A sensitivity analysis of the modeling of Polycyclic Aromatic Hydrocarbon emission in galaxies

We have conducted a sensitivity analysis on the mid-infrared spectral decomposition of galaxies and the modeling of the PAH emission spectrum with the NASA Ames PAH Infrared Spectroscopic Database (PAHdb) to assess the variance on the average galaxy PAH population properties under a grid of different modeling parameters. We find that the SL and SL+LL Spitzer-IRS decomposition with PAHFIT provides consistent modeling and recovery of the 5-15 $\mu$m PAH emission spectrum. For PAHdb modeling, application of a redshift to the calculated spectra to account for anharmonic effects introduces a $15\%$-$20\%$ variance on the derived parameters, while its absence improves the fits by $\sim13\%$. The 4.00-$\alpha$ release of PAHdb achieves the complete modeling of the 6-15 $\mu$m PAH spectrum, including the full 6.2 $\mu$m band, improving the average fitting uncertainty by a factor of 2. The optimal PAHdb modeling configuration requires selection of pure PAHs without applying a redshift to the bands. Although quantitatively the PAHdb-derived parameters change under different modeling configurations or database versions, their variation follows a linear scaling, with previously reported trends remaining qualitatively valid. PAHdb modeling of JWST observations, and JWST observations smoothed and resampled to the Spitzer-IRS resolution and dispersion have consistent PAHdb derived parameters. Decomposition with different codes, such as PAHFIT and CAFE, produce PAH emission spectra with noticeable variation in the 11-15~$\mu$m region, driving a $\sim7\%$ difference in the neutral PAH fraction under PAHdb modeling. A new library of galaxy PAH emission templates is delivered to be utilized in galaxy SED modeling.
Comment: 25 pages, 17 figures, 6 tables. Accepted for publication in ApJ