Thermobaric Instability, Cabbeling, OCAPE, and Abrupt Climate Change

We define ocean convective available potential energy (OCAPE) as the maximum stored potential energy (PE) that may be converted into kinetic energy (KE) under an adiabatic vertical rearrangement of ocean parcels. We propose that the release of OCAPE is an important mechanism for ocean deep convection and the resulting deep-water formation in polar regions. We present a scheme that efficiently calculates OCAPE for arbitrary profiles. We run numerical models of fully resolved deep convection events associated with the release of OCAPE and we present a theory of its energetics—how thermobaricity, cabbeling, and the background stratification interact to determine what fraction of the initial OCAPE is converted into kinetic energy, what fraction is destroyed by turbulent diffusion, and what fraction remains after the convection. In one version of our model, we develop strategies to conserve total energy exactly—to the round off error of the machine—even in the presence of viscosity and thermohaline diffusion (Su and Ingersoll. Ocean convective available potential energy. Part I: Concept and calculation, and Part II: Energetics of thermobaric convection. J. Phys. Oceanogr., in revision, 2015).

We apply these concepts to the last deglaciation, which was a two-step process: an abrupt warming taking place over a few years to decades (Bolling-Allerod), followed by a gradual cooling over a 3000 year period (Younger Dryas), followed by another abrupt warming to near-modern conditions. The ice sheets have been implicated in this rapid climate change, but a recent study of ancient corals has implicated the deep oceans (Thiagarajan et al. Abrupt pre-Bolling-Allerod warming and circulation changes in the deep ocean. Nature 511, 75-81, 2014). These data support the hypothesis that OCAPE was present in the North Atlantic during the last glacial maximum (LGM). We propose a mechanism whereby the abrupt release of OCAPE can explain the abruptness of the two warming events. We have run our model with realistic parameters, and the results compare favorably with the deglaciation record. We also show that OCAPE exists in state-of-the-art climate models of the LGM. By using the output from a transient simulation with CCSM3 (He et al. Nature 494, 81–85, 2013) of the climate from the LGM to present, we found a basin-scale OCAPE pattern of large amplitude (~0.01 J/kg) in the North Atlantic at the end of the stadials, before the sudden warming events. This confirms our hypothesis about the deglaciations and highlights the need to parameterize deep convection properly in ocean GCMs.