Unobscured Central Broad-line Regions in Type-1.9 AGN SDSS J1241+2602

Both broad emission lines from central broad emission-line regions (BLRs) and narrow emission lines from extended narrow emission-line regions (NLRs) are fundamental spectroscopic characteristics in the optical band of Type-1 active galactic nuclei (broad emission-line AGNs) (Osterbrock & Mathews 1986; Sulentic et al. 2000; Oh et al. 2015). Meanwhile, strong narrow emission lines from central NLRs but no apparent broad emission lines are fundamental spectroscopic characteristics of Type-2 AGNs (narrow emission-line AGNs). The well-known AGN unified model (Antonucci 1993; Netzer 2015; Balokovic et al. 2018; Kuraszkiewicz et al. 2021; Zhang 2022a) has been hypothesized to explain the different spectroscopic phenomena between Type-1 and Type-2 AGNs, after mainly considering severe obscurations of central BLRs in Type-2 AGNs. In the framework of the AGN unified model, Type-2 and Type-1 AGNs have intrinsically similar fundamental structures of central accretion disks around black holes (BHs), BLRs, dust torus, and NLRs, but Type-2 AGNs have central accretion disks around BHs and BLRs that are heavily obscured by the central dust torus due to its orientation with respect to the line of sight. The AGN unified model has been strongly supported by clearly detected polarized broad emission lines and/or clearly detected broad infrared emission lines in some Type-2 AGNs (Tran 2003; Savic et al. 2018; Moran et al. 2020).

However, even considering different properties expected of both the central dust torus and the central BH accretion process, some challenges to the AGN unified model have been reported. Franceschini et al. (2002) have discussed probably different evolutionary patterns in Type-1 and Type-2 AGNs. Villarroel & Korn (2014) have reported different neighbors around Type-1 and Type-2 AGNs. Zou et al. (2019) have reported lower host galaxy stellar masses in X-ray-selected Type-1 AGNs than in Type-2 AGNs. Bornancini & Garcia Lambas (2020) have shown significantly different properties of UV/optical and mid-infrared color distributions of different AGN types. More recently, Zhang (2022b) has shown that direct measurements of stellar velocity dispersion can lead to statistically larger stellar velocity dispersions in Type-1 AGNs than in Type-2 AGNs, with a confidence level higher than 10σ, even after considering the necessary effects of different redshifts and different physical properties related to the central BH accretion processes in AGNs. As discussed in Netzer (2015), the AGN unified model has been successfully applied to explain different observed spectroscopic features between Type-1 and Type-2 AGNs in many different ways; however, the AGN family with many other features considering the reported challenges to the AGN unified model is far from homogeneous.

Besides Type-1 and Type-2 AGNs explained by the AGN unified model, there is a special kind of optically selected AGN, a Type-1.9 AGN (first discussed in Osterbrock 1981), which has apparent broad Hα but no apparent broad Hβ. Commonly, the disappearance of broad Hβ (or quite large broad Balmer decrements, i.e., a large flux ratio of broad Hα to broad Hβ) in Type-1.9 AGNs is mainly attributed heavily obscured central BLRs, and can be applied to test the AGN unified model. However, as discussed in Canfield & Puetter (1981), Kwan & Krolik (1981), and Goodrich (1990), BLRs modeled with relatively low optical depths and low ionization parameters can reproduce large broad Balmer decrements in Type-1.9 AGNs, indicating there are rare Type-1.9 AGNs that have central BLRs with large broad Balmer decrements that are intrinsic but not due to serious obscuration. Barcons et al. (2003) have discussed that H1320+551 (a Type-1.9 AGN) with no apparent broad Hβ but apparent and strong broad Hα is not consistent with being an obscured Type-1 AGN, through its unabsorbed X-ray properties. More recently, Hernandez-Garcia et al. (2017) have discussed that Type-1.9 and Type-2 AGNs have different different variability properties in the UV and X-ray domains, indicating that pure obscurations of central regions should be disfavored to explain different features between Type-1.9 AGNs and Type-1/2 AGNs. In this paper, the interesting Type-1.9 AGN, SDSS J124131.46+260233.57 (= SDSS J1241+2602), is first reported for its unobscured central BLRs with strong evidence from optical spectroscopic results.

The paper is organized as follows. Section 2 shows the main hypotheses. Section 3 presents the spectroscopic results of SDSS J1241+2602 at redshift 0.0159. Section 4 describes our necessary discussions. Section 5 gives our final conclusions. We have adopted the cosmological parameters H₀ = 70 km s⁻¹ Mpc⁻¹, Ω_Λ = 0.7, and Ω_m = 0.3.

