clouds. Using an artificial neural network, the cloud patterns over the MC can be classified into nine categories, largely dominated by the MJO-phase. Active MJO activity is shown by systematic propagation around the cloud categories, with one category associated with the inactive MJO phase. The inhomogenous propagation of the MJO can partially be revealed in the generated patterns, which can be physically explained by the enhanced/suppressed convection over the Indian Ocean. This work has implications for understanding the MJO-scale variation in precipitation and diabatic heating associated with different cloud regimes, and its representation in mesoscale and climate scale modelling systems.