Satellite observations of clouds in the tropics reveal a hierarchy of organized cloud clusters ranging from the planetary-scale Madden-Julian Oscillation (MJO), through synoptic-scale convectively coupled equatorial waves (CCEWs), to mesoscale convective systems (MCSs, S. S. Chen et al., 1996;Mapes & Houze, 1993;Nakazawa, 1988). These various types of convective organization are not necessarily independent from one another, and MCSs, with scales of 100s of kilometers, are often considered building blocks of those larger-scale tropical circulations (Mapes et al., 2006). The most intense MCSs often cause high-impact weather events. The frequency of these rainfall extremes has increased in the past few decades and is projected to rise with climate change (Z. Feng et al., 2016;Tabari, 2020;Taylor et al., 2017). While MCSs are difficult to predict in current global weather and climate models due to their smaller scales and dependence on subgrid-scale processes, larger-scale disturbances such as CCEWs can be predicted with higher skill beyond 1 week (