In order to test heavily obscured central BLRs in a Type-1.9 AGN, the properties of virial BH mass can be applied as follows.

Adopting the virialization assumption regarding central BLRs as discussed in Vestergaard (2002), Peterson et al. (2004), and Shen et al. (2011), the virial BH mass of a broad-line AGN can be conveniently estimated as

$$\begin{eqnarray}&&\displaystyle \frac{{M}_{{\rm{BH}}}}{1.2131\times {10}^{-3}\,{M}_{\odot }}=5.5\times \displaystyle \frac{\tfrac{{R}_{{\rm{BLRs}}}}{100\,\text{light-days}}\times {\left(\tfrac{\sigma }{1000\,{\rm{km}}\,{{\rm{s}}}^{-1}}\right)}^{2}}{G},\end{eqnarray} \tag{ 1 }$$

with G = 6.672 × 10⁻¹¹ N m² kg^–2 as the gravitational constant, R_BLRs in units of 100 light-days as the distance of BLRs from the central BH, and σ in units of km s⁻¹ as the line width (second moment) of broad emission lines to trace rotational velocities of broad-line emission clouds in BLRs. The factor 5.5 is the virial factor, discussed in Onken et al. (2004), Woo et al. (2010, 2015), Graham et al. (2011), and Park et al. (2012). R_BLRs can be simply estimated from the continuum luminosity using the improved empirical relation in Bentz et al. (2013) after necessary corrections for host galaxy contaminations. Moreover, considering the strong linear correlation between continuum luminosity and broad Hα luminosity as discussed in Greene & Ho (2005) and Mejia-Restrepo et al. (2022), the virial BH mass of a broad-line AGN can be estimated from the line width (second moment, σ_Hα ) and line luminosity (L_Hα ) of broad Hα as

$$\begin{eqnarray}&&{M}_{{\rm{BH}}}=15.6\times {10}^{6}{\left(\displaystyle \frac{{L}_{{\rm{H}}\alpha }}{{10}^{42}\,{\rm{erg}}\,{{\rm{s}}}^{-1}}\right)}^{0.55}{\left(\displaystyle \frac{{\sigma }_{{\rm{H}}\alpha }}{1000\,{\rm{km}}\,{{\rm{s}}}^{-1}}\right)}^{2.06}\,{M}_{\odot }\end{eqnarray} \tag{ 2 }$$

in order to ignore the effects of uncertainties of the measured continuum luminosities in a broad-line AGN with strong host galaxy contributions. Here, the second moment rather than the FWHM of the broad Hα is applied, mainly due to some effects of sharp features around the peak of the broad-line profile on the calculated second moment.

Meanwhile, through measured stellar velocity dispersions of host galaxy stellar bulges, the M_BH–σ relation discussed in Ferrarese & Merritt (2000), Gebhardt et al. (2000), Kormendy & Ho (2013), Batiste et al. (2017), and Bennert et al. (2021) can also be conveniently applied to estimate central BH mass in both quiescent galaxies and active galaxies, without the effects of obscurations on central BLRs. If there were intrinsically serious obscurations of central BLRs leading to the disappearance of broad Hβ in a Type-1.9 AGN, the virial BH mass of the Type-1.9 AGN obtained through properties of observed broad Hα should be significantly smaller than the value expected from the M_BH–σ relation, which is the main point for testing obscured/unobscured BLRs in a Type-1.9 AGN. In the next section, the stellar velocity dispersion of SDSS J1241+2602 can be measured through apparent absorption features, leading to the BH mass of SDSS J1241+2602 being determined by the measured M_BH–σ relation. Therefore, in SDSS J1241+2602, interesting results are reported and discussed in the following sections on properties of the BH mass determined by the M_BH–σ relation and on properties of the virial BH mass with and without considerations of obscurations of central BLRs.

