Modeling the unresolved NIR-MIR SEDs of local ($z<0.1$) QSOs

To study the nuclear ($\lesssim1\,$kpc) dust of nearby ($z<0.1$) type 1 Quasi Stellar Objects (QSOs) we obtained new near-infrared (NIR) high angular resolution ($\sim0.3$ arcsec) photometry in the H and Ks bands, for 13 QSOs with available mid-infrared (MIR) high angular resolution spectroscopy ($\sim7.5-13.5\,\mu$m). We find that in most QSOs the NIR emission is unresolved. We subtract the contribution from the accretion disk, which decreases from NIR ($\sim35\%$) to MIR ($\sim2.4\%$). We also estimate these percentages assuming a bluer accretion disk and find that the contibution in the MIR is nearly seven time larger. We find that the majority of objects ($64\%$, 9/13) are better fitted by the Disk+Wind H17 model \citep[][]{Hoenig17}, while others can be fitted by the Smooth F06 \citep[$14\%$, 2/13,][]{Fritz06}, Clumpy N08 \citep[$7\%$, 1/13,][]{Nenkova08a,Nenkova08b}, Clumpy H10 \citep[$7\%$, 1/13,][]{Hoenig10b}, and Two-Phase media S16 \citep[$7\%$, 1/13,][]{Stalev16} models. However, if we assume the bluer accretion disk, the models fit only 2/13 objects. We measured two NIR to MIR spectral indexes, $\alpha_{NIR-MIR(1.6,8.7\,\mu\text{m})}$ and $\alpha_{NIR-MIR(2.2,8.7\,\mu\text{m})}$, and two MIR spectral indexes, $\alpha_{MIR(7.8, 9.8\,\mu\text{m})}$ and $\alpha_{MIR(9.8, 11.7\,\mu\text{m})}$, from models and observations. From observations, we find that the NIR to MIR spectral indexes are $\sim-1.1$ and the MIR spectral indexes are $\sim-0.3$. Comparing the synthetic and observed values, we find that none of the models simultaneously match the measured NIR to MIR and $7.8-9.8\,\mu$m slopes. However, we note that measuring the $\alpha_{MIR(7.8, 9.8\,\mu\text{m})}$ on the starburst-subtracted {\it Spitzer}/IRS spectrum, gives values of the slopes ($\sim-2$) that are similar to the synthetic values obtained from the models.
Comment: Accepted in ApJ