Physics Colloquium

Long ago, magnetic fields in astrophysics used to be something
for specialists. This has drastically changed over the past few
decades. Nowadays, many scientists invoke magnetic fields in simulations
of astrophysical flows. In fact, the concept of purely hydrodynamic
turbulence without magnetic fields seems to be an idealization that
hardly exists anywhere. We now understand that this is because of the
universality of dynamo action in many types of flows. Dynamos convert
kinetic energy into magnetic; they are self-excited and work without
wires, but without short-circuiting themselves in spite of their high
conductivity everywhere.

In my talk, I will start with a historical perspective, going back to
the days when the existence and origin of magnetic fields was still
very obscure. We knew about the Earth's magnetic field since the 1600s,
and astronomical observations have revealed magnetic fields in sunspots
and eventually in other stars and galaxies during the last century. To
understand their origin, people had to struggle with Cowling's anti-dynamo
theorem that magnetic fields cannot be generated from kinetic energy
in a simple axisymmetric geometry. Gradually, it became clear that in
three-dimensional settings, self-excited dynamos do actually work.

Meanwhile, with the emergence of three-dimensional simulations, where
the plasma motions tend to be turbulent, dynamo action appears to
be a natural by-product. Dynamos have also been realized in the lab
in various configurations. But some basic questions in astrophysical
applications are still troubling us: why exactly is the Sun's magnetic
field exhibiting equatorward migration and why do the most realistic
simulations not yet reproduce the large-scale magnetic fields observed
in spiral galaxies by the present time. I will finish with applications
to the early Universe, where decaying magnetic turbulence governs the
entire radiation-dominated era and many of the relationships can be
understood from dimensional arguments.

Join via Zoom:
https://caltech.zoom.us/j/81866929019
Meeting ID: 818 6692 9019

The colloquium is held in Feynman Lecture Hall, 201 E. Bridge.