Impact of a Stochastic Nonorographic Gravity Wave Parameterization on the Stratospheric Dynamics of a General Circulation Model

Plain Language Summary Present-day climate models are not able to simulate small-scale phenomena, which are represented by means of simplified descriptions. Among these processes, atmospheric gravity waves are important for the upper atmosphere, as they influence its structure, interacting with the mean flow. These small waves are often associated to convection, but their intermittent nature is not always present in climate models. Here we describe how this property can be included in the representation of gravity waves and which are the effects on the simulated upper atmosphere.

The general circulation of the middle atmosphere, particularly of the mesosphere, is strongly dependent on the forcing arising from gravity wave processes. Their sources in the troposphere are both orographic and nonorographic, the latter being strongly intermittent. In climate models, the effects of gravity waves need to be parameterized, often assuming that their properties are constant. In this work we focus on the changes of the middle atmosphere due to the introduction of intermittency in a parameterization of nonorographic gravity waves, using a stochastic version of the Hines scheme. The stochastic approach is tailored to the diagnosed sensitivity of the model to this forcing, and peculiar changes emerge even if a relatively small amount of intermittency is prescribed. We analyze in detail the changes of the stratospheric dynamics in the tropical region and the global circulation of the middle atmosphere, when the stochastic parameterization is employed in place of the deterministic one. The mean state and variability of the model, realistic also in the default version, are preserved when stochasticity is added. Significant changes are observed in the mesosphere due to an enhanced poleward transport, leading to warming in the winter season. While there are some improvements of the mean state, the interannual variability is not significantly affected in the extratropics. The impacts on the simulated equatorial stratosphere are evident, as the stochasticity reduces the overall period of the quasi-biennial oscillation but also leads to a net reduction of the variability between cycles.