Content uploaded by L. Spinoglio
Author content
All content in this area was uploaded by L. Spinoglio on Sep 11, 2021
Content may be subject to copyright.
arXiv:1708.08563v1 [astro-ph.GA] 29 Aug 2017
Emission Line Properties of Seyfert Galaxies
In the 12 Micron Sample
Matthew A. Malkan1, Lisbeth D. Jensen1, David R. Rodriguez1, Luigi Spinoglio2, and
Brian Rush1,3
malkan@astro.ucla.edu
ABSTRACT
We present optical and ultraviolet spectroscopic measurements of the emission
lines of 81 Seyfert 1 and 104 Seyfert 2 galaxies which comprise nearly all of
the IRAS 12µm AGN sample. We have analyzed the emission-line luminosity
functions, reddening, and other diagnostics. For example, the narrow-line regions
(NLR) of Seyfert 1 and 2 galaxies do not significantly differ from each other in
most of these diagnostics. Combining the Hα/Hβratio with a new reddening
indicator–the [SII]6720/[OII]3727 ratio, we find the average E(B−V) is 0.49 ±
0.35 for Seyfert 1’s and 0.52 ±0.26 for Seyfert 2’s. The NLR of Sy 1 galaxies has
only insignificantly higher ionization level than in the Sy 2’s. For the broad-line
region (BLR), we find that the C IV equivalent width correlates more strongly
with [O III]/Hβthan with UV luminosity. Our bright sample of local active
galaxies includes 22 Seyfert nuclei with extremely weak broad wings in Hα, known
as Seyfert 1.9’s and 1.8’s, depending on whether or not broad Hβwings are
detected. Aside from these weak broad lines, our low-luminosity Seyferts are
more similar to the Sy2’s than to the Sy 1’s. In a a BPT diagram we find
that Sy 1.8’s and Sy 1.9’s overlap the region occupied by the Sy 2 galaxies. We
compare our results on optical emission lines with those obtained by previous
investigators using AGN subsamples from the Sloan Digital Sky Survey. The
luminosity functions of forbidden emission lines [OII]λ3727˚
A, [OIII]λ5007˚
A, and
[SII]λ6720˚
A in Seyfert 1’s and 2’s are indistinguishable. They all show strong
downward curvature. Unlike the LF’s of Seyfert galaxies measured by the Sloan
Digital Sky Survey, ours are nearly flat at low luminosities. The larger number of
faint Sloan “AGN” is attributable to their inclusion of weakly emitting LINERs
1Physics and Astronomy Department, University of California, Los Angeles, CA 90095
2Istituto di Fisica dello Spazio Interplanetario, INAF, Via Fosso del Cavaliere 100, I-00133 Roma, Italy
3Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109
– 2 –
and H II+AGN “composite” nuclei, which do not meet our spectral classification
criteria for Seyferts.
In an Appendix, we have investigated which emission line luminosities can
provide the most reliable measures of the total non-stellar luminosity, esti-
mated from our extensive multi-wavelength database. The hard X-ray or
near-ultraviolet continuum luminosity can be crudely predicted from either the
[O III]λ5007˚
A luminosity, or the combination of [O III]+Hβ, or [N II]+Hαlines,
with a scatter of ±4 times for the Sy 1’s and ±10 times for the Sy 2’s. Although
these uncertainties are large, the latter two hybrid (NLR+BLR) indicators have
the advantage of predicting the same HX luminosity independent of Seyfert type.
Subject headings: galaxies: luminosity function — galaxies: Seyfert — quasars:
emission lines
1. Introduction
Many previous studies have measured emission line ratios in samples of Seyfert nuclei,
but the existence of strong selection effects (such as searches based on host galaxy proper-
ties, UV-excess or X-ray flux) raise questions about whether these results would hold for a
representative sample of the full unbiased Seyfert population. A common limitation of most
surveys is lack of complete data for the less luminous Seyfert nuclei. Another limitation of
surveys at optical, ultra-violet, or soft X-ray wavelengths is their inability to find more red-
dened AGN. These limitations can be overcome with a nonstellar-flux-limited all-Sky survey
at long wavelengths, which includes most of the bright Seyfert galaxies in the local universe.
We pursue this approach in this paper. One of our motivations is that a better observational
understanding of the nearest and best observed Seyfert galaxies will help us to understand
the population of AGN at high redshifts. High-redshift studies are now seeking to measure
the cosmic evolution of AGN, but with much less complete data than we have locally. By
emphasizing quality of the data over raw quantity, we hope to use extensive observations of
local AGN to help calibrate the emission-line diagnostics in high-redshift samples.
1.1. The Extended 12µm and CfA AGN Samples
The extended 12µm galaxy catalogue of Spinoglio & Malkan (1989) and Rush et al.
(1993), is a 12µm flux-limited sample containing 893 galaxies selected from the IRAS Faint
Source Catalog, Version 2. The galaxies in this catalogue have galactic latitude of |b| ≥ 25o
– 3 –
to decrease the extinction and avoid stellar contamination from the plane of our Galaxy.
The 12µm galaxy sample contains 9 percent Sy 1’s and quasars, and 11 percent Sy 2’s.
These percentages are of course far higher than the percentages of Seyferts among ordinary
optically-selected galaxies such as in the Sloan Digital Sky Survey (SDSS). This is because,
by design, the 12µm galaxies were selected at a wavelength where the continuum emission
from warm dust in the Seyfert nucleus is especially bright relative to the normal emission
from the underlying host galaxy. The 12µm flux is a constant fraction of ∼1/5 of the
bolometric flux in Seyfert 1 and 2 galaxies, three times more than in normal spiral galaxies
(Spinoglio & Malkan 1989). The 12µm Seyferts are excellent representatives of the entire
class, since they span nearly 6 orders of magnitude in luminosity. With a log(N)−log(S)
test Spinoglio & Malkan (1989) showed that the sample is complete down to 0.30 Jy, and
has a level of incompleteness of ∼40% at 0.22 Jy, the chosen flux limit.
The extended 12µm sample includes the brightest nearby Seyferts in the local universe.
It has been subjected to extensive observational follow-up across the entire electromagnetic
spectrum, and thus has the most complete multi-wavelength dataset available for any AGN
sample. The redshifts range from z=−0.0001 to z= +0.1884, with the majority at z≤
0.05. The redshifts are obtained from the the NASA/IPAC Extragalactic Database (NED)1.
Throughout we adopt H0= 72 km/s/Mpc when computing distances to the Seyferts. 2.
To further control possible selection bias, we supplement the data with Seyfert galaxies
from the Center for Astrophysics (CfA) galaxy sample. The CfA galaxies are a host-galaxy
flux-limited, spectroscopically selected sample defined by Huchra et al. (1983). They come
from 2399 galaxies with m≤14.5, with cuts in band δto avoid contaminations from the
galactic plane (Huchra & Burg 1992). Thuan & Sauvage (1992) provide IRAS fluxes for
1544 galaxies in the CfA sample that are detected in the IRAS Faint Source Catalog. The
overlap between the CfA and the 12µm sample is 47 Seyfert galaxies (∼25%), as described
in Rush et al. (1993).
Our data are compiled from multiple literature sources, the CfA sample, and our own
previously unpublished data. The full sample, containing 185 Seyfert galaxies, 81 Seyfert
1’s and 104 Seyfert 2 galaxies, is listed in Tables 1 and 2. It lists spectral classifications,
redshift, and which galaxies are common to both the 12µm and CfA samples.
For simplicity we have adopted the spectral type classification given in NED’s “Basic
1The NASA/IPAC Extragalactic Database (NED) is operated by the Jet Propulsion Laboratory, Cali-
fornia Institute of Technology, under contract with the National Aeronautics and Space Administration.
2The one exception is NGC 3031/M81, for which we use the measured Cepheid distance of 3.63 Mpc
(Freedman et al. 1994).
– 4 –
Data”. These are based on spectra obtained from a wide variety of publications. The ability
to discern faint Balmer-line wings–required for an Sy 1 classification–depends on the quality
the spectrum. The classification for weak Sy 1’s and also composite Sy 1+HII galaxies is
therefore sometimes ambiguous. But we only have one or two active galaxies for which our
data indicate a different classification from NED. These special cases are mentioned below.
1.2. Spectroscopic Measurements
We have averaged over multiple measurements and conservatively assign an uncertainty
of 30% in the line fluxes, although ratios of nearby lines are usually more accurate. Our
combined optical and ultraviolet emission lines and their corresponding rest wavelengths can
be found in Table 3. Optical line fluxes along with literature references are in Table 4, and
the UV data are summarized in Table 5, also with literature references.
We have supplemented our observational results with data from Sloan Digital Sky Sur-
vey Data Release 7 from “The MPA-JHU DR7 release of spectrum measurements”3. This
data base contain 927,552 AGN galaxies. We excludeed two-thirds of those which lacked
S/N emission-line ratios of at least >10. We have also used the SDSS DR6 SkyServer
Explore Tool4to provide line fluxes for 5 Sy1’s and 11 Sy 2’s, including four objects (IRAS
13354+3924, IRAS 16146+3549, NGC 833, and UGC 6100) for which no prior data ex-
isted. We calculate the line-flux ratios of [O III]/Hβ, [O III]/[O II], [O II]/[O I], [N II]/Hα,
[O II]/[S II], [O II]/[N II], [O III]/[S II], and [N II]/[S II], and find that the average difference
between our sample and that of SDSS for the ratios is −0.03 ±0.25, consistent with no
systematic difference.
This paper is organized as follows. In §2 we describe our emission line luminosity
functions. In §3 we use emission lines to diagnose properties in the narrow line region, while
in §4 we consider the broad line region. We summarize our results in §5. In Appendix A we
investigate, which, if any, emission line luminosities can provide the most reliable estimates
of the total non-stellar luminosity,
3Obtained from http://wwwmpa.mpa-garching.mpg.de/SDSS/DR7/. Raw data from
http://wwwmpa.mpa-garching.mpg.de/SDSS/DR7/raw data.html.
4The SDSS DR6 SkyServer Explore Tool can be found at http://cas.sdss.org/astro/en/tools/explore/.
– 5 –
2. Line Luminosity Functions
We construct emission-line luminosity functions (LFs) for each Sy 1 and Sy 2 galaxy
by taking Vmax values from the 12µm flux sample and binning them according to their
individual emission-line luminosities. Since our sample is defined and limited by the 12µm
flux, we apply correction factors where necessary to account for incompleteness previously
determined by Rush et al. (1993).
Emission-line LFs are derived for [O I], [O II], [O III], [N II], [S II], Hα, and Hβlines. A
double power-law in Logarithmic space is fitted to the luminosity functions:
Φ(L) = Φ⋆
L
L⋆α
+L
L⋆α+β(1)
where L⋆is the emission-line luminosities of the characteristic break in the LF (in erg/s),
and Φ⋆is twice the number density at that break (in Mpc−3). The points always require a
bend in the LF, i.e. a steeper slope at high luminosities. However, the strength of this bend
is not well determined due to the small of high- and low-luminosity galaxies in our sample.
We therefore assumed a low-luminosity slope of α=−0.1 and a slope steepening break of
β= +1.5. These broken power laws, based on the shape of the 12µm LF, match all the
emission line LFs adequately.
The best-fit parameters of the LFs are given in Table 6 and displayed in Figure 1. We
plot the LF for the narrow lines [O II] and [O III], and the Hαand Hβlines (with the broad
and narrow components combined) in Figure 1.
The Balmer-line luminosity function of the Sy 1s extends to higher luminosities than
that of the Sy 2s. Because the Sy 1 permitted lines are brightened by the contribution
from their broad line region (BLR), they are more numerous at the high line luminosities.
However, for the [O III] and [O II] LFs, there is only a small difference between the Sy 1 and
Sy 2 galaxies. Thus at a given narrow line luminosity, the two types of Seyferts have similar
space densities. There is also little difference in the 12µm Sy 1 and Sy 2 continuum LFs
(Rush et al. 1993; Toba et al. 2014), at luminosities below that of quasars. So if the narrow
line luminosity is emitted approximately isotropically, the similarity of these Sy1 and Sy2
space densities suggests that the 12µm continuum emission is also relatively isotropic.
Figure 1a, 1c, and 1d compares the Sy 1 and Sy 2 emission-line Luminosity Functions for
our 12µm selected sample with the optically selected AGN sample from the SDSS (Hao et al.
2005; Simpson 2005). The optical and IR selection methods show good agreement in the de-
rived LFs around the ‘knee’, i.e. the line-luminosities around 1039−41 erg/s, for both Seyfert
types. However, compared with the 12µm selection, the SDSS finds relatively more low-
– 6 –
luminosity AGN and relatively less high-luminosity AGN. Thus the SDSS LFs are steeper
at low luminosities than our 12µm sample. The SDSS LFs thus show much weaker breaks
to high luminosities. We attribute this difference to the inclusion in SDSS of more “compos-
ite” Sy 2’s. These low-luminosity AGN have substantial line emission contributed by star
formation, which tends to prevent them for being classified in NED as Seyfert Galaxies. On
the other hand, the 12µm selection is particularly efficient for finding luminous AGN, which
tend to be at high redshifts. Figure 1 indicates that it is the optical selection of SDSS that
becomes significantly incomplete at high luminosities (emission-lines &1041). We note that
the [O III] LF found by Bongiorno et al. (2010) from zCOSMOS, at slightly higher redshifts
(0.15 < z < 0.3) agrees closely with ours, and is similarly flatter than that of Hao et al.
(2005).
We have also compared our Seyfert galaxy LF’s to the those of the local normal galaxies
in Hαand [O III] (not shown), using data from Gallego et al. (1995, 1996). For most line
luminosities, the space density of the galaxies without Seyfert nuclei exceeds that of the
Seyfert by up to two orders of magnitude. However, the Hαand [O III] LF’s of non-AGN
cut off exponentially above Lline >1041 erg/sec. The result is that with Hαor [O III]
luminosities of 1041 erg/sec or above, most of the galaxies are Seyferts.
3. Seyfert 1 and 2 Narrow Line comparison
The main observational difference between the two types of Seyfert galaxies is the pres-
ence or absence of broad permitted lines with widths of 103km/s or higher. The Seyfert
1 galaxies are further divided into Seyfert 1.2, 1.5, 1.8, and 1.9’s, based on the increasing
relative strength of the Narrow to Broad line components (Osterbrock 1981).
