Martian atmosphere breeding: elucidating atmospheric instabilities

Breeding is a simple technique for identifying and explaining the physical origin of instabilities in a dynamical system. In chaotic systems, such as planetary atmospheres, two initial states that are originally close together will drift apart in the presence of dynamical instabilities. The breeding method introduces random perturbations to a base state, and evolves both states forward in time nonlinearly. At regular intervals, the perturbations are rescaled so that they maintain their original global amplitude. The bred vectors (similar to Lyapunov vectors) give the dominant patterns of instability.

Here, we apply the breeding method to the NASA/NOAA Mars Global Circulation Model. This technique is extended through the application of mean and perturbation energy equations, which are applied exactly to the base state and bred vector. This allows the specific diagnosis of instability types, clarifies conversions between kinetic and potential energy, and denotes favored locations for error growth, bolstering physical insight into the atmospheric dynamics of Mars. Although the current model only allows for constant dust profiles, with dust lifting parameterizations the bred vectors should provide a tool to assess the instability processes in dust storms. Breeding experiments can also assist in the optimal tuning of an ensemble data assimilation system, such as the selection of inflation and localization schemes.