The structures observed in the solar atmosphere, such as loops and prominences, are the result of the magnetic field of the Sun exhibitng itself on (relatively) small scales.
The dynamics of the plasma in the atmosphere are strongly influenced by the magnetic field, which exerts pressure on the plasma much larger then the thermodynamic pressure. (This is known as a low-beta plasma). The exciting events we see in the Sun's atmosphere, such as Flares's [LINK], CME's, and prominence eruptions, are a direct consequence of the evolution of the magnetic field, be it by instabilities, or reconnection of field lines, or other phenomena.
It has long been conjectured (originally by the Physicist Eugene Parker in the 60's) that the origins of the magnetic field in the solar atmosphere actually lie in the interior of the Sun. The Sun magnetic field exhibits behaviour that suggests that it is continaully beging generated in a cyclic nature (the solar dynamo). The most likely place for the generation of magentic field is in the interior of the Sun, at the boundary between the convection zone and the inner radiative zone. It is believed that large scale magnetic field can 'break up' at the base of the convection zone and rise up to the surface. The unstable layers of the convection zone allow the buoyant rise of 'magnetic flux tubes' to the surface. The manifestation of these tubes at the surface are sunspots[LINK], which are dark, cool regions where energy transport is inhibited by magnetic field.
Although this model of sunspot formation is well accepted, it is unclear as to how the magnetic field reaches the outer atmosphere, or corona of the Sun. As direct observations of the magnetic field at and below the surface are difficult, the use of numerical simulations can gives us an insight as to the mechanims which drive field into the corona.
The majority of my work invloves diagnosing the mechanism which forces field through the stable photpshere and chromosphere and into the corona. This is done by using large scales MHD simulation of emerging flux tubes. A likely candidate for emergence is the Magnetic Buoyancy instability. This can be thought of as similar to the Rayleight Taylor instability, which is a result of a denser fluid resting on a lighter one causing an unstable boundary. However, the strength of the magnetic field in the tubes, and more inportantly, the twist of ield lines may change the mechansim. By varying the free parameters of the problem we can investigate these mechanisms.
Results form 2D simulations can be found in my accepted A&A paper paper.pdf