In this Section we investigate differences and similarities in the emission line properties
of Seyfert 1 and 2 galaxies in our sample. Although we consider a galaxy with any broad
lines to be a Sy 1, we keep in mind the possibility that Seyferts with only very weak wings
on Hβ(Sy 1.8), or only on Hα(Sy 1.9), may turn out to be more similar to Seyfert 2 galaxies
in most observational respects.
3.1. Nuclear Reddening
Various physical models have been proposed to connect Sy 1 and Sy 2 galaxies. The
“unification model” assumes that the sole difference is the viewing orientation of the Seyfert
nucleus with respect to our line of sight (Antonucci 1993). Thus there would be no dif-
– 7 –
ference between the Sy 1 and the Sy 2 nuclei if these galaxies were viewed from the same
direction. There are, however, alternate hypotheses. For example, in the Galactic Dust
Model (Malkan et al. 1998), Sy 2’s have intrinsically larger dust covering fractions due to
the presence of galactic dust lanes. This dust obstructs our view of the inner nuclear regions
where the engine and broad line region (BLR) are located. One consequence is that the
emission lines in Sy 2’s should on average have a greater degree of reddening than Sy 1’s.
To test this we compare two optical emission line-ratios in our data that are widely
separated in wavelength–the Balmer decrement (Hα/Hβ) and the narrow line region (NLR)
ratio [O II]/[S II]. The positive correlation between these two reddening-sensitive line ratios
is shown in Figure 2. We note that both the Sy1’s and the Sy2’s appear to lie along the
same correlation. Furthermore, this trend looks similar to the one defined by emission-line
galaxies measured in the SDSS, which are shown by the cloud of small grey/green points.
There are 135,116 galaxies of any origin (AGN or starburst) with spectra from DR7, which
have all four emission lines detected at greater than the 10-sigma level.
To quantify this trend, we compute proper least-squares fits for these groups of galaxies
separately. For this and all subsequent correlation analyses, we used a proper least-squares
(LSQ) fit (to account for the comparable errors in both line ratios). We use a FORTRAN fitting
program (Linear regression with measurement errors and scatter 5) written and described by
Akritas & Bershady (1996), to compute the orthogonal proper LSQ fits. The slope of the fit
for the 12µm sample is 0.44 ±0.14 for the Sy 1 Galaxies and 0.44 ±0.22 for the Sy 2’s. The
large sample of SDSS emission line galaxies also shows a similar line-ratio correlation, with a
slope of 0.33 ±0.004 and intercept of 0.64 ±0.08. For comparison we calculate the predicted
slope in this line ratio correlation that should be produced solely by reddening of two fixed
(constant) intrinsic line ratios. Adopting the standard reddening law of Cardelli et al. (1989)
for R= 3.1, the predicted reddening slope should be 0.44. We placed the straight-line
reddening vector in the upper left in Figure 2 with tick marks showing increasing amounts
of E(B−V).
The quantitative fits confirm the visual impression of the figure: all three of the groups
of galaxies (Sy1s, Sy2s, and SDSS emission line galaxies) show a consistent correlation. The
positive slopes of each correlation are all consistent with each other, and with the prediction
from a standard reddening law. The Seyfert galaxies have more widely ranging line ratios
than those normally seen in the SDSS spiral galaxies, although the distributions mostly
overlap. We also compared the Hα/Hβ and [S II]/[O II] ratios with two other emission-line
ratios that should be sensitive to reddening. However, these other line ratios–[N II]/[O II]
5Obtained from the website for statistical packages: http://www2.astro.psu.edu/statcodes/sc correlregr.html.
– 8 –
and [O III]/[Ne III]–did not correlate so well the others, suggesting that they are influenced
by other factors beyond mere reddening. Thus, their intrinsic value can not be considered
constant, and we do not consider them valid reddening indicators. The results of all the
correlations are presented in Table 8.
There is reasonable consistency with the position of the lower-left (bluest) extent of
the line ratios in all three galaxy groups. This limit should correspond to galaxies with
essentially unreddened emission-line regions. We therefore interpret this lower left boundary
as indicating that both types of Seyferts, as well as normal star-forming galaxies, have
roughly the same intrinsic (unreddened) line ratios. Specifically, the line ratios in both Sy1s
and Sy2s have unreddened Hα/Hβratios which appear consistent with Case B’ (log Hα/Hβ
= 0.5), (Gaskell & Ferland 1984), also recommended by Malkan (1983)). The normal spiral
galaxies from SDSS have slightly lower Balmer decrements–they are consistent with the
normal Case B value of log Hα/Hβ= 0.45, although this small difference is only marginally
significant.
Supported by the consistency of the (Hα/Hβ) correlation with [S II]/[O II], we will now
assume that the primary determinant of where each galaxy lies in Figure 2 is its amount
of internal dust reddening. We therefore estimated E(B−V) values in each Seyfert galaxy
assuming the standard reddening law and intrinsic line ratios found above. The average of
the two line ratio estimators gives the E(B−V) values for individual Seyfert galaxies plotted
in Figure 3. This histogram shows that the average reddening is the same in both Seyfert
types: < E(B−V)>= 0.49 ±0.35 for the Sy1’s, and < E(B−V)>= 0.52 ±0.26 for the
Sy2’s. A Kolmogorov-Smirnov (K-S) test yields a probability of p(0.31) that the two groups
are drawn from the same distribution. In both Seyfert classes, E(B−V) ranges uniformly
from 0.0 to 1.0 mag. Furthermore, there is no trend for reddening to vary systematically
with the luminosity of the Seyfert galaxy, over more than two orders of magnitude in 12 µm
luminosity. For comparison, the average E(B−V) inferred from normal SDSS spirals is 0.63
mag. As shown in Figure 2, the Seyfert reddenings strongly overlap with these, except for
the larger scatter which may be due to observational uncertainties.
Other studies have been done on reddening in Seyfert galaxies. Rhee & Larkin (2005) re-
port substantial reddening in the narrow-line regions of 11 Seyfert 2 galaxies6and Tsvetanov & Iankulova
(1989) also finds that Sy 2’s are more reddened than the Sy 1’s for their sample of 24 Seyferts.
However, Malkan (1983) found only a marginal tendency for Seyfert 1’s to have smaller red-
dening in their forbidden line region, consistent with our new findings. Our finding of no
significant difference between the reddening of the NLRs in Seyfert 1 and Seyfert 2 galaxies
6These include Seyfert 1.8 and 1.9 galaxies.
– 9 –
is consistent with the expectation of the strong unification hypothesis.
We note that some narrow line regions appear to be optically thick at optical wave-
lengths, with E(B−V)>0.8 mag, which implies AV>2.5. If this reddening is also applied
to the non-stellar continuum, then the UV would be more then 99% extinguished – i.e.
virtually obliterated. It appears that some minority of the Seyfert nuclei we uncover in the
12 µm sample–unlike many other samples–are so reddened that some special explanation is
required for their observed UV continuum emission. The two possibilities are:
1. The UV continuum is purely from the stellar photospheres in the host galaxy. If the
UV continuum is particularly bright, this would be coming from a young population
of hot stars.
2. The observed UV continuum does come from the non-stellar engine (thought to be an
accretion flow around the central massive black hole), but suffers from less extinction
than the NLR. For example, we might have indirect unobscured views to the central
engine, which we see through scattered light (Antonucci 1993).
In most cases neither of these possibilities can be ruled out.
3.2. Degree of Gas Ionization
The blue portion of the optical spectrum contains several emission line ratios which are
sensitive to the level of ionization in the gas. Table 3 summarizes the ionization potential of
our emission lines. We select line ratios that differ in their ionization potential by 20 eV or
more. These diagnostic line ratios are observable by CCD spectrographs up to redshifts of
z∼0.9.
As summarized in Table 7, our averages for log([Ne V]/[Ne III]) are −0.07 ±0.24 and
−0.16±0.40 (individual scatter) for 32 Sy 1’s and 26 Sy 2’s, respectively. For log([O III]/[O II])
the averages are 0.59 ±0.51 and 0.43 ±0.52 for 50 Sy 1’s and 51 Sy 2’s, respectively. For
both line ratios, the K-S test shows that the small differences between the Sy1’s and Sy 2’s
are not significant at the 95% level 7.
7This finding contradicts Schmitt (1998). A possible explanation is the Schmitt (1998)’s selection of 52
Sy 1 galaxies from the literature may have missed some low-luminosity Sy 1’s. They are included in our
complete sample, and tend to have less highly ionized narrow lines. The only possible significant ionization
difference seen in our sample is that log([Ne III]/[O III]) is −0.17 ±0.40 in the Sy 1’s and −0.48 ±0.30
– 10 –
In Figure 4 we plot log[O III]/Hβversus log[Ne III]/[O II]. The dotted line is a model
for the NLR from Groves et al. (2004). The numbers along the line indicate the ionization
parameter U=S⋆/(nc), where S⋆is the flux of ionizing photons and nis the number
density of hydrogen atoms. Their model is of a dusty, radiation pressure-dominated region
surrounding a photo-evaporating molecular cloud, which in turn is surrounded by a coronal
halo where the dust has been largely destroyed (Dopita et al. 2002). In this model only
the NLR is included. The particular model that matches our data is un-reddened and un-
depleted. Groves et al. (2004) adopt a power-law ionizing continuum, Fν∝ναwith the best
slope of α=−1.4, and a number density of 103cm−3. The chemical abundances are solar.
As can be seen in Figure 4, the Sy 1’s are generally in the lower/right of the dotted line in
the graph, while the Sy 2’s are to the left and above. Since Hβis a broad line and the BLR
is not included in this particular model, we expect the Sy 1’s to have lower log[O III]/Hβ
values. The NLR ratios indicates an ionization parameter of log U=−2.5 for the average of
the Sy 1’s, and −2.8 for the average of the Sy 2’s.
3.3. [S II]/[N II] Ratio
We found that the log([S II]/[N II]) line ratio has almost the same value for Sy 1’s and
Sy 2’s, −0.23 ±0.24 and −0.28 ±0.20 respectively, with a K-S probability of p(0.97). For
the previously selected subset of Seyfert 2 galaxies we took from DR7, we find an average
value of log([SII ]/NI I]) of −0.19 ±0.1. This is also reasonably consistent with our values,
allowing for observational uncertainties (Figure 6). An average value of −0.23 ±0.18 for
this ratio can be used regardless of whether the galaxy is a type 1 or type 2 Seyfert. These
lines have about the same ionization potential (10.4 and 14.5 eV), but their critical densities
are different (∼103cm−3for [S II] and ∼105cm−3for [N II]). The fact that this ratio
is the same for both Seyfert types implies that the density structure of the narrow line
region is the same at least up to 105cm−3. This is consistent with the result found by
Nagao, Murayama, & Taniguchi (2001). They find that the ratio [S II]/[N II], among other
low-ionization line ratios, shows no difference between Sy 1’s and Sy 2’s. Our [S II]/[N II]
ratio is the same as that found in Sy 1 galaxies in (Stern & Laor 2013). They obtained
log([S II]/[N II]) = -0.25, independent of bolometric luminosity, for the mean stellar mass
in their sample <log M∗∼10.8>. We caution that, in the least-massive galaxies with
in the Sy 2’s. However, many of the measurements of “[Ne III]” in the Sy 1’s were obtained with low-
resolution spectroscopy, with moderate SNRs. We therefore suspect that some of the line flux attributed to
[Ne III]λ3869 in some Sy 1’s in our study, and in Schmitt’s, may instead be contaminated from the weak
broad line emission line He Iλ3889.
– 11 –
<log M∗∼10.4>, (Stern & Laor 2013) found slightly higher ratios of log([S II]/[N II])
&−0.20. Since we do not generally know M∗for all Seyfert host galaxies, this introduces
a small uncertainty that might contribute to the scatter we observe. Nonetheless, the near
constancy of the [S II]/[N II] can be useful in de-blending [N II] from broad Hαin Sy 1’s.
Some studies use other forbidden line fluxes to remove the [N II] which is blended with Hα
(Lacy et al. 1982).
3.4. Warm Dust
Our IRAS data give the ratio f25µm/f60µmintegrated over each entire galaxy, which in-
creases when the proportion of warm dust (heated by the AGN) increases. The [O III]/[O II]
ratio was previously used as a measure of the the relative strength of the Seyfert nucleus with
respect to the H II regions in the host galaxy. We plot [O III]/[O II] vs. f25µm/f60µm(Fig-
ure 5), and find a positive correlation for both Seyfert types. A Kendall’s Tau test reveals
that this gas ionization/dust temperature correlation holds for both Sy 1’s and Sy 2’s sepa-
rately, with confidence levels of CL = 99.9% and CL = 97.5% respectively. The individual
regression fits for the Sy galaxies are log([O III]/[O II]) = (0.97 ±0.12)log(f25µm/f60µm)
+ (1.03 ±0.63) for the Sy 1 and log([O III]/[O II]) = (0.69 ±0.16)log(f25µm/f60µm) +
(0.85 ±0.76) for the Sy 2 types.
3.5. Correlation of [OI] and [OIII]
The ionization potential of O Iis 13.6 eV and the critical density of [O I]λ6300˚
A is
1.8×106cm−3(De Robertis & Osterbrock 1986). The [O I] line is formed beyond the
classical ionization front, in a partially ionized region heated by X-rays from the AGN
(Veilleux & Osterbrock 1987; Spinoglio & Malkan 1992; Groves et al. 2004). We therefore
test the possibility that [O I] correlates better with the high-ionization fine structure emis-
sion lines of the Seyfert NLR, than with the low-ionization emission lines from H II regions.
In Figure 7 we plot the narrow line ratios log[O II]/[O III] vs. log[O I]/[S II]. As we found
in other NLR plots, the vertical axis is inversely proportional to the average ionization level
of the gas, and therefore increases in the relatively “weak” Seyferts, which have larger con-
tributions to their [O II] line emission from H II regions in their host galaxies. If the main
difference along the horizontal axis is also degree of gas ionization, then we would expect
a positive correlation in this graph. Instead, we find the opposite–an inverse correlation.
