The structure and dynamics of the outer solar atmosphere are reviewed with emphasis on the role played by the magnetic field. Contemporary observations that focus on high resolution imaging over a range of temperatures, as well as UV, EUV and hard X-
ray spectroscopy, demonstrate the presence of a vast range of temporal and spatial scales, mass motions, and particle energies present. By focussing on recent developments in the chromosphere, corona and solar wind, it is shown that small scale processes, in particular magnetic reconnection, play a central role in determining the large-scale structure and properties of all regions. This coupling of scales is central to understanding the atmosphere, yet poses formidable challenges for theoretical models.
A major unexplained feature of the solar atmosphere is the accumulation of magnetic shear, in the form of filament channels, at photospheric polarity inversion lines (PILs). In addition to free energy, this shear also represents magnetic helicity, wh
ich is conserved under reconnection. In this paper, we address the problem of filament channel formation and show how they acquire their shear and magnetic helicity. The results of 3D simulations using the Adaptively Refined Magnetohydrodynamics Solver (ARMS) are presented that support the model of filament channel formation by magnetic helicity condensation developed by citet{Antiochos13}. We consider the supergranular twisting of a quasi-potential flux system that is bounded by a PIL and contains a coronal hole (CH). The magnetic helicity injected by the small-scale photospheric motions is shown to inverse-cascade up to the largest allowable scales that define the closed flux system: the PIL and the CH. This process produces field lines that are both sheared and smooth, and are sheared in opposite senses at the PIL and the CH. The accumulated helicity and shear flux are shown to be in excellent quantitative agreement with the helicity-condensation model. We present a detailed analysis of the simulations, including comparisons of our analytical and numerical results, and discuss their implications for observations.
