Rethinking Offshore Propagation of Tropical Convection
How can thunderstorms born over tropical mountains vanish near the coast, only to reappear hours later over the open ocean and continue traveling for hundreds of kilometers? For decades, this remarkable offshore propagation has often been viewed as a single, continuous journey. My research suggests a very different picture. Rather than one uninterrupted convective system, offshore propagation over New Guinea unfolds as a two-stage relay: a land mode, in which afternoon convection develops over the mountain slopes and propagates toward the coast, and an ocean mode, in which convection regenerates over the warm ocean after sunset. This distinction helps connect the familiar double peak of tropical island rainfall–one over land and another offshore at night–to two physically distinct propagation processes. It also changes the central question from how storms keep moving continuously to how one convective mode gives way to another.
This reframing requires a different way of thinking about the coastal flows that organize tropical convection. In this work, I reinterpret sea breezes and land breezes not merely as daily wind reversals, but as coastal density currents with distinct structures, asymmetries, and interactions with convective outflows. The sea breeze is not simply the daytime counterpart of the land breeze: it can interrupt the land mode by carrying cool marine air inland, while the nocturnal land breeze behaves as a shallow offshore density current capable of merging directly with storm-generated cold pools. This interaction produces what I call a hybrid land breeze–a cold-pool-enhanced nocturnal density current that couples with moist boundary-layer thermodynamics, organizes moist patches along its leading edge, supports convective regeneration over the ocean, and helps transform a local diurnal circulation into a mechanism for long-distance offshore propagation.
This study also introduces a novel two-and-a-half-variable framework that identifies the spatial relationship between old and newly developing convective cells and tracks how cold pools connect them directly from high-resolution simulations, extending my previous diagnostic work while avoiding computationally expensive Lagrangian particle tracking. It assesses whether hybrid land breezes exhibit density-current characteristics and contain embedded cold-pool components as well.
More broadly, this work emphasizes the nonlinear processes that emerge after convective initiation. Once storms form, their cold pools, moist patches, and gust fronts actively reshape the boundary layer, influencing where convection regenerates and how long it remains organized over the ocean. From this small-scale perspective, tropical offshore propagation becomes a cross-scale interaction problem in which convective-scale cold pools, island-scale land-sea breezes, and the lower-tropospheric cross-equatorial monsoon interact within the moist tropical boundary layer. Together, these processes suggest that tropical offshore propagation is best understood not as the evolution of an individual storm, but as an emergent phenomenon arising from interactions across convective, island, and monsoon scales. For more details, check our recent publication Tang et al. (2025). For a more accessible analogy, see this plain-language explanation of the storm relay from the mountains to the deep ocean.