The orthogonal regression fit for the Sy 2 galaxies is: log([O II]/[O III]) = (−1.2±0.12)
log([O I]/[S II)] - (1.11 ±1.72). Although the spread is large, we find that a strong cor-
– 12 –
relation exists (Kendall’s Tau >99%). The Sy 1 sample also gives a fit with a negative
slope: log([O II]/[O III]) = (−0.36 ±0.16)log([O I]/[S II]) - (0.71 ±1.17). Here the Kendall’s
Tau is <90% and the null hypotheses can not be rejected, i.e. the Sy 1 correlation is not
significant. But for Seyferts, overall, the [O I] line tracks [O III] more closely than the [O II]
line.
Since this conclusion comes from the small dataset of our 12µm Seyferts, we sought
confirmation of this same inverse correlation in the much larger database of the SDSS DR7.
From this database of >900,000 galaxies, we restricted our consideration to 239,795 galaxies
in which the [NII], Hα, and [OIII] emission lines were detected at the 10-σlevel or better,
while the [OII] line was detected at the 20-σlevel. Using TOPCAT8to plot these line ratios
in the standard BPT diagram, we used the Select Tool feature to construct a subset of
15,190 “pure” Seyfert galaxies, whose emission spectra are dominated by the NLR. These
DR7 Seyferts are plotted in the figure with small grey/green dots, which were fitted by
the solid gray/green line with the proper LSQ. This fit is indistinguishable from what we
obtained fitting our much smaller Sy2 sample, but with a larger scatter (log([O II]/[O III]) =
(−1.3±0.87)log([O I]/[S II]) - (1.10 ±0.42)). Thus SDSS confirms the trend we found that
[O I]λ6300˚
A tracks the high-ionization gas in Seyferts more closely than the low-ionization
gas, because both are produced primarily by the AGN, not HII regions. 9
3.6. BPT Diagram
The BPT diagrams (Baldwin, Phillips, & Terlevich 1981) were developed to identify dif-
ferent photoionization mechanisms in galaxies, using ratios of lines at similar wavelengths,
to minimize the effects of reddening. In general, galaxies dominated by stellar photoioniza-
tion in H II regions have relatively stronger emission lines from less ionized gas, like [O II]
and [N II]. Active galaxies, in contrast, have a power-law ionizing continuum, which tends to
produce more lines from highly ionized gas. In addition, X-rays from the AGN can penetrate
into neutral or partially ionized zones to produce low-ionization lines that would typically
not be produced in H II regions. Thus AGN emission line spectra show a wider range of
ionization.
The most common BPT diagram uses the ratio of [O III]/Hβto [N II]/Hα, though
sometimes [S II] or [O I] is used in place of [N II]. These diagrams are designed for narrow-line
8Available at http://www.star.bris.ac.uk/∼mbt/topcat/.
9This finding holds for Seyfert nuclei, not for LINERs (see also Netzer (2009)).
– 13 –
objects, and therefore the Sy 1 galaxies are generally not displayed because of the presence
of broad permitted lines. However, we decided to include the weak broad-line objects to
determine where they would be located in the BPT diagram, for situations in which the
BLR and NLR are not separated.
Our BPT diagram is shown in Figure 8. The gray/green dots represent all 239,795
SDSS DR7 galaxies with highly significant detection in all four emission-lines. We include
the heavy dashed-dot line from Kauffmann et al. (2003) defined by: log([OII I]/Hβ)>
0.61/[log([NI I]/Hα)−0.05]+1.3, which separates H II regions from active galaxies. The light
dotted line from Kewley et al. (2001) defined by: log([OII I]/Hβ)>0.61/[log([NII ]/Hα)−
0.47)] + 1.19, also excludes “composite” galaxies which include too much emission form H II
regions to be classified as “pure” AGN.
The Sy 2 Galaxies cluster in the upper right of the BPT diagram, around (x, y) =
(0.05, 0.85). However, a significant minority of the Sy 2 galaxies fall close to the Kauffman
boundary: 17 out of a total of 82 Sy 2’s (21%), would be classified as “composite” (Seyfet
2 + HII) galaxies. This contamination of narrow emission line fluxes from the host spiral
galaxy is especially serious for less luminous AGN, when they are observed with relatively
poor spatial resolution (Theios et al. 2016).
3.7. BPT Classification of Broad Line AGN
The BLR contamination should in principle be removed before using the narrow line for
BPT classification. However, in some studies the spectra do not or cannot have the broad
Balmer components removed. This is especially true if the BLR component is very faint com-
pared to the NLR. We therefore plot the total line-flux ratios (broad + narrow components)
for Sy 1.5, 1.8, 1.8 and Sy 2 galaxies in Figure 8. In addition to the AGN/HII boundary
line from Kauffmann et al. (2003) we also plot the boundary line defined by Kewley et al.
(2001). The Sy 1.9 galaxies occupy the same region as the Sy 2’s (except for NGC 7314
which has a very weak nucleus and is located below the AGN/HII boundary line). This is
because their broad Balmer line components are so weak. Thus the BLR hardly alters the
ratio of forbidden lines to the permitted lines, away from the NLR values. This agrees with
the findings of Simpson (2005), who showed that his “Sy 1.x galaxies” having broad Hα
wings (what we call Sy 1.8 and Sy 1.9’s) are indistinguishable from the Sy 2’s in the BPT
diagram. Indeed, Simpson’s NLR mixing line - also plotted in Figure 8 - goes through most
of our Sy 1.8’s and Sy 1.9’s, with the latter lying further away from the AGN/HII boundary.
Our few Sy 1.8’s are all at an intermediate location in the BPT diagram. Their combination
of NLR and BLR compontents make them appear like composite AGN - mixtures of HII and
– 14 –
Seyfert 2 line emissions. The distribution of our own 12 µm Seyferts in the BPT diagram is
somewhat similar to that of the hard X-ray selected Seyferts from the BAT survey (Oh 2017).
However, our sample includes more low-luminosity Sy 2’s. Their line emission is more dom-
inated by HII regions. The relatively weaker NLR emission corresponds to lower hard X-ray
luminosity (see Appendix A). Thus they are likely to be missed by the BAT survey. Another
difference is that a substantial fraction of the BAT Seyfert 1’s fall left of the NLR area in
the BPT diagram. Their [N II]/Hαratios are anomalously low. This could have resulted
from mistakenly attributing some of their broad Hαemission to the NLR component.10
(Stern & Laor 2013) used SDSS spectra of 3175 Sy 1 galaxies to separate out the narrow
line components of Hαand Hβ. This allowed them to place the pure NLR line ratios on the
BPT diagram. Since we did not make this BLR/NLR decomposition, we can only compare
the their results to our own Sy 2’s and Sy 1.9’s, since they have negligible contribution from
the broad Balmer lines. (Stern & Laor 2013) found that the BPT ratios classify 5% of the
NLRs in their Sy 1 sample as “Star-Forming galaxies” (i.e. below the Kauffman line) and
15% of them as HII/AGN ( i.e. “Composites”, between the Kauffman and Kewley lines ).
Our narrow-line Seyferts show exactly the same distribution: three out of 65 are classified
as Star-Forming, while 12 out of 65 are classified as Composites. Thus the BPT distribution
of our NLR AGN sample is indistinguishable from SDSS Seyfert 1’s.
3.7.1. Quantative Decomposition of AGN Components in the BPT Diagram
To quantitatively interpret the location of AGNs in this BPT diagram, we make the
simplifying assumption that each galaxy has observed emission lines which are the sums of
three “pure” Seyfert 1, 2, and LINER components. To accomplish this crude separation we
have adopted the values for log([N II]/Hα) of -1.1, 0.05, 0.2 and log([O III]/Hβ) of -0.1, 0.85,
0.1, for the three respective components. The solid lines in Figure 8 connect each of these
three components, showing the ‘ mixing curves” obtained by combining varying proportions
of two of our “pure” components. Along the perimeter we have omitted the third emission
component; for the interior points we compute the contribution from all three types.
To determine the relative contribution from each particular type, we take the distance,
10The BPT diagram is supposed to include only the narrow emission lines. When their broad emission lines
are not removed, their contamination would push Sy galaxies to the left downward as in Figure 8. Indeed, the
BLR-dominated Seyferts nearly all fall below the AGN/HII separation line. For all the 32 Seyfert galaxies
classified as any type of broad line AGN (1, 1.2, 1.5, 1.8 and/or 1.9) we find that according to the definition
of the BPT diagram 59% and 75% of them lie below the Kauffman and Kewley lines, respectively.
– 15 –
diin the BPT diagram of each Seyfert to the three points defined as pure Seyfert 1, 2, and
LINER. The contribution is defined as
Ci=1/di
X
i
1/di
(2)
where irefers to Seyfert 1, 2, and LINER emission components, and
d2
i=log [OI II]
Hβ −log [O III ]
Hβ i2
+log [NII ]
Hα −log [N I I ]
Hα i2.(3)
Our three-component decompositions of the relative AGN contribution are shown graph-
ically in Figure 8. The resulting components for individual galaxies are shown in Figure 9.
The average BPT contributions of the “BLR” and “NLR” components change from
72%/6% in pure Sy 1’s to 56%/17% in Sy 1.5’s, to 43%/12% in Sy 1.8’s, and to 20%/58%
in Sy 1.9’s. In this highly simplified BPT decomposition, all types of Seyfert 1 nuclei show
an average contribution of 24% from a “LINER” component, regardless of the BLR/NLR
ratios.
As Figure 8 indicates, actual AGN show a continuously ranging mixture of NLR and
LINER components. There is no clear-cut separation between the two components. However,
an emission line galaxy cannot be reliably classified as predominantly a LINER unless it
shows a >50% contribution from the LINER component. The dot-dashed “mixing curve”
in Figure 8 shows the locus along which the LINER component, CLIN ER, equals 50%. As we
expected, the only galaxies lying to of the right of the boundary are classified as LINERs (the
rightmost filled circle is NGC 2639, classified in NED basic data as a Sy 1.9, but based on its
Activity Type it is considered a LINER). Interestingly, this line corresponds to the transition
between LINERs and Seyfert galaxies in Groves et al. (2004). That is, our CLI N E Rs = 50%
line accurately divides most of the Groves et al. (2004) Seyferts from the LINERs–the blue
and the red datapoints, respectively in their Figure 1. As expected from our use of NED
classification, hardly any of our Seyfert galaxies turn out to be LINER-dominated. The same
result applies to the BAT-selected survey (Oh 2017).
4. Broad Line Region, Eigenvector 1, and the Baldwin Effect Relationship
The luminosities of broad emission lines generally increase linearly with the non-stellar
continuum luminosity, but there are some exceptions. Baldwin (1977) found a negative
– 16 –
correlation between the equivalent width of C IVλ1549˚
A, and UV continuum luminosity
(LUV =νLνλ1449˚
A) in quasars, commonly referred to as the Baldwin effect. This less-
than-linear increase in C IV with underlying luminosity has been confirmed in many different
quasar and Sy 1 samples (Jensen et al. 2016). One interpretation is that the wavelength
peak of accretion disk luminosity shifts from the UV in quasars to the EUV in the less
luminous Seyferts, because the latter have smaller black holes with hotter accretion disks
(Zheng & Malkan 1993).
Our sample includes 23 Sy 1 galaxies with measured C IV emission lines (Table 5).
We exclude the galaxies MKN 231 and NGC 2841 galaxies because of their anomalously
weak C IV. Figure 10a shows a plot of log(C IV EW) versus log(LU V /1040) (color coding is
described in the following section). The solid line shows our proper least-squares fit to this
Baldwin relation
log(C IV EW ) = (−0.12 ±0.05) log LUV
1040 + (2.4±0.16).(4)
The amplitude of this slope is much smaller than what Baldwin (1977) found for his
sample of 20 quasars (−0.63). These is so much scatter in our sample that the anti-correlation
is only weakly significant.
4.1. Improving the Baldwin Effect with Eigenvector 1
Many studies have attempted to connect the Baldwin effect with some other parameter,
which could be more astrophysically fundamental than LUV (Zheng & Malkan 1993). The
most significant way in which the emission-line regions of quasars differ from each other is
in the strength of the so called ‘Eigenvector 1’ (Boroson & Green 1992). This parameter is
associated with stronger [O III] from the NLR, which is in turn strongly anti-correlated with
the permitted Fe II emission lines from the BLR. Boroson & Green (1992) also reported that
EW of [O III] and the luminosity of [O III] are both correlated with Eigenvector 1. They
suggested that a weaker Eigenvector 1 results when the Eddington ratio (L/LEdd) is larger.
Our sample does not contain reliable Fe II measurements. We therefore use the [O III]/Hβ
ratio instead to indicate the strength of Eigenvector 1.
We coded the Seyfert galaxy points in Figure 10a by their [O III]/Hβratio. Those with
log([O III]/Hβ)>0.28 are shown as red squares; Seyferts with weaker [O III]/Hβare shown
as blue diamonds. The strong segregation between the red and blue points shows that the
[O III]/Hβis indeed a good a predictor of EW(C IV). To find out how good, we plot log(C IV
– 17 –
EW) versus log([O III]/Hβ) in Figure 10b. It shows a significant positive correlation:
log(C IV EW ) = (0.42 ±0.1) log [OIII]
Hβ + (2.11 ±0.27).(5)
with a χ2/dof = 1.11 for 22 degrees of freedom. This new “substitute Baldwin relation”
works much better than the original (χ2/dof = 1.48, 21 degrees of freedom). We can also
see this in Figure 10b by the color coding. Seyferts with log(LUV )>43.15 as blue diamonds,
while the less luminous galaxies are shown as red squares. The large overlap of red and
blue points in this graph demonstrates that LU V is not so good predictor of EW(C IV) as is
[O III]/Hβ. Our finding of a positive correlation between EW (CIV) and Eigenvector 1 in
Seyfert 1 nuclei is fully consistent with Baskin and Laor’s finding of a similar correlation in
the PG Quasars (Baskin & Laor 2004).
Could both LU V and [O III]/Hβbe combined to make an even better Baldwin relation?
