Stochastic Modelling of Jupiter's Magnetosphere

With the Juno spacecraft in orbit about Jupiter, an extensive data set of magnetic field measurements now exists for the Jovian field. This has improved the coverage of the planet both in space and time by reaching greater latitudes as well as providing a continuous timeseries over a nine year period (2016-present). Such improved sampling of Jupiter has allowed for far more in-depth studies of its magnetic field, producing more complex field models to those found from earlier missions and the first claims of secular variation at a body other than Earth [1,2].

Although these models have revealed interesting new features of the Jovian magnetic field, they all suffer from large misfits in comparison with intrinsic satellite uncertainties: typical rms misfits of 500-1000 nT give weighted rms misfits of the order of ~10 (when published error budgets are adopted) [3]. Statistically, this underfitting strongly suggests that there are currently unmodelled processes that need to be considered.

Given the size of the Jovian magnetosphere, we believe this external field is responsible for the observed mismatch. The majority of field models take this external field as a low order l ~ 1, 2 spherical harmonic expansion, corresponding physically to Jupiter’s magnetodisc (a ring current, tilted with respect to the equatorial plane). This (near-)dipolar magnetosphere describes a simplified system in which small scale effects are neglected. A deterministic, higher order spherical harmonic expansion would provide a more accurate representation of the external field. However, there is still insufficient data to carry out this process to the desired level of accuracy.

The remaining approach to consider is a stochastic description of the system [4,5]. To do so, we model a distribution of randomly oriented currents around a planet, in a simplified spherical geometry. We find that measurements made within this region have non-zero, large scale correlations. This indicates that the presence of stochastic currents in the magnetosphere acts to relate measurements that would otherwise be considered independent.

We apply this method to the Juno data set for the case of a spherical magnetosphere geometry as described above. By taking into account these newfound correlations, we “pre-whiten” the Juno measurements (removing the effects of these stochastic currents in the magnetosphere). The resulting dataset is composed of entries that are more independent than the raw data itself. Through regularised inversion of this “pre-whitened” data, we find new models of Jupiter’s internally sourced magnetic field. For appropriately chosen current variance, these models fit the data well and account for the misfit seen in previous cases.

This is an encouraging result that shows the importance of a more full description of Jupiter’s magnetosphere. There is, however, more to consider in regards to further magnetosphere geometries and the interplay between these more complex magnetospheres and other contemporary models of Jupiter’s magentic field.

 

Acknowledgments:

The Planetary Data System (PDS) has proved invaluable in carrying out this work, to source data from both Juno and the earlier missions mentioned used throughout.

 

References:

[1] K. Moore et al. (2019) Nature, 561, 76-78

[2] J. Bloxham et al. (2022) JGR: Planets, 127

[3] S. Sharan et al. (2022) Geophysical Research Letters , 49

[4] R. Parker (1988) JGR: Solid Earth, 93, 3105

[5] A. Jackson (1990) GJI, 103, 657