Benchmarking Dust Emission Models in M101

TL;DR

This study compares four physical dust models and two modified blackbody models by fitting them to multi-wavelength data of M101, revealing significant variations in dust mass estimates and the importance of normalization for consistency.

Contribution

It provides a systematic comparison of dust models using identical data and fitting techniques, highlighting differences and potential ways to reconcile them.

Findings

Dust mass estimates vary by up to a factor of 3 across models.

Physical models show consistent spatial variations in aromatic grain abundance.

Renormalization aligns dust mass estimates with physical and metallicity constraints.

Abstract

We present a comparative study of four physical dust models and two single-temperature modified blackbody models by fitting them to the resolved WISE, Spitzer, and Herschel photometry of M101 (NGC 5457). Using identical data and a grid-based fitting technique, we compare the resulting dust and radiation field properties derived from the models. We find that the dust mass yielded by the different models can vary by up to factor of 3 (factor of 1.4 between physical models only), although the fits have similar quality. Despite differences in their definition of the carriers of the mid-IR aromatic features, all physical models show the same spatial variations for the abundance of that grain population. Using the well determined metallicity gradient in M101 and resolved gas maps, we calculate an approximate upper limit on the dust mass as a function of radius. All physical dust models are found to exceed this maximum estimate over some range of galactocentric radii. We show that renormalizing the models to match the same Milky Way high latitude cirrus spectrum and abundance constraints can reduce the dust mass differences between models and bring the total dust mass below the maximum estimate at all radii.