We made a bivariate least-squares fit:
log(C IV EW ) = (2.16 ±0.1) + (−0.02 ±0.04) log LU V
1040 + (0.37 ±0.12) log [OI II]
Hβ
(6)
The coefficient of the UV luminosity is only slightly negative, not significantly different from
zero. In contrast, the coefficient of log([O III]/Hβ) differs from zero by 3σ, indicating that
the Baldwin effect may be tightened by either adding this optical line ratio or replacing LUV
altogether. We use the Bayesian Information Criterion difference (∆BIC) to compare the
bivariate fit in Eqn. 6. The parameters from the original Baldwin effect (Eqn. 4) give a
BI C = 96.93 and the new “substitute Baldwin relation” gives BI C = 89.11. The inclusion
of the [O III]/Hβin the Baldwin relation give χ2/dof = 1.15 and BIC = 85.79. The ∆BI C
between LUV and the bivariate fit is >10. Thus the original Baldwin relation, using LUV
is significantly inferior. The ∆BI C obtained when adding LUV to the regression (Eqn. 6)
is 3.32, which is considered only marginal evidence in favour of adding LU V . Evidently,
Eigenvector 1 may be a more fundamental driver of the Baldwin effect. Perhaps LU V is
only a secondary parameter–it might correlate inversely with EW(CIV) merely because it
correlates inversely with log([O III]/Hβ). We graph this relation in Figure 10c, and indeed
there is a strong anticorrelation: log(LU V )= (−2.57 ±0.66)log([O III]/Hβ)+(2.57 ±1.51)
(Kendall’s Tau >99% significance).
Our data also show a weak positive correlation between EW(C III]1909) and [O III]/Hβ
. However, there is so much scatter that, unlike with the EW(C IV), this does not provide
a meaningful Baldwin relation. We also plot log(Mg II/[O II]) versus log([O III]/Hβ) in
Figure 11. The solid black line shows the fit to the broad-line objects only (Sy 1’s are
– 18 –
indicated by red traingles). The best fit slope is very significantly non-zero (−0.93 ±0.12),
i.e. the anti-correlation is significant at the 99% level. Since we use [O III]/Hβas our proxy
for “Eigenvector 1”, this anti-correlation shows that the relative strength of Mg II decreases
in strong “Eigenvector 1” Seyfert 1 galaxies. Although we lack sufficient Fe II data for our
12µm sample, this suggests the UV Mg II and optical Fe II emission line strengths should
be closely correlated. The average log(Mg II/[O II]) ratio in the NLR alone (Seyfert 2’s) is
−0.99 ±0.63.
5. Conclusions
We have shown that the narrow-line regions of Seyfert 1’s and Seyfert 2’s galaxies have
little systematic difference in many properties:
1. The luminosity functions of the narrow lines are the same for both types of Seyferts.
Therefore the space densities of Seyfert 1 and Seyfert 2 galaxies are roughly equal.
(Only the Hαand Hβluminosity functions differ in Sy 1’s at higher luminosities.)
2. Measured by two independent emission-line ratios (Hα/Hβand [S II]/[O II]), the Sy
2’s are not more reddened than the Sy 1’s. This indicates that the amount of dust
present in the narrow line region in both types of Seyferts is similar. Nor do the Sy1’s
show a significantly higher ionization in their NLRs, also consistent with the premise
of simple geometric unification. We have demonstrated that a value of −0.25 ±0.16
can be used for log([S II]/[N II]), regardless of whether the galaxy is a type 1 or 2
Seyfert. This can be useful in determining the flux of [N II] when it is blended in Hα.
3. We identify several ratios indicative of the ratio of the AGN luminosity to that of
the host galaxy (the “Seyfert dominance”). For example the dust in Seyfert galaxies
is warmer for those objects more dominated by their AGN contribution than by their
starburst or H II region contributions. And we have found that [O I]λ6300˚
A correlates
most closely with the high-ionization lines, both being powered primarily by the AGN.
In the BPT diagram, we find that 15 % of our Seyfert galaxies would be classified
by Kauffmann et al. (2003) as “HII/AGN Composites”. A further 5 % have such a
weak Seyfert nuclei that their spectra are not distinguishable from those of normal
star-forming spiral galaxies.
4. By separating the Seyfert galaxies into sub-types (1+1.2 +1.5 and 1.8 +1.9) we showed
that Seyfert 1.8 and 1.9 galaxies have similar ionization levels to Sy 2’s. The Sy 1.8
and Sy 1.9’s have very weak broad lines, so their narrow line ratios are similar to that
– 19 –
of Sy 2’s. The BPT diagram demonstrates that these objects lie in the same area as
Sy 2’s as well.
5. We use the BPT diagram to make a simple decomposition of emission line ratios into
three “pure” components –Sy1, Sy2, and LINERs. As expected, the relative importance
of the Sy 1 component falls steadily from the Sy 1.2’s to the Sy 1.5’s, the Sy 1.8’s,
and finally the Sy 1.9’s. Although AGNs show a continuous range in NLR/LINER
ratios, we find that the LINER component needs to be >50% for an AGN to be
specroscopically classified as a LINER.
6. In the broad-line region we studied the Baldwin effect and found that C IV equivalent
width correlates more strongly with [O III]/Hβ, rather than with UV luminosity. This
may imply the Baldwin effect is more strongly dependent on the Eddington ratio,
L/LEdd . An additional implication was that broad Mg II emission correlates with
Fe II.
This paper has benefited greatly from the extensive valuable comments of an anonymous
referee, whom we thank. This research has made use of the NASA/IPAC Extragalactic
Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute
of Technology, under contract with the National Aeronautics and Space Administration.
Funding for the Sloan Digital Sky Survey (SDSS) has been provided by the Alfred P.
Sloan Foundation, the Participating Institutions, the National Aeronautics and Space Ad-
ministration, the National Science Foundation, the U.S. Department of Energy, the Japanese
Monbukagakusho, and the Max Planck Society.
A. APPENDIX A. Emission Line Proxies for the Nonstellar Luminosity
Many previous studies pursued the goal of estimating the total non-stellar power of an
AGN by simply measuring one of the strongest emission lines in its spectrum. Any such
“shortcut”, if reliable over a wide range of Seyfert galaxies, would be of value in analyzing
large samples, where good data at all wavelengths may not be available. However, the
danger of correlating any two luminosities against each other in astronomical samples is well
known. When a wide range of luminosities is present, as in our Seyfert sample, an apparent
correlation of one luminosity with another luminosity is often found, even when there may
be little physical connection between the two quantities. Correlations of line luminosities
with continuum luminosities in the 12µm Seyfert sample should be expected, and do not
necessarily prove how they are linked.
– 20 –
Keeping this caveat in mind, we now examine these line/continuum correlations to de-
termine the intrinsic scatter, possible non-linearities and systematic differences depending on
which quantities and which AGN types are included. Our dataset–a representative sampling
of a very wide range of local AGN properties–has advantages for finding and testing various
“scaling relations” between emission lines and the broadband AGN continuum. The various
AGN-powered luminosities we are seeking to correlate span almost six orders of magnitude
in our Seyfert galaxies. Our estimates of non-stellar luminosities from the X-ray and UV
to the IR are given in Table 10. It is important to base correlations on a complete sample,
where the observed fluxes for nearly all the AGN are detections–not upper limits.
The key to all proposed scaling relations is finding easily measured quantities which
are dominated by the non-stellar (AGN) component. We start with the most obvious non-
stellar continuum – hard X-rays – since they are thought to be produced almost entirely by
the central engine of the AGN, except at the lowest luminosities. As Brightman & Nandra
(2011a) and others point out, it is very difficult for any galaxy lacking a Seyfert nucleus to
produce (through normal stellar processes, including X-ray binaries) more than 1040 erg/sec
in hard X-rays. In Figures 12a to 15a (top left in these panels) we compare the luminosities
of several of the strongest optical emission lines with the hard X-ray luminosities (LHX )
reported by Brightman & Nandra (2011a,b); Panessa et al. (2008).
In the Sy 1’s, LH X is reasonably well correlated with LHα . The plotted lines assumed a
linear correlation, i.e., fixing the slope in log(line) versus log(continuum) to be 1. There are
no cases when a deviation from this linear slope gives a statistically superior fit. The result
that LHX ∼15LH α is not surprising, since it has long been known that the broad Balmer
emission line luminosities are closely correlated with the non-stellar continuum (Yee 1980;
Malkan & Sargent 1982). As summarized in Table 9, we find an individual scatter of a
factor of 3 (solid blue squares in Figure 13a). This is far from perfect, but might be useful
in cases where only a rough individual estimate, or average of a sample, is wanted.
A limitation of that relation is that a high-quality spectrum of the Hαregion may not
be available, so it is sometimes unclear whether an AGN should be classified as Sy 1 or Sy
2. Then this correlation runs into trouble, since Sy 2’s lack any directly detectable broad
Hαemission. They produce more hard X-rays for a given (narrow) Hαluminosity, and the
scatter is so large it nearly destroys the HX/Hαcorrelation in Sy 2’s.
To include both Sy 1’s and Sy 2’s in one single AGN correlation, we have followed previ-
ous studies, such as Tommasin et al. (2010). We considered whether the strongest emission
from the NLR – the [O III]λ5007˚
A line – could instead serve as a proxy to measure the
non-stellar luminosity (Mulchaey et al. 1994; Dasyra et al. 2008; Goto et al. 2011). Indeed,
the blue symbols in Figures 14a and 15a do show that the [O III] luminosity can roughly
– 21 –
predict the HX luminosity, with an uncertainty of 4 times for Sy 1’s and 10 times for Sy 2’s.
For a given λ5007 line luminosity, the Sy 1’s tend to be 30% brighter in hard X-rays. That
is a small difference compared with the large intrinsic scatter.
Our LHX/L[OIII] correlation for the Sy 1’s has nearly the same normalization and scatter
as Heckman (2005) found for local AGN selected by O IIIλ5007˚
A (LHX/L[OIII] = 101.64 com-
pared with our LHX/L[OIII] = 101.59). This is also the same correlation found by Xu et al.
(1999) (LHX/L[OIII] = 101.60), as well as (Stern & Laor 2012), (their Equation 6). We find
that the Sy 2’s have relatively weaker hard X-rays and larger scatter in the L[OIII] correla-
tion. But this Sy 1/2 difference is larger in the Hao et al. (2005) sample, which includes
a substantial tail of Sy 2’s with very weak hard X-rays (log(L[OIII]) = 39 −41). Hao et al.
(2005) points out that the X-ray selected AGN will have relatively larger LHX/L[OIII] ratios
than our sample selected at longer wavelengths. Indeed this trend is seen in the luminous
AGN sample found in Chandra Deep Field South Survey at 0.3< z < 0.8. The best fit
correlation for AGN, found by Netzer (2006) is shown by the dashed line in Figures 14a and
15a. In contrast, up to LHX ∼1044 we see no evidence of non-linearities in the correlation
with any of the emission-line fluxes.
But we can improve on using LOIII as a general measure of AGN power. We are mo-
tivated by the Stern and Laor studies (papers II and III) of 3175 SDSS spectra of Sy 1
galaxies. (Stern & Laor 2012) showed that the luminosity of the broad component of Hα
increase more rapid then the luminosities of any of the narrow lines. We propose to account
for this simply by adding measures of NLR and BLR luminosities together. To include
Seyferts of Type 1 and Type 2, we favor a “hybrid” indicator of non-stellar luminosity, the
sum of the [O III] and Hβlines - shown by the open symbols in Figures 14a and 15a.
This empirical compromise captures the AGN luminosity in Sy 2’s emerging in the NLR,
but also the BLR luminosity in Sy 1’s. And as shown by Table 10, one single relation,
LHX ∼25L([OIII ]+H β ), gives a rough estimate for any Seyfert galaxy, regardless of type.
The usefulness of this hybrid (NLR+BLR) AGN luminosity indicator lead us to also
consider correlations with the blend of Hα+[N II] (open symbols in Figures 13a and 14a),
which can be difficult to disentangle at low spectral resolution. The correlation with HX
shows an extreme amount of scatter for the Sy 2’s, but is still consistent with the same
(tighter) correlation we find for Sy 1’s. We therefore consider the combined Hα+[N II] line
luminosity a possible “backup” predictor of non-stellar hard X-rays, which is also indepen-
dent of Seyfert type: LH X ∼11L([N I I ]+H α). Since the broad Hαand Hβin Seyfert 1 galaxies
vary with time, the non-simultaneity of our spectroscopy with the X-ray observations intro-
duces some artificial scatter into these diagrams. However, the amplitude of variability is
usually small enough that this increase in scatter is small compared to what is observed in
– 22 –
Runco, J., et al. (2016).
A.0.1. Less Reliable Continuum Proxies for the total Nonstellar Luminosity
The hard X-rays only carry a minority (≤10%) of the bolometric luminosity of most
AGN, and this fraction tends to be significantly lower in more luminous objects. We therefore
also searched for emission line proxies which could predict non-stellar luminosity at longer
wavelengths closer to the bolometric peak output of typical AGN. The near-UV luminosity
of Sy 1’s (around the peak of the “Big Blue Bump” (Malkan & Sargent 1982)), tends to be
about four to five times the HX. As we found for the hard X-rays, the NUV luminosities
can also be predicted by our combined [O III] and [O III]+Hβluminosities, with similar
scatter (±4 times for the Sy 1’s, and ±9 for the Sy 2’s), shown in Figures 12b to 15b. Our
normalization of LU V ∼200L[OI I I ]and LU V ∼100L([OI I I ]+H β)for Sy 1s are similar to the
range of models in Netzer (2006), assuming bolometric luminosity is a few times LU V .
However, these emission line correlations with NUV luminosity have very different nor-
malizations: for a given line luminosity, the Sy 2’s produce 3 to 8 times more NUV than the
Sy 1’s. Although we are correlating observed luminosities, this discrepancy is not explained
by extinction differences, since we did not find strong Sy 1/Sy 2 differences in optical es-
timates of E(B−V), and the non-stellar UV continuum extinction should be even larger.
Instead, a plausible explanation is that the UV continuum in Sy 2’s is dominated by recently
formed hot stars which were also included in the large GALEX beam, outshining a possible
AGN contribution.
Next, we made an even less reliable effort to isolate the non-stellar nuclear continuum,
using the 1.2 µm, 1.6 µm, and 2.2 µm luminosities of the central pixels in 2MASS images of
each Seyfert nucleus (Figures 12c to 15c). We decomposed the near-IR luminosities of the
nucleus into contributions from the AGN and starlight, both assumed to have constant colors.
