A New Model for Jovian Weather Layer Dynamics

Juno has revealed long-lived clusters of circumpolar cyclones (CPCs) in Jupiter's polar regions. These observations were not predicted, challenging our basic understanding of fluid dynamics, and evade simulation by global numerical models, preventing us from testing theories for their existence. 

Global, three-dimensional weather layer General Circulation Models (GCMs) of Jupiter's atmosphere are useful tools for bridging theory and observations. However, they fail to produce CPCs. Two potential causes for this are: (1) low spatial resolution, which precludes GCMs from representing thin, anti-cyclonic 'shields', that are believed to be important for CPC formation; and (2) poor representation of interior-weather layer interaction, which may be important for generating the CPCs. The aim of my research is to design a new Jupiter GCM that is capable of simulating small-scale meteorological features in the polar regions. 

The new model has been constructed using FV3, an open-source finite-volume 'dynamical core' (fluid dynamics solver) developed by the NOAA Geophysical Fluid Dynamics Laboratory (GFDL). FV3 solves the fully-compressible Euler equations for a shallow atmosphere (appropriate for the weather layer) using a cubed-sphere horizontal grid and a Lagrangian, flow-following coordinate in the vertical. FV3 offers several advantages for the simulation of Jovian (and giant planet) atmospheric dynamics. FV3 can be run in a standard 'global-only' configuration, but also has functionality to embed high-resolution regional grids ('nests') within a global GCM. This capability will enable its use for conducting high-fidelity simulation of the Jupiter’s polar regions, overcoming issue (1) above. In addition, FV3 was designed for operational weather forecasting, and is thus designed to be extremely computationally efficient while preserving solution accuracy. Finally, the importance of vorticity dynamics is emphasised in FV3's design, and its advection scheme has the property that the shallow-water potential vorticity (PV) and the helicity are advected as scalars. This allows FV3 to excel in the simulation of cyclones on Earth, and is a major advantage for this research project.

Here, I report results from global-only benchmarking simulations run using this model. Specifically, FV3 has been adapted to incorporate the 'physics packages' (e.g., convective adjustment, radiative transfer) implemented by Schneider and Liu (SL; 2009) and Young et al. (2019) to study Jupiter. Diagnostics obtained from FV3 using these configurations, including: the circulation morphology; the zonal-mean structure of the temperature and zonal wind; eddy statistics; and energy spectra; will be compared against the original results reported in these studies. The aim of this comparison is to quantify the impact of dynamical-core choice on the simulated atmospheric circulation (SL use the GFDL pseudospectral dynamical core, and Young et al. use the MITgcm on an Arakawa-C grid). Provisional results using a nested-grid set-up (with high-resolution grids included at the poles) may also be presented, dependent on the progress of model development between now (April 2026) and EPSC! Future development of the model will focus on adding a stochastic element to the lower-boundary interior heat flux, as an initial step towards overcoming issue (2) above.

References: 1. Schneider and Liu, 2009. J. Atmos. Sci., 66, 3, pp. 579-601. 2. Young, Read, and Wang, 2019. Icarus, 326, pp. 225-252.