First GCM simulations of Uranus and Neptune atmospheres with realistic radiative transfer

Uranus and Neptune"s atmospheres are active worlds, with vigorous meteorological activity and strong zonal winds occurring despite small absorbed solar radiation and internal heat fluxes. A few 3-D General Circulation Models (GCM) of their atmospheres exist in the literature, focusing mostly on understanding their zonal jet structure [1,2] or the evolution of large disturbances [3,4].Building a complete and realistic GCM is a challenging task, given the long orbital and radiative timescales involved, along with the rather high spatial and temporal resolution needed when solving the atmospheric equations of motion on the rotating sphere. For this reason, existing GCMs include crude representation of radiative transfer (a simple relaxation scheme to an equilibrium temperature profile) and/or neglect seasonal variations. We are currently developing a GCM for Uranus and Neptune"s atmospheres, building on our existing expertise on Jupiter and Saturn GCMs [5,6]. Compared to other existing GCMs for ice giants, our model includes state-of-the art parametrization of radiative transfer. The radiation scheme is a full radiative transfer using correlated-k distributions. Seasonal variations of the incoming solar flux are taken into account. Opacity sources include gaseous opacity from methane, ethane, acetylene, H2-H2, H2-He continua along with opacity from two aerosol layers: one optically thick cloud with a base at the 2-bar level and one optically thin haze layer with a base at 300 mbar. These layers are consistent with the putative H2S and CH4 clouds reported by many observational studies (eg [7,8]). Simulations at radiative-equilibrium are discussed in a companion abstract [9] ; in this one we focus on dynamical aspects. We will present results from first 3D GCM simulations performed at a horizontal resolution up to 256x192 in longitude x latitude (corresponding to 1.4°x0.9°), extending from 3 bars to 0.3 mbar. A broad equatorial retrograde jet develop on both Uranus and Neptune and two prograde jets emerge near 50° latitude in the Neptune simulation. This is in qualitative agreement with the observed zonal wind structure on Neptune, although the zonal jet wind speeds are much smaller than the observed ones. We are able to show that acceleration by eddies is an important contributor to the two prograde jets in the Neptune simulation.However, the Uranus simulation does not exhibit high-latitude prograde jets that have been reported by cloud-tracking observations. In other words, the zonal jet structure currently obtained in our simulations differs significantly between the two planets, which is puzzling and at odds with their qualitatively similar observed zonal wind structures. This might indicate that important processes governing the atmospheric circulation of ice giants is missing in our GCM.Another outcome of these simulations is that all tropospheric zonal jets are slowed down to near zero wind speeds in the lower stratosphere. The reason behind this behaviour is under investigation, as is the associated meridional circulation. Next steps will include the study of the role of Uranus and Neptune respective axial tilts and internal heat fluxes (or lack thereof) on their circulation. Furthermore, our GCM is still lacking important processes, such as latent heat release from water and other condensing species, and is lacking a realistic parametrization for convective processes. This might explain the observation-model mismatches in their zonal wind structure and will be the subject of future developments. [1] Lian and Showman, Icarus, Vol. 207, Issue 1, p. 373-393, 2010.[2] Liu and Schneider, Journal of the Atmospheric Sciences, Vol. 67, issue 11, pp. 3652-3672, 2010.[3] Lebeau and Dowling, Icarus, Vol. 132, Issue 2, pp. 239-265, 1998.[4] Hammel et al., Icarus, Vol. 201, Issue 1, p. 257-271, 2006.[5] Spiga et al., Icarus, Vol. 335, article id. 113377, 2020.[6] Guerlet et al., accepted in Icarus, 2020.[7] Irwin et al., Icarus, Vol. 227, p. 37-48, 2014.[8] Sromovsky et al., Icarus, Vol. 317, p. 266-306, 2018.[9] Vatant d"Ollone et al., EPSC 2020