The uncertainties in deriving the 2.2 µm luminosities of the weaker AGN are at least a factor
of 2 by this so-called “color given” method (Malkan & Filippenko 1983). The correlations
of line luminosities with this non-stellar 2 µm luminosity are even weaker than with the
HX and NUV, with very large scatter for both Sy 1’s and Sy 2’s. The only line luminosity
predictor worth possibly considering is the sum of [O III]+Hβ, which when multiplied by
15 gives a rough prediction for the non-stellar 2.2 µm luminosity of any Seyfert nucleus,
regardless of type. And our same backup line ratio, the sum of Hα+[N II], when multiplied
by by 5.5, gives a rough prediction, with the Sy 2 nuclei being on average 40% brighter at
2.2 µm .
– 23 –
Our final estimator of the non-stellar AGN luminosity was made from small-beam pho-
tometry at 10 µm, from Gorjian et al. (2004) (Figures 12d to 15d). The correlations with
the line luminosities are too poor to be of much use, with worse than a factor of ten scatter.
The 10 µm correlation with [O III] luminosity is the same for Sy 1’s and Sy 2’s, perhaps
because both are thought to originate in the NLR.
Five of the broad-line AGN in our sample are LINERs which have been found to exhibit
weak broad-line components, usually faint extended wings under Hα(Ho et al. 1997b). We
have included these with the Seyfert 1’s in Figure 13, but plotted with stars and ’x’ symbols,
not boxes. We were surprised to see that these AGN, which Ho et al. denotes as “LINER 1’s”
also follow roughly the same correlations between emission lines and non-stellar continuum
as do the normal Sy 1’s.
Similarly, a dozen of the narrow-line AGN in our multi-wavelength sample are now best
classified as LINERs, and we included them in the panels of Figures 13 and 15, along
with the normal Sy 2’s, except plotted with star symbols. Within the large scatter of the
non-stellar line/continuum luminosity correlations, we again note that the LINERs lacking
any broad wings overlap entirely with the Sy 2s. Our results would hardly have changed
whether they were included or excluded in the linear fits. Given the large scatter in these
emission-line/continuum luminosity correlations, they cannot be used to separate various
AGN types, even LINERs.
REFERENCES
Akritas M. G. & Bershad M. A.1996 ApJ, 470, 706
Antonucci, R., 1993, ARA&A, 31, 473
Armus, L., Heckman, T. M., & Miley, G. K. 1989, ApJ, 347, 727
Baldwin, J. A. 1977, ApJ, 214, 679
Baldwin, J. A., Phillips, M. M., Terlevich, R., 1981, PASP, 93, 5
Barr, P., Willis, A. J., & Wilson, R. 1983, MNRAS, 203, 201
Baskin, A. & Laor, A. 2004, MNRAS, 350, 31
Bonatto, C. J., & Pastoriza, M. G. 1997, ApJ, 486, 132
Bongiorno, A., Mignoli, M., Zamorani, G., Lamareille, F., et al. 2010, A&A, 510, A56
– 24 –
Boroson, T. A. & Green, R. F. 1992, ApJS, 80, 109
Boroson, T. A., & Meyers, K. A. 1992, ApJ, 397, 442
Brightman, M. & Nandra, K, MNRAS, 413, 1206
Brightman, M. & Nandra, K, MNRAS, 414, 3084
Buson, L. M. & Ulrich, M.-H. 1990, A&A, 240, 247
Cardelli, J. A., Clayton, G. C., Mathis, J. S. 1989, ApJ, 345, 245
Clavel, J., Benvenuti, P., Cassatella, A., Heck, A., Penston, M. V., Selvelli, P. L., Beeckmans,
F., & Macchetto, F. 1980, MNRAS, 192, 769
Clavel, J., & Joly, M. 1984, A&A, 131, 87
Colbert, E. J. M., Baum, S. A., Gallimore, J. F., O’Dea, C. P., Lehnert, M. D., Tsvetanov,
Z. I., Mulchaey, J. S., & Caganoff, S. 1996, ApJS, 105, 75
Coziol, R., Pena, M., Demers, S., & Torres-Peimbert, S. 1993, MNRAS, 261, 170
Cruz-Gonzalez, I., Carrasco, L., Serrano, A., Guichard, J., Dultzin-Hacyan, D., & Bisiacchi,
G. F. 1994, ApJS, 94, 47
Dasyra, K. M., Ho, L. C., Armus, L., Ogle, P.; Helou, G., Peterson, B. M., Lutz, D., Netzer,
H., Sturm, E. 2008, ApJ, 674L, 9
de Grijp, M. H. K., Keel, W. C., Miley, G. K., Goudfrooij, P., & Lub, J. 1992, A&A, 96, 389
De Robertis, M. M. & Osterbrock, D. E., 1986, ApJ, 301, 727
De Robertis, M. M., Hutchings, J. B., & Pitts, R. E. 1988, AJ, 95, 1371
Dopita, M. A., Groves, B. A., Sutherland, R. S., Binette, L., & Cecil, G. 2002, ApJ, 572,
753
Farrah, D., Surace, J. A., Veilleux, S., Sanders, D. B., & Vacca, W. D. 2005, ApJ, 626, 70
Ferland, G. J., & Osterbrock, D. E. 1986, ApJ, 300, 658
Freedman, W. L., Hughes, S. M., Madore, B. F., et al. 1994, ApJ, 427, 628
Gallego, J., Zamorano, J., Aragon-Salamanca, A., & Rego, M. 1995, ApJ, 455, L1
Gallego, J., Zamorano, J., Rego, M., Alonso, O., & Vitores, A. G. 1996, A&AS, 120, 323
– 25 –
Gaskell, C. M. & Ferland, G. J., 1984, PASP, 96, 393
Gorjian, V., Werner, M. W., Jarrett, T. H., Cole, D. M., Ressler, M. E. 2004, ApJ, 605, 156
Goto, T., Arnouts, S., Malkan, M., , et al., 2011, MNRAS, 414, 1903
Grandi, S. A. 1983, ApJ, 268, 591
Groves, B. A., Dopita, M.A., Sutherland, R.S. 2004, ApJS, 153, 9
Groves, B. A., Heckman, M.H., Kauffman, G. 2006, MNRAS, 371, 1559
Hao, L., Strauss, M. A., Fan, X., , et al., 2005, AJ, 129, 1795
Heckman, T. M, Ptak, A., Hornschemeier, A., Kauffmann, G. 2005, ApJ, 634, 161
Ho, L. C., Filippenko, A. V., & Sargent, W. L. W. 1996, ApJ, 462, 183
Ho, L. C., Filippenko, A. V., & Sargent, W. L. W. 1997, ApJS, 112, 315
Ho, L. C., Filippenko, A. V., Sargent, W. L. W., & Peng, C. Y. 1997, ApJS, 112, 391
Ho, L. C., & Peng, C. Y. 2001, ApJ, 555, 650
Huchra, J., Davis, M., Latham, D., Tonry, J. 1983, ApJS, 52, 89
Huchra, J. & Burg, R., 1992, ApJ, 393, 90
Jensen, T.W., Vivek, M., Dawson, K.S., Anderson, S.F., Bautista, J., et al. 2003, ApJ, 833,
199
Kauffmann, G., Heckman, T. M., Tremonti, C., et al., 2003, MNRAS, 346, 1055
Kewley, I., Dopita, M.A., Sutherland, R. S., Heiser, C. A., & Trevena, J. 2001, ApJ, 556,
121
Kim, D.-C., Sanders, D. B., Veilleux, S., Mazzarella, J. M., & Soifer, B. T. 1995, ApJS, 98,
129
Kirhakos, S. D., & Steiner, J. E. 1990, AJ, 99, 1722
Kishimoto, M., Antonucci, R., Cimatti, A., Hurt, T., Dey, A., van Breugel, W., & Spinrad,
H. 2001, ApJ, 547, 667
Kraemer, S. B., Wu, C.-C., Crenshaw, D. M., & Harrington, J. P. 1994, ApJ, 435, 171
– 26 –
Kuraszkiewicz, J. K., Green, P. J., Forster, K., Aldcroft, T. L., Evans, I. N., & Koratkar, A.
2002, ApJS, 143, 257
Kuraszkiewicz, J. K., Green, P. J., Crenshaw, D. M., Dunn, J., Forster, K., Vestergaard, M.,
& Aldcroft, T. L. 2004, ApJS, 150, 165
Lacy, J. H., Malkan, M., Becklin, E. E., Soifer, B. T., Neugebauer, G., Matthews, K., Wu,
C.-C., Boggess, A., & Gull, T. R., 1982, ApJ, 256, 75
Lumsden, S. L., Heisler, C. A., Bailey, J. A., Hough, J. H., & Young, S. 2001, MNRAS, 327,
459
Malkan, M. A. & Sargent, W. L. ApJ, 254, 22
Malkan, M. A., 1983, ApJ, 264, L1
Malkan, M. A. & Fillipenko, A.V. 1983, ApJ, 268, 562
Malkan, M. A., & Oke, J. B. 1983, ApJ, 265, 92
Malkan, M. A., Gorjian, V., & Tam, R. 1998, ApJS, 117, 25
McQuade, K., Calzetti, D., & Kinney, A. L. 1995, ApJS, 97, 331
Misselt, K. A., Clayton, G. C., & Gordon, K. D. 1999, PASP, 111, 1398
Moran, E. C., Halpern, J. P., & Helfand, D. J. 1996, ApJS, 106, 341
Morris, S. L., & Ward, M. J. 1989, ApJ, 340, 713
Morton, D. C., 2003, ApJS, 149, 205
Mulchaey, J. S., Koratkar, A., Ward, M. J., Wilson, A. S., Whittle, M., Antonucci, R. R. J.,
Kinney, A. L., & Hurt, T. 1994, ApJ, 436, 586
Nagao, T., Murayama, T., & Taniguchi, Y. 2001, PASJ, 53, 629
Netzer, H., 2006, A&A, 453, 525
Netzer, H., 2009, MNRAS, 399, 1907
Oh, K., 2017, MNRAS, 464, 1466
Osterbrock, D. E. 1981, ApJ, 249, 462
Osterbrock, D. E., & Martel, A. 1993, ApJ, 414, 552
– 27 –
Osterbrock, D. E. & Ferland, G. J., 2006, Astrophysics of Gaseous Nebulae and Active
Galactic Nuclei, Second Edition (Sausalito, California: University Science Books)
Panessa, F., Bassani, L., de Rosa, A., et al. 2008, A&A, 483, 151
Poggianti, B. M., & Wu, H. 2000, ApJ, 529, 157
Polletta, M., Bassani, L., Malaguti, G., Palumbo, G. G. C., & Caroli, E. 1996, ApJS, 106,
399
Reynolds, C. S., Ward, M. J., Fabian, A. C., & Celotti, A. 1997, MNRAS, 291, 403
Rhee, J. H. & Larkin, J. E., 2005, ApJ, 620, 151
Rodr´ıguez-Ardila, A., Pastoriza, M. G., & Donzelli, C. J. 2000, ApJS, 126, 63
Runco, J., Cosens, M., Bennert, V. N., et al. 2016, ApJ, 821, 33
Rush, B., Malkan, M. A., Spinoglio, L. 1993, ApJS, 89, 1
Schmitt, H. R. 1998, ApJ, 506, 647
Simpson C., 2005, MNRAS, 360, 565
Sosa-Brito, R. M., Tacconi-Garman, L. E., Lehnert, M. D., & Gallimore, J. F. 2001, ApJS,
136, 61
Spinoglio, L. & Malkan, M. A., 1989, ApJ, 342, 83
Spinoglio, L. & Malkan, M. A., 1992, ApJ, 399, 504
Stearn, J.& Laor, A., 2013, MNRAS, 426, 2703
Stearn, J.& Laor, A., 2013, MNRAS, 431, 836
Storchi-Bergmann, T., Kinney, A. L., & Challis, P. 1995, ApJS, 98, 103
Theos, R., Malkan, M. & Ross, N. 2016, ApJ, 822, 45
Thuan, T. X. 1984, ApJ, 281, 126
Thuan, T. X. & Sauvage, M. 1992, A&AS, 92, 749
Toba, Y., Oyabu, S., Matsuhara, H., Malkan, M. A., et al. 2014, ApJ, 788, 45
Tommasin, S., Spinoglio, L., Malkan, M. A., Fazio, G. 2010, ApJ, 709, 1257
– 28 –
Tsvetanov, Z. I. & Iankulova, I. M., 1989, MNRAS, 237, 707
Veilleux, S. & Osterbrock, D. E., 1987, ApJS, 63, 295
Veilleux, S., Kim, D.-C., Sanders, D. B., Mazzarella, J. M., & Soifer, B. T. 1995, ApJS, 98,
171
Veilleux, S., Kim, D.-C., & Sanders, D. B. 1999, ApJ, 522, 113
Wang, T.-G., Zhou, Y.-Y., & Gao, A.-S. 1996, ApJ, 457, 111
Wang, T.-G., Lu, Y.-J., & Zhou, Y.-Y. 1998, ApJ, 493, 1
Winkler, H. 1992, MNRAS, 257, 677
Xu, C., Livio, M., Baum, S., et al. ApJ, 118, 1169
Yee, H. K. (1980), ApJ, 241, 894
Young, S., Hough, J. H., Efstathiou, A., Wills, B. J., Bailey, J. A., Ward, M. J., & Axon,
D. J. 1996, MNRAS, 281, 1206
Zheng, W. & Malkan, M. A., 1993, ApJ, 415, 517
This preprint was prepared with the AAS L
A
T
E
X macros v5.2.
– 29 –
-10
-9
-8
-7
-6
-5
-4
-3
-2
36 37 38 39 40 41 42 43 44
Log(Φ(L)) Mpc-3 mag-1
Log(L(H
α
)) erg s-1
Sy 1
Sy 1 Best fit
Sy 2
Sy 2 Best fit
Sy 1 + Sy 2 (NLR+BLR) Hao (2005, Fig 3) -10
-9
-8
-7
-6
-5
-4
-3
-2
36 37 38 39 40 41 42 43 44
Log(Φ(L)) Mpc-3 mag-1
Log(L(H
β
)) erg s-1
Sy 1
Sy 1 Best fit
Sy 2
Sy 2 Best fit
-10
-9
-8
-7
-6
-5
-4
-3
-2
36 37 38 39 40 41 42 43 44
Log(Φ(L)) Mpc-3 mag-1
Log(L[OIII]) erg s-1
Sy 1
Sy 1 Best fit
Sy 2
Sy 2 Best fit
Sy 1 Hao (2005, Fig 6)
Sy 2 Hao (2005, Fig 6)
Sy 2 Bongiorno (2010) -10
-9
-8
-7
-6
-5
-4
-3
-2
36 37 38 39 40 41 42 43 44
Log(Φ(L)) Mpc-3 mag-1
Log(L[OII]) erg s-1
Sy 1
Sy 1 Best fit
Sy 2
Sy 2 Best fit
Sy 1 Hao (2005, Fig 6)
Sy 2 Hao (2005, Fig 6)
Fig. 1.— Emission Line Luminosity Functions for Hα, Hβ, [O III]λ5007, and [O II], where Φ
has units of Mpc−3mag−1and L has units of ergs s−1. The solid line is the fit to the Seyfert 1
LF’s, while the blue line is the fit to the Seyfert 2’s. In the graphs of LF for Hα, [O III], and
[O II] we have overplotted the Luminosity Functions of SDSS Seyfert galaxies from Hao et al.
