Bringing the Sun Down to Earth

The sun is often portrayed as a “laboratory for astrophysics” but it is a laboratory over which we have no control. Instead, solar physicists employ a wide range of theoretical, computational, and observational tools to study the magnetically driven activity on the sun. The challenge for these methods is that solar physics is extremely complex, making it difficult to isolate the key drivers of any one phenomenon. All of this is changing with the development of tailored, laboratory-scale plasma experiments here on earth. In Physical Review Letters, Eve Stenson and Paul Bellan at the California Institute of Technology in Pasadena describe just such an experiment [1], which is aimed at understanding a long-standing puzzle: why the solar corona—the halo of plasma that surrounds the sun’s atmosphere—is so hot. In their setup, they inject ionized gas into a horseshoe-shaped magnetic field, allowing them to simulate the dynamic behavior of plasma confined by the arched magnetic fields associated with sunspots. These fields are believed to play an important role in providing the energy to heat the corona, but the mechanism by which this occurs has been unclear. By demonstrating that they can create, within a plasma, magnetic fields with roughly the same shape and properties as those found in the corona, Stenson and Bellan offer an exciting opportunity to see solar magnetic forces at work in a controllable experiment. In the 1930s, scientists observing the solar corona found a mysterious spectroscopic line that was emitted from highly ionized iron. The finding revealed that the solar corona was a few million degrees kelvin, more than three hundred times hotter than the surface of the sun below, and flew in the face of what was expected from simple thermodynamics [2]; namely, that the corona, being furthest from the heat generated by nuclear fusion at the sun’s core, should be cooler than the underlying layers. For the next 80 years, solar physicists tried to understand what was making the corona so hot. It is now universally accepted that the reservoir of energy stored in the sun’s atmospheric magnetic field is what heats the localized plasma in the corona. In simplified terms, the field is generated in the solar interior as a result of large-scale rotational and convective motions of the charged plasma, which serve to produce a strong (100, 000 gauss) magnetic field some 200, 000 km below the solar surface [3]. At this depth, the field is in the form of a concentrated tube that encircles the sun, but as it makes its way to the surface, it can emerge as a pair, or group, of sunspots connected by arches of magnetic field that extend hundreds of thousands of kilometers into the solar atmosphere [Fig. 1(a)]. These arched fields vary in strength from 2–3 kilogauss (kG) at either foot, to about 10–100 gauss (G) at their apex. What is not known, and remains under considerable debate even now, is how the energy stored in the magnetic fields is converted into heating the corona [4]. Several authors have argued that the corona itself is not directly heated, but is a consequence of energy released in the much cooler chromosphere (∼ 50, 000 K)—a highly dynamic region of the solar atmosphere that separates the corona from the sun’s surface (photosphere)—that both heats the chromospheric plasma to coronal temperatures and then drives the heated plasma into the corona along the preexisting arched magnetic fields [5]. As an explanation, this idea has some advantages in that the energy dissipation is more efficient in the dense chromosphere and more naturally explains the density of the coronal material confined by the magnetic arches, which is much higher than that of the background corona. The drawback is we are left needing to understand the dissipation process, how the plasma is transported to the corona, and the impact on the preexisting magnetic field. Another puzzle is the shape of the coronal magnetic arches. The observed structures have a much more uniform (collimated) width across their entire length than would be expected from a simple dipolar field extending