A New Titan GCM and Stratospheric Superrotation

We will present numerical simulations of the atmospheric circulation predicted by a newly developed Ashima/MIT Titan general circulation model (GCM), which uses the MITgcm dynamical core and the physics package from TitanWRF. The main differences between the two Titan GCMs are: (1) The Ashima/MIT Titan GCM uses a finite-volume dynamical core while the TitanWRF GCM uses a finite-difference dynamical core, therefore conservation properties, particularly of momentum and energy, are different between these two GCMs. (2) Unlike the implicit numerical diffusion associated with the finite difference solver in TitanWRF, the Ashima/MIT Titan GCM has an explicitly adjustable Shapiro filter. (3) The Ashima/MIT Titan GCM includes a variety of advection schemes that may benefit transport-related problems.

Despite these differences, the Ashima/MIT Titan GCM produces results that are generally similar to those from TitanWRF. The similarities include the development and maintenance of stratospheric superrotation, and the overall seasonal and spatial variation of the wind and temperature field. However, there are also noticeable differences between the two GCMs. First, the strength of superrotation in the Ashima/MIT Titan GCM is generally weaker (about 15-20% weaker depending on the season). Second, the Ashima/MIT Titan GCM is able to produce a zonal wind minimum similar to (though at a different altitude to) the Huygens' probe measurement in the lower stratosphere, while TitanWRF does not capture this feature.

We speculate that the differences between the two GCMs may relate to the behavior of wave-mean interactions. For example, it seems that both models have little difficulty capturing interactions between the wind and fast traveling waves (e.g., the waves generated by barotropic instabilities travel at speed over 100m/s and provide equatorward flux of eastward momentum that drives the stratospheric superrotation), however, the strength of predicted instabilities may be different and thus affect the strength of superrotation. In the case of the wind minimum in the lower stratosphere, however, we speculate that this feature is a result of diurnal tides (or other slow-traveling waves), which travel slightly westward with phase speed of several meters per second and provide westward momentum to the zonal flow. If true, this would suggest that the Ashima/MIT Titan GCM captures interactions between the zonal wind and slowly-propagating waves slightly better than TitanWRF.

We will present detailed diagnostics of various instabilities and their effect on the mean flow in the two GCMs, focusing on explanation for the differences, which should shed further light on the challenges of reproducing Titan's observed wind field in GCMs.