(2005). In the panel for [O III] we have also added LF data from Bongiorno et al. (2010)
(zCOSMOS). Although data from Hao et al. (2005) covers our zrange (0 < z < 0.15), the
zCOSMOS data (0.15 < z < 0.35) agrees better with our LF’s.
– 30 –
0
0.5
1
1.5
2
-1 -0.5 0 0.5 1 1.5
Log(H
α
/H
β
)
Log([SII]/[OII])
E(B-V) = 0.0
0.2 0.4 0.6 0.8 1.0
| |
SDSS - Data
SDSS - Fit
12
µ
m Sample Sy 1
CfA Sample Sy 1
12
µ
m Sample Sy 2
CfA Sample Sy 2
Theoretical E(B-V)
Fig. 2.— Nuclear reddening-sensitive emission-line ratios log(Hα/Hβ) vs. log([S II]/[O II]).
Open symbols represent galaxies added data from the CfA sample. The best fit slopes for
the Sy 1’sand Sy 2’s are identical; Sy 1: 0.44 ±0.12 (red line), Sy 2: 0.44 ±0.22 (blue
line). The best fit to the SDSS data is y= (0.33 ±0.004)x−(0.64 ±0.08). The theoretical
reddening vector, derived from Cardelli et al. (1989), is arbitrarily offset for clarity. Our
Sy 1 and Sy 2 galaxies are plotted with red triangles and blue squares, respectively, with
typical uncertainty shown by the black error bar. The large number of (135,116) of SDSS
DR7 galaxies with strong emission-lines (of any origin–AGN or starburst) are plotted as a
cloud of light green/gray dots.
– 31 –
0
2
4
6
8
10
12
14
16
18
-0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Number Density
<E(B-V)>
Sy 2 - 12 µm
Sy 1 - 12 µm
SDSS/5000
Fig. 3.— Histogram of the average E(B-V) using Hα/Hβand [S II]/[O II] line ratios for the
12 µm Seyferts and the SDSS galaxies. The average reddening is essentially the same in both
Seyfert types: < E(B−V)>= 0.49 ±0.35 for the Sy1’s, and < E(B−V)>= 0.52 ±0.26
for the Sy2’s. The average reddening for the SDSS galaxies is < E(B−V)>= 0.63 ±0.21.
– 32 –
-1
-0.5
0
0.5
1
1.5
2
-2 -1.5 -1 -0.5 0 0.5 1
Log([OIII]/Hβ)
Log([NeIII]/[OII])
-4.0
-3.6
-3.3
-3.0
-2.6 -2.3 -2.0 -1.6 -1.3
Ionization Models, log(U)
Sy 1
Sy 1.2
Sy 1.5
Sy 1.8
Sy 1.9
Sy 2
Fig. 4.— Logarithmic line ratios of[O III]/Hβvs. [Ne III]/[O II]. The dashed line is a
sequence of models for the narrow line region from Groves et al. (2004) with the numbers
along the line indicating the ionization parameter U=S⋆/(nc), where S⋆is the flux of
ionizing photons and nis the number density of hydrogen atoms. Since this model only
includes the NLR, most of the Sy 1 galaxies, especially Seyfert 1, 1.2, and 1.5’s, falls in the
lower right section of the graph, because of their additional Hβ contribution from the BLR.
– 33 –
Number Density
Log([SII]/[NII]λ6731)
Gaussian Fit:
Sy 1: µ = -0.23, σ = 0.24
Sy 2: µ = -0.28, σ = 0.20
SDSS - Sy 2: µ = -0.19, σ = 0.15
0
4
8
12
16
20
24
28
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6
Fig. 5.— Histogram of [S II]/[N II] line raions for our 12 µm Seyfert 1 and Seyfert 2 galaxies
(red and blue hatching, respectively). These distributions are quite similar to the line ratios
observed in our SDSS DR7 sample of 16,708 Sy 2 galaxies, shown in light green. Each of these
distributions can be fitted approximately by a Gaussian. The mean and standard deviation
values are given. All of the distributions overlap substaintially, justifying our adoption of
a fairly “universal” of log([S II]/[N II]])= −0.23. The histogram widths are likely to be
dominated by observational uncertainties.
– 34 –
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
-1.25 -1 -0.75 -0.5 -0.25 0 0.25 0.5
Log([OIII]/[OII])
Log(f25
µ
m/f60
µ
m)
Sy 1
Sy 1 Best fit
Sy 2
Sy 2 Best fit
Fig. 6.— Logarithmic line-ratios [O III]/[O II] vs. f25µm/f60µm, which both measure the
relative Seyfert nucleus power, compared with the lower-ionization gas and the cooler dust
in HII regions. Although this positive correlation shows large scatter, it is significant, with
a Sy 1 regression fit of log([O III]/[O II]) = (0.97 ±0.12)log(f25µm/f60µm) + (1.03 ±0.63)
and log([O III]/[O II]) = (0.69 ±0.16)log(f25µm/f60µm) + (0.85 ±0.76) for Sy 2 types. A
Kendall’s Tau significant test confirms that the gas ionization/dust temperature correlation
holds for both Sy 1’s and Sy 2’s, at a CL >97.5%.
– 35 –
-2
-1.5
-1
-0.5
0
0.5
1
-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Log([OII]/[OIII])
Log([OI]/[SII])
Sy 1 - 12 µm Fit: y=(-0.36 ± 0.16)x-(0.71 ± 1.17)
Sy 2 - 12 µm Fit: y=(-1.2 ± 0.12)x-(1.11 ± 1.72)
SDSS - Sy 2 Fit: y=(-1.3 ± 0.87)x-(1.10 ± 0.42)
Fig. 7.— Correlation of log([O II]/[O III]) vs. log([O I]/[S II]). We again show our Sy 1 and
Sy 2 galaxies as red triangles and blue squares, respectively. The light green/grey points is a
subsample of 13,688 Seyfert 2 galaxies from the DR7 of the SDSS. The inverse correlation is
the opposite of an ionization effect, and indicates that [O I] follows [O III] more closely than
[S II]. This implies that the [O I] emission comes predominantly from the Seyfert nucleus,
not from HII regions. The Kendall’s Tau significant test reveals that it is significant for the
Sy 2 galaxies but not for the Sy 1’s, CL = 99.9% and CL <90% respectively. The slope of
the Sy 2 fit is −1.2±0.12 and the slope for the SDSS fit is −1.3±0.87.
– 36 –
-1
-0.5
0
0.5
1
1.5
-2 -1.5 -1 -0.5 0 0.5 1
Log([OIII]/Hβ)
Log([NII]/Hα)
Sy 2 - 12µm sample
Sy 2 CfA Sample
Sy 1.5
Sy 1.8
Sy 1.9
Sy2/LINERs
Sy1/LINERs
Sy 1
Sy 2
LINERs
50% LINERs mixing-line
Simpson Sy 2 Mixing Ridge
Fig. 8.— BPT line-ratio diagram of our Sy 1.5, 1.8, 1.9, and Sy 2 galaxies, shown by the
same color symbols as before. The green/grey dots represent the data form emission-line
galaxies from SDSS DR7. We include the boundaries separating AGN and SFG/HII regions
from Kauffmann et al. (2003)) (dashed light-grey) and Kewley et al. (2001) (solid light-grey).
Most of the Sy 2’s fall in the upper right corner as expected, as do Sy 1.9 ’s. Because of
their BLR contimination of the H line-fluxes, the Sy 1.8’s fall in the SFR/AGN “composite’
region of the BPT diagram, while the Sy 1.5’s scatter below the HII/AGN boundary. The
mixture of the pure Sy 2 line emmision and HII regions seen in our 12 µm Seyferts are well
matched by the Sy 2 “ridge line” from Simpson (2005). The three vertices of the black
triangle show our adopted line rations for pure “NLR” (Sy 2, upper right), “BLR” (Sy 1,
lower left), and ”LINERs” (lower right). The curved line inside the triangle shows a 50%
mix of LINER emissions and Seyfert lines. This curve corresponds closely to the boundary
between Seyfert 2 (above) and LINER (below) galaxies.
– 37 –
Fig. 9.— Emisson line components of 12 µm of Seyfert 1 galaxies based on their locations
in the BPT diagram. White corresponds to the Seyfert 1 contribution, the light gray cor-
responds to the Seyfert 2 contribution, and the dark gray corresponds to LINER. AGN
components classified as Sy 1.8 and 1.9 are dominated by the narrow line component as in
the Sy 2’s, while the Sy 1.5’s are a mix of the Sy 1 and the Sy 2 line components. Spectro-
scopically classified LINERs have 50% or more of the emission from the LINER component.
– 38 –
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
−1 −0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Log(CIV EW)
Log(LUV/1040)
log[OIII]/Hβ > 0.28
log[OIII]/Hβ < 0.28
Best Fit: slope = −0.12 ± 0.06,
σ
= 0.26 1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
−1 −0.75 −0.5 −0.25 0 0.25 0.5 0.75 1
Log(CIV EW)
Log([OIII]/H
β
)
Best Fit: slope = 0.42 ± 0.13,
σ
= 0.19
log(LUV/1040) > 0.40
log(LUV/1040) < 0.40
−1
0
1
2
3
4
5
6
−1.5 −1 −0.5 0 0.5 1 1.5
Log(LUV/1040)
Log([OIII]/Hβ)
Best Fit: y = (−2.57 ± 0.66)x + (2.27 ± 1.51)
Fig. 10.— Top Panels: The classical Baldwin effect is displayed in the left panel and our
“substitute Baldwin relation” is shown in the right panel. The symbols for each 12 µm
Seyfert 1 are coded by the strength of [O III]/Hβratio (left) and LU V /1040 (right). The
strong segregation between the red and blue points in the left uppper panel indicates that
[O III]/Hβis a good predictor of EW(C IV). The classical Baldwin effect is weak, compared
to log(C IV EW) vs. log([O III]/Hβ) (right). The EW(CIV) is much better predicted from
log([O III]/Hβ). The LU V /1040 parameter hardly improves this. Bottom Panel: The strong
anticorrelation between log LU V /1040 and log([O III]/Hβ) suggests that the classical Baldwin
relation is a secondary effect, resulting from this correlation.
– 39 –
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
-1 -0.75 -0.5 -0.25 0 0.25 0.5 0.75 1 1.25 1.5
Log(MgII/[OII])
Log([OIII]/Hβ)
Sy 1
Sy 2
Fig. 11.— Plot of Mg II/[O II] vs. [O III]/Hβ. The best fitting line for the Sy 1 galaxies
(red triangles) is log(Mg II/[O II])=(−0.93 ±0.12)log([O III]/Hβ) + 0.84 ±0.32. We use
[O III]/Hβas a proxy for weak Fe II (Eigenvector 1), so an inverse correlation for the Sy 1’s
implies that Mg II and Fe II are correlated. The Seyfert 2 galaxies (blue squares) have no
BLR and therefore have much lower Mg II/[O II], and higher [O II]/Hβratios.
– 40 –
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(HX) erg s-1
H
α
: 1.17 ± 0.52
H
α
+[NII]: 1.06 ± 0.55
<y> intercept ± σ
Log(H
α
) Sy 1
Log(H
α
+[NII]) Sy 1
Log(H
α
) LINERs
Log(H
α
+[NII]) LINERs
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(NUV) erg s-1
H
α
: 1.76 ± 0.73
H
α
+[NII]: 1.62 ± 0.61
<y> intercept ± σ
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(2
µ
m) erg s-1
Log L(NUV) erg s-1
Log L(Emission Line) erg s-1
H
α
: 0.81 ± 0.85
H
α
+[NII]: 0.69 ± 0.88
<y> intercept ± σ
35 36 37 38 39 40 41 42 43 44 45 35
36
37
38
39
40
41
42
43
44
45
46
47
Log L(2
µ
m) erg s-1
Log L(10
µ
m) erg s-1
Log L(Emission Line) erg s-1
H
α
: 1.22 ± 1.11
H
α
+[NII]: 1.09 ± 1.08
<y> intercept ± σ
Fig. 12.— Correlation of non-stellar continuum luminosity with emission line luminos-
ity, either with Hα(filled in squares) or with Hα+[N II] (open squares), for Seyfert
1 galaxies. The filled and open triangles (in green) show observations of 12 µm Sy1’s
which were also classified as either LINERs or Starbursts. The plotted lines are fits of
log(LContinuum) = Alog(LEmission Line ) + Bwith a fixed slope of A= 1. The tightest correla-
tion is for the HX luminosities, with LHX ∼15LHα for the Sy 1’s and LHX ∼11.5L(Hα+[N II ]) ,
independent of Seyfert type. Those Sy 1 which have secondary classification as LINERs or
Starbursts are expected to harbor relative fainter Syfert nuclei. Nonetheless, they do not
deviate systematically from the correlation defined by the more AGN-dominated Sy 1’s
line/continuum.
