Impact of Submesoscale Structures on Global Climate

Present-day climate models explicitly take into account mesoscale (~50-300km) and larger-scale motions that drive most of the horizontal heat transport in the ocean. Because of computational limitations, oceanic submesoscale turbulence (≲ 50km) is traditionally parameterized by vertical heat transport from warm to cold waters. This, however, is challenged by recent studies that identify an opposite nature of submesoscales: their vertical heat transport is essentially from cold to warm waters. The impact of submesoscales on global ocean heat transport, e.g., on near-surface temperature and air-sea fluxes, is still unclear, but is implicitly suggested important by parameterization studies (e.g., Fox-Kemper et al. 2011. Ocean Modell., 39, 61–78).

Here we show that submesoscale turbulence (≲50km) transports heat upward towards the surface at an exceptional significant rate ubiquitously worldwide. Our results are based on an unprecedented global simulation with the highest resolution ever obtained (~2km horizontally). It explicitly resolves, for the first time, the temporal and spatial variabilities of global ocean turbulence down to a scale of ~5 km. Results highlight remarkable fast-rotating submesoscale turbulence and a strong associated vertical velocity field (~50-200 m/day) worldwide near the air-sea interface, especially during the winter. This occurs not only in the Gulf Stream and Kuroshio Current as already reported, but also in much broader unexplored regions that are critical for air-sea interaction, including the Southern Ocean, the high-latitude North Atlantic, the broad subtropical oceans, and the Mediterranean, Black and Arabian Seas. Submesoscales produce strikingly omnipresent, high-amplitude, systematically-upward heat fluxes over the global upper ocean, ~30-300 W/m² for winter-season averages and ~500-1000 W/m² for abrupt intermittencies at days to weeks’ timescales. These magnitudes are comparable to or much larger than all other heat transport mechanisms such as air-sea fluxes and mesoscale heat fluxes. Our control experiment indicates that submesoscale heat fluxes, when resolved just slightly (~14%) better, already cause a remarkable global response: a sea surface warming (~0.4 °C) balanced by a stronger upward ocean-atmosphere heat exchange (~20 W/m²). The regional temporary responses can reach ~3 °C and ~150 W/m², respectively. Our finding is demanding an accurate inclusion of submesoscales in improving the assessment of the overall rate, the location, and the timing of global ocean heat transport and the exchange with the atmosphere.

We also find that submesoscale dynamics provides an important kinetic energy source on a global scale that strongly strengthens large-scale upper-ocean circulation and hence influences the associated lateral tracer transports. Submesoscales in major ocean basins are often produced, not only by classical mixed-layer baroclinic instabilities, but also by frontal instabilities at much smaller wavelengths. They can be considerably influenced by surface wind.