Astrophysical Fluid Dynamics

Visiting Honorary Professor:
Professor Chris Jones

PhD Students:
Xiaoya Zhan

*Also a member of Exeter Climate Systems (XCS).

Like many other research groups around the world, we run large scale numerical dynamo calculations to try and gain insight into the processes which are involved in the regeneration of the Earth's magnetic field.

A self-consistent dynamo must simultaneously satisfy the Navier-Stokes equation, the induction equation and the equation of heat advection. The resulting system is highly non-linear and very difficult to solve, requiring highly complicated programs to run for long periods of time. Verifying that such codes give the correct results can be highly problematic.

Different MHD groups over the world, including the one at Exeter have verified the results of their dynamo codes by comparing their results against the Numerical Dynamo Benchmark (for details, see Physics of the Earth and Planetary Interiors, v128, pp.25-34, 2001), initiated by the geodynamo group at the Institut für Geophysik der Universität Göttingen, who have found a dynamo which is independent of time, except for a constant drift in longitude.

The following two plots are the field and velocity for the (Case 1) dynamo as calculated by the Exeter MHD group code.

Contours of radial field (r = r_o)

Contours of radial velocity (r = r_i + 0.64d)

Recent studies by Sreenivasan and Jones have revealed the role played by inertia in the current generation of spherical dynamos. The following pictures give the contour plots of the radial velocity (at r=0.8r_o) and magnetic field (at r=r_o) as the Prandtl (Pr) and magnetic Prandtl (Pm) numbers are varied together from 5 to 0.2. The onset of inertial dominance is apparent from the breakdown of well-defined columnar structures and the departure from a dipolar configuration.

Contours of radial velocity at Pr=Pm=5,1 & 0.2 (top row) and the corresponding radial magnetic fields (bottom row). As Pr=Pm is lowered, inertial effects set in, distorting the columnar structures and erasing the dipolar field.

We are also interested in how the behaviour of the geodynamo is influenced by the mantle, in particular through thermal core-mantle coupling.

It may be that the most significant cause of these observations is lateral variations in lower-most mantle composition, although we assume that low (high) S-wave velocities indicate relatively hot (cold) regions. We then observe how this thermal boundary condition then effects the behaviour of our models.