– 41 –
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(HX) erg s-1
H
α
: 1.34 ± 1.31
H
α
+[NII]: 1.05 ± 1.33
<y> intercept ± σ
Log(H
α
) Sy 2
Log(H
α
+[NII]) Sy 2
Log(H
α
) LINERs
Log(H
α
+[NII]) LINERs
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(HX) erg s-1
Log L(NUV) erg s-1
H
α
: 2.71 ± 0.85
H
α
+[NII]: 2.42 ± 0.85
<y> intercept ± σ
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(2
µ
m) erg s-1
Log L(NUV) erg s-1
Log L(Emission Line) erg s-1
H
α
: 1.14 ± 1.08
H
α
+[NII]: 0.85 ± 1.09
<y> intercept ± σ
35 36 37 38 39 40 41 42 43 44 45 35
36
37
38
39
40
41
42
43
44
45
46
47
Log L(2
µ
m) erg s-1
Log L(10
µ
m) erg s-1
Log L(Emission Line) erg s-1
H
α
: 2.07 ± 1.18
H
α
+[NII]: 1.74 ± 1.26
<y> intercept ± σ
Fig. 13.— Same non-stellar continuum/emission line correlation as shown in Figure 12,
except the Seyfert 2 galaxies. The tightest correlation is for the HX luminosities, with
LHX ∼22LH α for the Sy 2’s and LH X ∼11.5L(Hα+[N II]) , independent of Seyfert type.
– 42 –
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(HX) erg s-1
[OIII]: 1.64 ± 0.63
[OIII]+H
β
: 1.34 ± 0.62
<y> intercept ± σ
Log([OIII]) Sy 1
Log([OIII]+H
β
) Sy 1
Log([OIII]) LINERs
Log([OIII]+H
β
) LINERs
Netzer (2005)
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(HX) erg s-1
Log L(NUV) erg s-1
[OIII]: 2.31 ± 0.61
[OIII]+H
β
: 2.00 ± 0.62
<y> intercept ± σ
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(2
µ
m) erg s-1
Log L(NUV) erg s-1
Log L(Emission Line) erg s-1
[OIII]: 1.55 ± 0.92
[OIII]+H
β
: 1.22 ± 0.84
<y> intercept ± σ
35 36 37 38 39 40 41 42 43 44 45 35
36
37
38
39
40
41
42
43
44
45
46
47
Log L(2
µ
m) erg s-1
Log L(10
µ
m) erg s-1
Log L(Emission Line) erg s-1
[OIII]: 2.16 ± 1.23
[OIII]+H
β
: 1.84 ± 1.17
<y> intercept ± σ
Fig. 14.— Same non-stellar continuum/emission line correlation as shown in Figure 12,
but for [O III] and [O III]+Hβon the x-axis, for Seyfert 1 galaxies. The plotted lines are
fits of log(LContinuum) = Alog(LEmission Line ) + Bwith a fixed slope of A= 1. The best
correlation is for the HX luminosities, with a single relation of LHX ∼25L([OI I I ]+H β), and
with LHX ∼44L[OIII ]. The non-linear relation between the LHX and L(Emission Line)found
by Netzer (2006) is indicated by the steep solid line in the upper left panel.
– 43 –
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(HX) erg s-1
[OIII]: 1.50 ± 1.01
[OIII]+H
β
: 1.43 ± 1.02
<y> intercept ± σ
Log([OIII]) Sy 2
Log([OIII]+H
β
) Sy 2
Log([OIII]) LINERs
Log([OIII]+H
β
) LINERs
Netzer (2005)
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(HX) erg s-1
Log L(NUV) erg s-1
[OIII]: 2.76 ± 1.05
[OIII]+H
β
: 2.64 ± 0.99
<y> intercept ± σ
35
36
37
38
39
40
41
42
43
44
45
46
47
35 36 37 38 39 40 41 42 43 44 45
Log L(2
µ
m) erg s-1
Log L(NUV) erg s-1
Log L(Emission Line) erg s-1
[OIII]: 1.32 ± 1.61
[OIII]+H
β
: 1.18 ± 1.33
<y> intercept ± σ
35 36 37 38 39 40 41 42 43 44 45 35
36
37
38
39
40
41
42
43
44
45
46
47
Log L(2
µ
m) erg s-1
Log L(10
µ
m) erg s-1
Log L(Emission Line) erg s-1
[OIII]: 2.07 ± 1.31
[OIII]+H
β
: 2.00 ± 1.13
<y> intercept ± σ
Fig. 15.— Same non-stellar continuum/emission line correlation as shown in Figure 12,
but for [O III] and [O III]+Hβ, for Seyfert 2 galaxies. The best correlation is for the HX
luminosities, with a single relation of LH X ∼25L([OII I ]+H β), and LHX ∼32L[OI I I]. The steep
solid line in the upper left panel is the same as in Figure 14.
– 44 –
Table 1. Seyfert 1 Galaxies
Name of Galaxy RA (J2000) DEC (J2000) CfAaTypebRedshift (z)
MKN 334 00h03m09.62s +21d57m36.6s ∗1.8 0.022
MKN 335 00h06m19.52s +20d12m10.5s 1.2 0.026
ESO 540-G1 00h34m13.82s -21d26m20.6s 1.8 0.027
ESO 12-G21 00h40m46.26s -79d14m24.2s 1 0.030
IRAS 00488+2907 00h51m35.01s +29d24m04.5s ∗1 0.036
I ZW 1 00h53m34.94s +12d41m36.2s 1/STBRST 0.061
NGC 526A 01h23m54.39s -35d03m55.9s ∗1.5 0.019
MKN 993 01h25m31.46s +32d08m11.4s ∗1.5 0.016
IRAS 01527+0622 01h55m22.04s +06d36m42.6s ∗1.9 0.017
MKN 590 02h14m33.56s -00d46m00.1s ∗1.2 0.026
MKN 1034 02h23m20.45s +32d11m34.2s 1 0.034
M-3-7-11 02h24m40.62s -19d08m31.3s 1.8 0.034
NGC 931 02h28m14.48s +31d18m42.0s 1.5 0.017
NGC 1097 02h46m19.05s -30d16m29.6s 1 0.004
NGC 1365 03h33m36.37s -36d08m25.4s 1.8 0.005
IRAS F03450+0055 03h47m40.19s +01d05m14.0s 1 0.031
NGC 1566 04h20m00.42s -54d56m16.1s 1 0.005
3C 120 04h33m11.10s +05d21m15.6s 1 0.033
MKN 618 04h36m22.24s -10d22m33.8s 1 0.036
M-5-13-17 05h19m35.80s -32d39m27.3s 1.5 0.012
IRAS F05563-3820 05h58m02.00s -38d20m04.7s 1 0.034
MKN 6 06h52m12.25s +74d25m37.5s 1.5 0.019
MKN 9 07h36m56.98s +58d46m13.4s 1.5 0.040
MKN 79 07h42m32.80s +49d48m34.7s 1.2 0.022
IRAS F07599+6508 08h04m33.08s +64d59m48.6s 1 0.148
NGC 2639 08h43m38.08s +50d12m20.0s 1.9 0.011
NGC 2782 09h14m05.11s +40d06m49.3s ∗1/STBRST 0.008
MKN 704 09h18m26.01s +16d18m19.2s 1.5 0.029
NGC 2841 09h22m02.63s +50d58m35.5s ∗1/LINER 0.002
UCG 5101 09h35m51.60s +61d21m11.5s 1.5/LINER 0.039
NGC 2992 09h45m42.05s -14d19m35.0s 1.9 0.008
MKN 1239 09h52m19.10s -01d36m43.5s 1.5/LINER 0.020
NGC 3031 09h55m33.17s +69d03m55.1s 1.8 -0.0001
MKN 1243 09h59m55.84s +13d02m37.8s ∗1 0.035
3C 234 10h01m49.52s +28d47m09.0s 1 0.185
NGC 3227 10h23m30.58s +19d51m54.2s 1.5 0.004
NGC 3511 11h03m23.77s -23d05m12.4s 1 0.004
– 45 –
Table 1—Continued
Name of Galaxy RA (J2000) DEC (J2000) CfAaTypebRedshift (z)
NGC 3516 11h06m47.49s +72d34m06.9s 1.5 0.009
MKN 744 11h39m42.55s +31d54m33.4s ∗1.8 0.009
NGC 4051 12h03m09.61s +44d31m52.8s 1.5 0.002
UCG 7064 12h04m43.32s +31d10m38.2s 1.9 0.025
NGC 4151 12h10m32.58s +39d24m20.6s 1.5 0.003
NGC 4235 12h17m09.88s +07d11m29.7s ∗1.5 0.008
NGC 4253 12h18m26.51s +29d48m46.3s 1.5 0.013
MKN 205 12h21m44.22s +75d18m38.8s ∗1 0.071
NGC 4395 12h25m48.86s +33d32m48.9s ∗1.8/LINER 0.001
3C 273 12h29m06.70s +02d03m08.6s 1 0.158
NGC 4565 12h36m20.78s +25d59m15.6s ∗1.9 0.004
NGC 4579 12h37m43.52s +11d49m05.5s 1.9/LINER 0.005
NGC 4593 12h39m39.43s -05d20m39.3s 1 0.009
NGC 4594 12h39m59.43s -11d37m23.0s 1.9/LINER 0.003
NGC 4602 12h40m36.85s -05d07m58.8s 1.9 0.008
M-2-33-34 12h52m12.46s -13d24m53.0s 1 0.015
MKN 231 12h56m14.23s +56d52m25.2s 1 0.042
NGC 5033 13h13m27.47s +36d35m38.2s 1.9 0.003
M-6-30-15 13h35m53.71s -34d17m43.9s 1.2 0.008
IRAS F13349+2438 13h37m18.73s +24d23m03.4s 1 0.108
IRAS 13354+3924 13h37m39.87s +39d09m17.0s ∗1.8 0.020
NGC 5252 13h38m15.96s +04d32m33.3s ∗1.9 0.023
NGC 5273 13h42m08.34s +35d39m15.2s ∗1.9 0.004
IC 4329A 13h49m19.27s -30d18m34.0s 1.2 0.016
MKN 279 13h53m03.45s +69d18m29.6s ∗1.5 0.030
NGC 5548 14h17m59.53s +25d08m12.4s 1.5 0.017
MKN 471 14h22m55.37s +32d51m02.7s ∗1.8 0.034
NGC 5674 14h33m52.24s +05d27m29.6s ∗1.9 0.025
MKN 817 14h36m22.07s +58d47m39.4s 1.5 0.031
MKN 841 15h04m01.20s +10d26m16.2s ∗1.5 0.036
IRAS F15091-2107 15h11m59.80s -21d19m01.7s 1 0.045
NGC 5905 15h15m23.32s +55d31m02.5s 1 0.011
CGCG 022-021 15h38m44.74s -03d22m48.2s ∗1.9 0.024
IRAS 16146+3549 16h16m30.69s +35d42m29.0s ∗1.5 0.028
ESO 141-G55 19h21m14.14s -58d40m13.1s 1 0.036
NGC 6860 20h08m46.89s -61d06m00.7s 1 0.015
MKN 509 20h44m09.74s -10d43m24.5s 1.2 0.034
– 46 –
Table 1—Continued
Name of Galaxy RA (J2000) DEC (J2000) CfAaTypebRedshift (z)
NGC 7213 22h09m16.31s -47d09m59.8s 1.5/LINER 0.006
ESO 344-G16 22h14m42.01s -38d48m22.9s 1.5 0.040
3C 445 22h23m49.53s -02d06m12.9s 1 0.056
NGC 7314 22h35m46.19s -26d03m01.7s 1.9 0.005
UGC 12138 22h40m17.05s +08d03m14.1s ∗1.8 0.025
NGC 7469 23h03m15.62s +08d52m26.4s 1.2 0.016
NGC 7603 23h18m56.62s +00d14m38.2s 1.5 0.030
aAn asterisk marks if this object was added from the CfA sample.
bAs listed under ‘BASIC DATA - Classifications’ in Nasa/Ipac Extragalactic Database
(NED), https://ned.ipac.caltech.edu.
Note. — This Table, as are all the data tables, are sorted by increasing RA starting
with the Seyfert 1’s then doing the same for the Seyfert 2’s.
