Properties of spicules inferred from IRIS observations via approximate Bayesian computation

Solar spicules remain a rather enigmatic phenomenon which is highly important for understanding of the physics of the outer solar atmosphere. Given the enormous number of spicules emanating at any time from the solar surface, it is difficult to make quantitative conclusions about the spicule properties. Most of the recent studies tend to investigate individual spicules that crop out of the forest of other overlapping spicules. However, such an approach comes with the risk that the studied spicules do not reveal the typical spicule properties due to the applied selection criteria. In our approach we consider the large amount of visually overlapping and constantly changing spicules forming the spicular forest as an observational advantage and we study them using advanced statistical techniques. We construct a generative radiative-transfer model of spicules and use it for inference of the spicule parameters via approximate Bayesian computation. We analyze the MgII k line profiles observed by IRIS on Feb 21, 2016 in a solar coronal hole and we infer the approximate posterior distribution of a number of spicule parameters; namely the inclinations, velocities, opacities, geometrical widths, and lengths. In addition, we derive the distributions of the absorption profile width of the MgII k line and of the spatial density of spicules on the solar surface.