SDSS J1241+2602 has its Sloan Digital Sky Survey (SDSS) spectrum (plate-mjd-fiberid = 2660-54504-0446) with signal-to-noise ratio of about 56 shown in Figure 1 with apparent broad Hα and apparent stellar absorption features. In order to measure the emission lines, the commonly accepted simple stellar population (SSP) method is applied to determine host galaxy contributions. More detailed descriptions of the SSP method can be found in Bruzual & Charlot (2003), Kauffmann et al. (2003), Cid Fernandes et al. (2005), and Cappellari (2017). The SSP method has been applied in our previous papers Zhang (2021a, 2021b, 2021c, 2022a, 2022b). Here, we show simple descriptions of the SSP method as follows. The 39 SSP templates from Bruzual & Charlot (2003) and Kauffmann et al. (2003) have been exploited, which can be used to describe the characteristics of almost all the SDSS galaxies. Meanwhile, there is an additional fifth-order polynomial component applied to describe intrinsic AGN continuum emissions. Here, as was shown for properties of the composite spectrum of SDSS quasars in Vanden Berk et al. (2001), AGN continuum emissions can be fitted by two power laws with a break at 5000 Å, indicating that a simple power-law component is not preferred. Moreover, higher-order polynomial functions are also checked, leading to no variability of the subsequently determined χ²/dof where χ² is the sum of squared residuals and dof denotes degrees of freedom. Therefore, a fifth-order polynomial component is preferred in this paper. When the SSP method is applied, optical narrow emission lines are masked out by the full width at zero intensity of about 450 km s⁻¹, and the spectrum with rest wavelength ranging from 6250 to 6750 Å is also masked out due to the strongly broad Hα. Then, through the Levenberg–Marquardt least-squares minimization technique (the known MPFIT package), the SDSS spectrum with emission lines masked out can be described and shown in Figure 1 with corresponding χ²/dof ∼ 0.96 and with the determined stellar velocity dispersion of about 117 ± 3 km s⁻¹ and determined continuum luminosity at rest wavelength 5100 Å (from the determined fifth-order polynomial component) of about (1.09 ± 0.09) × 10⁴² erg s⁻¹.

Zoom In Zoom Out Reset image size

Before proceeding further, one point is noted. The stellar velocity dispersion reported by the SDSS pipeline is about 144 ± 3 km s⁻¹ in SDSS J1241+2602, without considering AGN continuum emissions in the pipeline. If the fifth-order polynomial component is not considered, then a similar procedure applied to describe the SDSS spectrum with only narrow emission lines masked out can lead to a determination of the stellar velocity dispersion as 146 ± 3 km s⁻¹, which is consistent with the value reported by the SDSS pipeline. However, considering the apparent broad Hα in SDSS J1241+2602, the fifth-order polynomial component should be preferred.

Moreover, the Ca ii triplet (Ca T) from 8300 Å to 8800 Å, the Ca ii H + K absorption features from 3750 Å to 4200 Å, and the Mg i absorption features from 5050 Å to 5250 Å are applied to remeasure the stellar velocity dispersions of SDSS J1241+2602 through the same SSP method as discussed above in order to describe the whole SDSS spectrum. The best-fitting results are shown in the bottom panels of Figure 1 with the determined stellar velocity dispersions in units of km s⁻¹ being about 116 ± 7, 99 ± 4, and 110 ± 8 for the Ca T, Ca ii H + K, and Mg i absorption features, respectively. Therefore, in this paper, the mean value σ_⋆ = 110 ± 12km s⁻¹ is accepted as the stellar velocity dispersion of SDSS J1241+2602.

After subtractions of the host galaxy contributions and the AGN continuum emissions, emission lines in the line spectrum can be measured, similarly to what we have previously done in Zhang (2021b, 2021c, 2022a, 2022b, 2022c). For the emission lines within the rest wavelength range from 4750 Å to 5100 Å, there is one Gaussian function applied to describe narrow Hβ and four Gaussian functions applied to describe the [O iii] λ λ4959, 5007 doublet (two for the core components and two for the blueshifted wings). When the Gaussian functions above are applied, only two criteria are accepted. First, each Gaussian component has line intensity not smaller than zero. Second, the core (blueshifted) components of the [O iii] doublet have the same redshift and the same line width, and have their flux ratio fixed to the theoretical value 3. Then, through the Levenberg–Marquardt least-squares minimization technique, the best-fitting results to the emission lines and the corresponding residuals (calculated from the line spectrum minus the best-fitting results then divided by the uncertainties of the SDSS spectrum) are shown in the left panels of Figure 2 with ${\chi }_{0}^{2}/{\mathrm{dof}}_{0}=209.8/344\sim 0.61$. Besides the discussed Gaussian components, we tried to apply one additional broad Gaussian function to describe probable broad Hβ; however, the fitting procedure led to the additional broad component having a determined line flux and line width smaller than their corresponding determined uncertainties. Therefore, it is not necessary to consider broad Gaussian components to describe the broad Hβ in SDSS J1241+2602.