– 47 –
Table 2. Seyfert 2 Galaxies
Name of Galaxy RA (J2000) DEC (J2000) CfAaTypebRedshift (z)
NGC 34 00h11m06.55s -12d06m26.3s 2 0.020
IRAS F00198-7926 00h21m53.61s -79d10m07.5s 2 0.073
NGC 262 00h48m47.14s +31d57m25.1s 2 0.015
IRAS F00521-7054 00h53m56.15s -70d38m04.2s 2 0.069
ESO 541-IG12 01h02m17.55s -19d40m08.7s 2 0.057
NGC 424 01h11m27.63s -38d05m00.5s 2 0.012
NGC 513 01h24m26.85s +33d47m58.0s 2 0.020
MKN 573 01h43m57.80s +02d20m59.7s ∗2 0.017
IRAS F01475-0740 01h50m02.70s -07d25m48.5s 2 0.018
NGC 833 02h09m20.84s -10d07m59.1s ∗2/LINER 0.013
NGC 839 02h09m42.93s -10d11m02.7s 2/LINER 0.013
UGC 2024 02h33m01.24s +00d25m15.0s 2 0.022
NGC 1052 02h41m04.80s -08d15m20.8s ∗2/LINER 0.005
NGC 1068 02h42m40.71s -00d00m47.8s 2 0.004
NGC 1125 02h51m40.27s -16d39m03.7s 2 0.011
NGC 1144/1143 02h55m12.2s -00d11m01s 2 0.029
M-2-8-39 03h00m30.59s -11d24m56.6s 2 0.030
NGC 1194 03h03m49.11s -01d06m13.5s 2 0.014
NGC 1241 03h11m14.64s -08d55m19.7s 2 0.014
NGC 1320 03h24m48.70s -03d02m32.2s 2 0.009
NGC 1386 03h36m46.18s -35d59m57.9s 2 0.003
IRAS F03362-1642 03h38m33.58s -16d32m15.6s 2 0.037
NGC 1433 03h42m01.55s -47d13m19.5s 2 0.004
ESO 420-G13 04h13m49.69s -32d00m25.1s ∗2 0.012
IRAS 04259-0440 04h28m26.05s -04d33m49.5s ∗2/LINER 0.016
NGC 1614 04h33m59.85s -08d34m44.0s ∗2/STBRST 0.016
NGC 1667 04h48m37.14s -06d19m11.9s ∗2 0.015
NGC 1672 04h45m42.50s -59d14m49.9s ∗2 0.004
ESO 033-G002 04h55m58.96s -75d32m28.2s 2 0.018
NGC 1808 05h07m42.34s -37d30m47.0s 2 0.003
IRAS F05189-2524 05h21m01.39s -25d21m45.4s 2 0.043
NGC 2655 08h55m37.73s +78d13m23.1s ∗2 0.005
NGC 2683 08h52m41.33s +33d25m18.3s 2/LINER 0.001
IRAS F08572+3915 09h00m25.39s +39d03m54.4s 2/LINER 0.058
NGC 3079 10h01m57.80s +55d40m47.2s ∗2/LINER 0.004
NGC 3147 10h16m53.65s +73d24m02.7s 2 0.009
NGC 3362 10h44m51.72s +06d35m48.2s ∗2 0.028
– 48 –
Table 2—Continued
Name of Galaxy RA (J2000) DEC (J2000) CfAaTypebRedshift (z)
NGC 3486 11h00m23.87s +28d58m30.5s 2 0.002
UCG 6100 11h01m34.00s +45d39m14.2s ∗2 0.030
NGC 3593 11h14m37.00s +12d49m03.6s ∗2/LINER 0.002
NGC 3627 11h20m14.96s +12d59m29.5s ∗2 0.002
M 0-29-23 11h21m12.26s -02d59m03.5s 2 0.025
NGC 3660 11h23m32.28s -08d39m30.8s 2 0.012
NGC 3735 11h35m57.30s +70d32m08.1s ∗2 0.009
NGC 3822 11h42m11.11s +10d16m40.0s 2 0.020
NGC 3976 11h55m57.29s +06d44m58.0s ∗2 0.008
NGC 3982 11h56m28.13s +55d07m30.9s 2 0.004
NGC 4303 12h21m54.90s +04d28m25.1s 2 0.005
NGC 4388 12h25m46.75s +12d39m43.5s 2 0.008
IC 3639 12h40m52.85s -36d45m21.1s 2 0.011
NGC 4628 12h42m25.26s -06d58m15.6s ∗2 0.009
NGC 4826 12h56m43.64s +21d40m58.7s 2 0.001
NGC 4922 13h01m24.90s +29d18m40.0s 2/LINER 0.024
NGC 4941 13h04m13.14s -05d33m05.8s 2 0.004
NGC 4968 13h07m05.98s -23d40m37.3s 2 0.010
NGC 5005 13h10m56.23s +37d03m33.1s 2/LINER 0.003
M-3-34-64 13h22m24.46s -16d43m42.5s 2 0.017
NGC 5135 13h25m44.06s -29d50m01.2s 2 0.014
NGC 5194 13h29m52.71s +47d11m42.6s 2 0.002
NGC 5248 13h37m32.02s +08d53m06.6s 2 0.004
NGC 5256 13h38m17.50s +48d16m37.0s ∗2/STBRST 0.028
MKN 270 13h41m05.76s +67d40m20.3s ∗2 0.010
NGC 5278 13h41m39.62s +55d40m14.3s 2 0.025
MKN 273 13h44m42.11s +55d53m12.7s 2/LINER 0.038
MKN 796 13h46m49.45s +14d24m01.7s ∗2 0.022
MKN 1361 13h47m04.36s +11d06m22.6s 2 0.023
MKN 461 13h47m17.75s +34d08m55.7s ∗2 0.016
NGC 5347 13h53m17.83s +33d29m27.0s 2 0.008
MKN 463 13h56m02.87s +18d22m19.5s 2 0.050
NGC 5395 13h58m37.98s +37d25m28.1s ∗2 0.012
NGC 5506 14h13m14.89s -03d12m27.3s 2 0.006
MKN 686 14h37m22.12s +36d34m04.1s ∗2 0.014
NGC 5899 15h15m03.22s +42d02m59.4s ∗2 0.009
NGC 5929 15h26m06.16s +41d40m14.4s ∗2 0.008
– 49 –
Table 2—Continued
Name of Galaxy RA (J2000) DEC (J2000) CfAaTypebRedshift (z)
NGC 5953 15h34m32.38s +15d11m37.6s 2/LINER 0.007
UGC 9913 15h34m57.12s +23d30m11.5s 2/LINER 0.018
UGC 9944 15h35m47.86s +73d27m02.5s ∗2 0.025
NGC 5995 15h48m24.95s -13d45m28.0s 2 0.025
IRAS F15480-0344 15h50m41.50s -03d53m18.3s 2 0.030
NGC 6217 16h32m39.20s +78d11m53.4s 2 0.005
NGC 6240 16h52m58.87s +02d24m03.3s 2/LINER 0.024
NGC 6552 18h00m07.23s +66d36m54.4s ∗2 0.026
NGC 6810 19h43m34.25s -58d39m20.1s 2 0.007
ESO 339-G11 19h57m37.58s -37d56m08.3s 2 0.019
NGC 6890 20h18m18.10s -44d48m24.2s 2 0.008
IC 5063 20h52m02.34s -57d04m07.6s 2 0.011
MKN 897 21h07m45.82s +03d52m40.4s 2 0.026
NGC 7130 21h48m19.52s -34d57m04.5s 2/LINER 0.016
NGC 7172 22h02m01.89s -31d52m10.8s ∗2 0.009
IRAS F22017+0319 22h04m19.17s +03d33m50.2s 2 0.061
IC 5169 22h10m09.98s -36d05m19.0s 2 0.010
M-3-58-07 22h49m37.15s -19d16m26.4s 2 0.031
NGC 7479 23h04m56.65s +12d19m22.4s ∗2/LINER 0.008
NGC 7496 23h09m47.29s -43d25m40.6s 2 0.006
ESO 148-IG2 23h15m46.75s -59d03m15.6s 2/STBRST 0.045
IC 5298 23h16m00.70s +25d33m24.1s ∗2/STBRST 0.027
NGC 7582 23h18m23.50s -42d22m14.0s 2 0.005
NGC 7590 23h18m54.81s -42d14m20.6s 2 0.005
NGC 7674 23h27m56.72s +08d46m44.5s 2 0.029
NGC 7678 23h28m27.90s +22d25m16.3s ∗2/STBRST 0.012
NGC 7682 23h29m03.93s +03d32m00.0s ∗2 0.017
NGC 7733 23h42m32.95s -65d57m23.4s ∗2 0.034
CGCG 381-051 23h48m41.72s 02d14m23.1s 2 0.031
MKN 331 23h51m26.80s +20d35m09.9s ∗2 0.018
aAn asterisk marks if this object was added from the CfA sample.
bAs listed under ‘BASIC DATA - Classifications’ in Nasa/Ipac Extragalactic Database
(NED), https://ned.ipac.caltech.edu.
Note. — This Table, as are all the data tables, are sorted by increasing RA starting with
– 50 –
the Seyfert 1’s then doing the same for the Seyfert 2’s.
– 51 –
Table 3. Emission Lines
Emission Line Wavelength Ionization Potential Ionization Potential Critical Density
(˚
A) Lower (eV) Upper (eV) (cm−3)
Lyman α1216 ··· 13.6 ···
N V 1240, 1243 77.5 97.9 ···
C IV 1549, 1551 47.9 64.5 ···
C III] 1907, 1909 24.4 47.9 1×1010
Mg II 2796, 2804 7.6 15.0 ···
[Ne V] 3426 97.1 126.2 1.6×107
[O II] 3727, 3729 13.6 35.1 4.5×103
[Ne III] 3869 41.0 63.5 9.7×106
Hβ4861 ··· 13.6 ···
[O III] 4959 35.1 54.9 7.0×105
[O III] 5007 35.1 54.9 7.0×105
[O I] 6300 ··· 13.6 1.8×106
[N II] 6549, 6583 14.5 29.6 8.7×104
Hα6563 ··· 13.6 ···
[S II] 6717, 6734 10.4 23.3 1.5×103, 3.9×103
Note. — Emission lines and their corresponding wavelength. The ionization potentials and
critical densities are from De Robertis & Osterbrock (1986), Morton (2003), Osterbrock & Ferland
(2006), and SPECTR-W3 (http://spectr-w3.snz.ru/ion.phtml).
– 52 –
Table 4. Optical Emission Line Data
Name [NeV]a[OII] [NeIII] Hβ[OIII] 4959˚
A [OIII] 5007˚
A [OI] Hα[NII] Hα+ [NII] [SII] Ref
Seyfert 1
MKN 334 0.71 7.03 1.23 4.48 2.38 7.87 1.29 28.35 16.1 ··· 7.42 1,2
MKN 335 6.3 3 11 91 6.3 22.45 ··· ··· ··· 264.8 ··· 1,3
E 540-G1 ··· 1.95 ··· 2.71 ··· 2.59 0.4 9.69 4.72 ··· 2.25 4,5
E12-G21 · · · · · · · · · 14.8 ··· 9.95 ··· 111 ··· ··· ··· 6,7
IRAS 00488+2907 ··· ··· ··· ··· ··· ··· ··· ··· ··· ··· ···
I ZW 1 6.3 2.7 4.3 31 4.6 7.43 ··· 74.8 11 164.8 ··· 1,8
NGC 526A ··· 11 3.41 2.15 8.4 24.7 1.6 6.44 4.54 ··· 3.62 3,7
MKN 993 · · · · · · · · · 1.13 1.02 5.05 0.51 7.71 3.97 ··· 1.81 2,9
IRAS 01527+0622 ··· ··· ··· ··· ··· ··· ··· ··· ··· ··· ··· 2
MKN 590 · · · · · · · · · 6.02 ··· 5.25 ··· ··· ··· 170 ··· 3,10
MKN 1034 ··· 0.65 0.21 0.46 0.81 1.1 0.66 ··· ··· 13.8 2.5 1,11
M-3-7-11 · · · · · · · · · 4.5 8.9 25 1.37 18.35 2.56 ··· 4.8 6,12
NGC 931 1.17 0.26 0.65 11.8 3.85 12.13 0.48 ··· ··· 67.1 2.09 1,3
NGC 1097 ··· 11.1 ··· 10.1 ··· 1.75 ··· 42.4 25.61 ··· 10.83 6,13,14
NGC 1365 ··· 16.6 ··· 6··· 18 ··· 143.1 ··· 130 ··· 6,10,15
F03450+0055 · · · · · · · · · 21.27 ··· 6.86 0.92 ··· ··· 114.25 1.42 1,16
NGC 1566 ··· 7.64 3.07 16.6 7.28 18.2 4.19 56.3 12 ··· 5.82 7,25
3C 120 · · · · · · 3.5 30.4 13.4 38.2 1.5 208.9 10 133.67 ··· 1,6,10,18
MKN 618 3.59 1.49 2.18 27.24 4.8 14.05 0.74 ··· ··· 118.15 2.07 1,3,10,16
M-5-13-17 ··· 5.43 6.6 20.4 14.7 36.45 3.88 81.45 10.51 ··· 3.1 7,8
F05563-3820 ··· 0.32 0.65 13.48 ··· 7.01 ··· 122 ··· 190 ··· 6,10,19
MKN 6 1.25 9.74 5.45 30 24 76 7 216 ··· ··· 22 1,25,20
MKN 9 5.4 0.65 2.6 26.65 3.3 10.05 ··· ··· ··· 84.6 ··· 1,3
MKN 79 3.22 7.89 4.04 68.5 10 35.1 ··· 208.8 ··· 261.7 ··· 1,3,25
F07599+6508 · · · · · · · · · 1.1 ··· 0.33 ··· 37 ··· ··· ··· 1,21
NGC 2639 0.68 3.48 ··· 0.7 1.09 2.34 1.17 2.96 10.71 ··· 5.23 1,22
NGC 2782 ··· 11.7 1.02 6.07 2.92 6.56 1.97 36.8 15.2 ··· 15.6 1,8,22
– 53 –
Table 4—Continued
Name [NeV]a[OII] [NeIII] Hβ[OIII] 4959˚
A [OIII] 5007˚
A [OI] Hα[NII] Hα+ [NII] [SII] Ref
MKN 704 3.77 1.41 2.77 28.65 ··· 14.06 ··· 226.95 ··· ··· ··· 1,3,6,25
NGC 2841 · · · · · · · · · 0.59 ··· 1.09 0.33 1.94 3.57 ··· 2.2 22
UGC 5101 ··· 0.81 ··· 0.2 0.24 0.45 0.29 3.87 4.91 ··· 1.56 1,23,30
NGC 2992 ··· 11.4 2.46 2.7 9.13 29.3 3.11 19.7 16.55 ··· 11.66 1,7,25
MKN 1239 6.27 3.61 5.56 22.05 7.92 27.5 1.93 ··· ··· 158.25 6 1,3,10
NGC 3031 1.09 5.43 5.27 19.8 8.2 22.6 23.5 113.8 41.8 ··· 25.8 26
MKN 1243 · · · · · · · · · 1.31 ··· 7.01 ··· ··· ··· ··· ··· 9
3C 234 0.85 1.41 ··· 2.12 ··· 15.7 0.21 ··· ··· ··· ··· 1
NGC 3227 1.8 14 3.2 22.8 19 61.8 10.8 144.4 62.2 167.6 31.8 1,3,22
NGC 3511 · · · · · · · · · 1.43 ··· 0.72 ··· 7.02 2.53 ··· 3.51 27
NGC 3516 4.2 2.4 3.5 87.33 14 36.31 ··· 264.45 16.4 324.85 ··· 1,3,8,18,22
MKN 744 ··· ··· ··· ··· ··· ··· ··· ··· ··· ··· ···
NGC 4051 7.7 21 7.1 24.8 14 39.45 ··· 80.4 20.7 140 11.3 1,3,22
UGC 7064 1.22 2.48 1.41 1.11 4.49 11.54 0.97 4.31 4.53 ··· 1.24 1
NGC 4151 96 210 110 502.9 1000 1264.1 ··· 1384.7 220 2064.5 ··· 1,3,18
NGC 4235 · · · · · · · · · 3.13 ··· 2.01 1.02 28.7 3.09 ··· 2.33 3,17,22
NGC 4253 3.7 3.82 3.46 16.8 11.8 51.75 1.32 92.8 ··· 87.6 3.52 1,3
MKN 205 1.2 0.55 0.96 25.9 1.1 3.8 ··· ··· ··· 74 ··· 1
NGC 4395 · · · · · · · · · 2.3 ··· 14.4 1.91 9.85 2.16 ··· 4.7 17,22
3C 273 · · · · · · · · · 15.9 ··· ··· ··· ··· ··· ··· ··· 1
NGC 4565 · · · · · · · · · 0.17 ··· 1.46 0.26 0.77 2.08 ···