[M-GI35] Computational sciences on the universe, galaxies, stars, planets, and their environments

[MGI35-01] Development of a non-hydrostatic global model of the Venus atmosphere: A simplified Venus simulation

*Hiroki Kashimura¹, Hisashi Yashiro², Seiya Nishizawa³, Hirofumi Tomita³, Masahiro Takagi⁴, Norihiko Sugimoto⁵, Kazunori Ogohara⁴, Takeshi Kuroda⁶, Kensuke Nakajima⁷, Masaki Ishiwatari⁸, Yoshiyuki O. Takahashi¹, Yoshi-Yuki Hayashi¹ (1.Department of Planetology, Graduate School of Science, Kobe University, 2.Center for Global Environmental Research, National Institute for Environmental Studies, 3.RIKEN Center for Computational Science, 4.Faculty of Science, Kyoto Sangyo University, 5.Department of Physics, Keio University, 6.Department of Geophysics Tohoku University, 7.Department of Earth and Planetary Sciences, Flculty of Sciences, Kyushu University, 8.Department of Cosmosciences, Graduate school of Science, Hokkaido University)

Keywords:Venus atmosphere, Non-hydrostatic global model, high-resolution simulation

Venus is fully covered by thick clouds of sulfuric acid, and its atmospheric circulation and inherent phenomena are not well understood. However, recent observations by the Venus Climate Orbiter/Akatsuki have revealed a variety of atmospheric phenomena from the planetary-scale bow-shaped structure (Fukuhara et al., 2017) and streak-structure (Kashimura et al., 2019) to a front-like structure to small-scale vortices and waves of several hundred kilometers (Limaye et al., 2018). There have also been active attempts to reproduce these phenomena using Venusian atmospheric general circulation models and to understand the mechanisms involved. In particular, AFES-Venus (Sugimoto et al., 2014), which was developed based on AFES (Ohfuchi et al., 2004; Enomoto et al., 2008), the atmospheric general circulation model highly optimized for the Earth Simulator, has realized high-resolution simulations of the Venus atmosphere, and the planetary-scale streak-structure, thermal tides, gravity waves, meridional circulations, and other structures have been analyzed (e.g., Kashimura et al., 2019; Takagi et al., 2018; Sugimoto et al., in revision; Takagi et al., in prep.). However, AFES is a hydrostatic model, which is not suitable for phenomena whose horizontal scales are less than several tens of kilometers and cannot explicitly express convections in the cloud layer. The convections in the cloud layer are not only interesting in themselves but are also very important in the Venusian atmosphere because the neutral or low-stability layer resulting from convections is closely related to the formation of the planetary-scale bow-shaped structure and the streak structure. Though non-hydrostatic regional models have been used to study convective activities in the cloud layer and the resulting gravity waves (e.g., Baker et al., 1998; Imamura et al., 2014; Yamamoto 2014, Lefèvre et al., 2017), due to the limitation of the domain size, effects of the convective activities on large-scale phenomena have not been investigated.

Then, we have started developing a non-hydrostatic Venusian atmospheric general circulation model to realize a global simulation of the Venusian atmosphere that explicitly represents convective activities in the cloud layer. We utilize SCALE-GM (http://r-ccs-climate.riken.jp/scale/) for the dynamical core, which determines the coordinate system and calculates atmospheric motions. As a first step in the development, the various planetary and atmospheric constants are changed to Venusian values, and the solar heating and Newtonian cooling functions used in AFES-Venus (Tomasko et al., 1980; Crisp et al., 1986; Sugimoto et al., 2014) are imported to SCALE-GM. We then tried a simplified Venus simulation similar to AFES-Venus in various resolutions to explore the necessary numerical properties such as time-step intervals and strength of the numerical diffusion and the sponge layer. The planetary-scale streak-structure are represented in the same way as in AFES-Venus, in case of the resolution of a horizontal grid spacing of dx ~ 52 km and 120 layers (dz = 1 km) in vertical with a second-order Laplacian for the horizontal diffusion whose e-folding time for the minimum scale is 100 s. However, in SCALE-GM, a strong sponge layer that dumps not only the eddies but also the mean flow is necessary for computational stability. Therefore, the superrotation strength and the angular momentum budget of the atmosphere are influenced by the artificial sponge layer, and we need to seek better ways to keep computational stability. We are attempting to calculate in higher horizontal resolution with 240 vertical layers (dz = 500 m), and we will also present these results.