Medium-resolution Optical and Near-Infrared Spectral Atlas of 16 2MASS-selected NIR-red Active Galactic Nuclei at $z \sim 0.3$

Dohyeong Kim,Myungshin Im,Gabriela Canalizo,Minjin Kim,Ji Hoon Kim,Jong-Hak Woo,Yoon Chan Taak,Jae-Woo Kim,Mariana Lazarova
DOI: https://doi.org/10.3847/1538-4365/aadfd5
2018-12-18
Abstract:We present medium-resolution spectra ($R \sim 2000$--4000) at 0.4--1.0\,$\mu$m and 0.7--2.5\,$\mu$m of 16 active galactic nuclei (AGNs) selected with red color in the near-infrared (NIR) of $J - K > 2.0$ mag at $z \sim 0.3$. We fit the H$\beta$, H$\alpha$, P$\beta$, and P$\alpha$ lines from these spectra to obtain their luminosities and line widths. We derive the $E(B-V)$ color excess values of the NIR-red AGNs using two methods, one based on the line-luminosity ratios and another based on the continuum slopes. The two $E(B-V)$ values agree with each other at rms dispersion $\sim$ 0.249. About half of the NIR-red AGNs have $g'-K < 5$ magnitude, and we find that these NIR red, but blue in optical-NIR AGNs, have $E(B-V) \sim 0$, suggesting that a significant fraction of the NIR color-selected red AGNs are unobscured or only mildly obscured. After correcting for the dust extinction, we estimate the black hole (BH) masses and the bolometric luminosities of the NIR-red AGNs using the Paschen lines to calculate their Eddington ratios ($\lambda_{\rm Edd}$). The median Eddington ratios of nine NIR-red AGNs ($\log (\lambda_{\rm Edd}) \simeq -0.654 \pm 0.176$) are only mildly higher than those of unobscured type 1 AGNs ($\log (\lambda_{\rm Edd}) \simeq -0.961 \pm 0.008$). Moreover, we find that the $M_{\rm BH}$--$\sigma_{\ast}$ relation for three NIR-red AGNs is consistent with that of unobscured type 1 AGNs at similar redshift. These results suggest that the NIR red color selection alone is not effective at picking up dusty, intermediate-stage AGNs.
Astrophysics of Galaxies
What problem does this paper attempt to address?