2018MNRAS.481..509E

The multifractal geometry remains an underexploited approach to describe and quantify the large-scale structure of interstellar clouds. In this paper, the typical tools of multifractal analysis are applied to Herschel far-infrared (70-500µm) dust continuum maps, which represent an ideal case of study. This dust component is a relatively optically thin tracer at these wavelengths and the size in pixel of the maps is well suitable for this statistical analysis. We investigate the so-called multifractal spectrum and generalized fractal dimension of six Hi-GAL regions in the third Galactic quadrant. We measure systematic variations of the spectrum at increasing wavelength, which generally correspond to an increasing complexity of the image, and we observe peculiar behaviours of the investigated fields, strictly related to the presence of high-emission regions, which in turn are connected to star formation activity. The same analysis is applied to synthetic column density maps, generated from numerical turbulent molecular cloud models and from fractal Brownian motion (fBm), allowing for the confrontation of the observations with models with well controlled physical parameters. This comparison shows that, in terms of multifractal descriptors of the structure, fBm images exhibit a very different, quite unrealistic behaviour when compared with Hi-GAL observations, whereas the numerical simulations appear in some cases (depending on the specific model) more similar to the observations. Finally, the link between mono-fractal parameters (commonly found in the literature) and multifractal indicators is investigated: the former appear to be only partially connected with the latter, with the multifractal analysis offering a more extended and complete characterization of the cloud structure..