2014MNRAS.440.2726R

We use anisotropic complex wavelet transforms to make a multiscale analysis of the distribution of fluctuations of dust emission at 250 µm from the Herschel infrared Galactic Plane Survey of the Herschel Space Observatory. By reproducing the Fourier power spectrum with complex wavelet transforms we show that different power distributions at each scale can have a significant effect on the measured power law. Moreover, with an iterative algorithm we separate the Gaussian and non-Gaussian part of wavelet coefficient distributions in each scale and azimuthal direction. The reconstructed map of non-Gaussian fluctuations is well correlated to small-scale structures of the 13CO emission map as well as its power spectrum which has a power-law index of -2.3. The Gaussian part of the map shows more diffuse structures with a steeper power law of -3.1. Non-Gaussian coefficients are almost undetectable at scales greater than ∼ 0.15/arcmin ( ∼ 4-26 pc). This characteristic scale could be associated with the size of the larger molecular cloud in the region. We also demonstrate that exponentiated Gaussian random field cannot reproduce the flatter power law of the non-Gaussian component observed in the thermal dust emission, which seems to be a property of interstellar clouds.