Zoom In Zoom Out Reset image size

Meanwhile, emission lines within the rest wavelength range from 6250 Å to 6800 Å can also be measured using multiple Gaussian functions. There is one Gaussian function applied to describe the narrow Hα, three broad Gaussian functions applied to describe the broad Hα, two Gaussian functions applied to describe the [N ii] doublet, two Gaussian functions applied to describe the [O i] doublet, and four Gaussian functions applied to describe the [S ii] doublet (two for the core components and two for the shifted wings). When the Gaussian functions above are applied, only three criteria are accepted. First, each Gaussian component has line intensity not smaller than zero. Second, the components of the [N ii] (the [O i], the [S ii]) doublet have the same redshift and the same line width, and the [N ii] doublet has its flux ratio fixed to the theoretical value 3. Third, the components in the narrow Hα and in the narrow Hβ have the same redshift and the same line width. Then, through the Levenberg–Marquardt least-squares minimization technique, the best-fitting results and the corresponding residuals are shown in the right panels of Figure 2 with χ²/dof = 160.6/398 ∼ 0.41. Besides the Gaussian components discussed above, we tried to apply additional Gaussian functions to describe probable blue-/redshifted wings of the [O i] and the [N ii] doublets; however, the fitting procedure led to the determined line fluxes of the additional Gaussian components being smaller than their corresponding determined uncertainties. Therefore, it is not necessary to consider additional blue-/redshifted wings in the [O i] and the [N ii] doublets in SDSS J1241+2602.

Before proceeding further, another point is noted. If different numbers (not three) of broad Gaussian functions were applied to describe the broad Hα of SDSS J1241+2602, were there different results for the line profiles? In order to test the effects of applying different numbers of broad Gaussian functions, the F-test is used, similar to what we have recently done in Zhang (2022c). For one, two, and four broad Gaussian functions applied to describe broad Hα of SDSS J1241+2602, the corresponding χ²/dof are about ${\chi }_{1}^{2}/{\mathrm{dof}}_{1}=238.8/404\sim 0.59$, ${\chi }_{2}^{2}/{\mathrm{dof}}_{2}=224.2/401\sim 0.56$, and ${\chi }_{4}^{2}/{\mathrm{dof}}_{4}=160.6/395\sim 0.406$, respectively. Then, based on the different χ² and dof for different model functions, the F-test technique can be applied to confirm that the confidence level is higher than 5σ to support the notion that three broad Gaussian functions describe the broad Hα better than one or two, and the F-test technique can be applied to confirm that the probability is only about 10⁻⁵ that four broad Gaussian functions describe the broad Hα better than three. Therefore, in the manuscript, three broad Gaussian functions are applied to describe the broad Hα of SDSS J1241+2602.

Based on the measured line parameters listed in Table 1, the observed line luminosities are (3.39 ± 0.11) × 10³⁹ erg s⁻¹ and (1.18 ± 0.02) × 10⁴⁰ erg s⁻¹ for the narrow Hβ and the narrow Hα, respectively, leading to a normal flux ratio of the narrow Hα to the narrow Hβ of 3.5. The line width is about 114 ± 2 km s⁻¹ for the narrow emission lines, consistent with the measured stellar velocity dispersion. The line luminosity and the line width (second moment) of the broad Hα (sum of the three broad Gaussian components) are about (5.69 ± 0.91) × 10⁴⁰ erg s⁻¹ and about 3150 ± 420 km s⁻¹, respectively. The uncertainties above are determined through the determined uncertainties of the Gaussian emission components. The results can be applied to confirm that SDSS J1241+2602 is a Type-1.9 AGN with apparent broad Hα but no broad Hβ.

Based on the calculated line width (second moment) and line luminosity of the broad Hα without considering any obscuration of the central BLRs, the virial BH mass in SDSS J1241+2602 can be estimated as (3.43 ± 1.25) × 10⁷ M_⊙ through Equation (2) above. The uncertainty of the virial BH mass is estimated from the uncertainties of the line width and the line luminosity of the broad Hα. Meanwhile, a similar virial BH mass of about 2.5 × 10⁷ M_⊙ in SDSS J1241+2602 has been reported in Liu et al. (2019) and Martin-Navarro et al. (2022). It is interesting that the virial BH mass is consistent with the BH mass expected from the M_BH–σ relation. The dependence of virial BH mass on stellar velocity dispersion in SDSS J1241+2602 is shown in Figure 3 with results reported previously in the literature for quiescent galaxies, active galaxies, and also tidal disruption events (TDEs), etc. The corresponding reference can be found in the caption of Figure 3.

Zoom In Zoom Out Reset image size

Before proceeding further, it is interesting to check whether a broad component related to shifted wings of narrow Balmer lines can also lead to so large a virial BH mass. If it is assumed that the summed component of the three Gaussian components listed in Table 1 for broad Hα is the shifted-wing-related extended component in narrow Hα, the line width (second moment) and line flux can be determined through the summed profile of the three Gaussian components. Then, comparing with the parameters of the Gaussian component in the narrow Hα listed in Table 1 as the core component, the line width ratio R_σ and the line flux ratio R_flux are about 27.6 and 4.8 of the shifted-wing-related extended component to the core component of the narrow Hα in SDSS J1241+2602. If it is simply assumed that similar properties of the broad components are related to the shifted wings of the narrow Hα as those discussed in blueshifted components in [O iii] lines of the 557 blue quasars in Zhang (2021b) with distributions of line width ratio R_σ and flux ratio R_flux of the broad component to the core component of [O iii] λ5007 shown in Figure 4, the probabilities are about 1.54 × 10⁻¹⁰ for $\mathrm{log}({R}_{\sigma })\gt \mathrm{log}(27.6)$ and about 1.21% for $\mathrm{log}({R}_{\mathrm{flux}})\gt \mathrm{log}(4.8)$. Therefore, the probability is only about 1.54 × 10⁻¹⁰ × 1.21% ∼ 1.86 × 10⁻¹² for a shifted-wing-related broad component in SDSS J1241+2602, leading to the similar virial BH mass. In other words, the confidence level is higher than 5σ to support the notion that the broad components in Balmer lines are not components related to shifted wings of narrow Balmer lines but really related to central BLRs in SDSS J1241+2602.

Zoom In Zoom Out Reset image size

However, if it is accepted that the disappearance of broad Hβ was due to serious obscuration of central BLRs in SDSS J1241+2602, the upper limits of line flux of the broad Hβ can be simply estimated through the F-test technique as follows, and then a different virial BH mass can be estimated. Assuming that the obscured broad Hβ overwhelmed in spectral noise has the same line profile as that ([λ(Hα_b ), f_λ (Hα_b )]) of broad Hα described by three Gaussian functions but has total line flux determined by the line flux of broad Hα divided by the expected broad Balmer decrement BD_b , then the intrinsic but obscured broad Hβ can be described by

$$\begin{eqnarray}&&[\lambda ({\rm{H}}{\beta }_{b}),\,{f}_{\lambda }({\rm{H}}{\beta }_{b})]=\left[\lambda ({\rm{H}}{\alpha }_{b})\times \displaystyle \frac{4862.68\,\mathring{{\rm{A}}}}{6564.61\,\mathring{{\rm{A}}}},\,\displaystyle \frac{\left.{f}_{\lambda }({\rm{H}}{\alpha }_{b}\right)}{{{\rm{BD}}}_{b}}\right]\end{eqnarray} \tag{ 3 }$$

For the observed line spectrum shown in the top left panel of Figure 2, contributions of obscured broad Hβ can be considered, and a new line spectrum can be created as

$$\begin{eqnarray}&&[\lambda ({\rm{H}}{\beta }_{b}),\,{f}_{\lambda }]\,=\,[\lambda ({\rm{H}}{\beta }_{b}),\,{f}_{\lambda ,{\rm{obs}}}+{f}_{\lambda }({\rm{H}}{\beta }_{b})].\end{eqnarray} \tag{ 4 }$$

Then, for a series of 1000 new line spectra including different contributions of obscured broad Hβ with different values of BD_b (larger than 3), χ² values can be calculated as

$$\begin{eqnarray}&&{\chi }_{i}^{2}\,=\,\sum \,{\left(\displaystyle \frac{{f}_{\lambda }-{Y}_{{\rm{fit}}}}{{y}_{{\rm{err}}}}\right)}^{2}\,(i=1,\,\ldots ,\,1000)\end{eqnarray} \tag{ 5 }$$

with Y_fit as the best-fitting results shown in the top left panel of Figure 2 and y_err as uncertainties of the SDSS spectrum. Then, through the F-test statistical technique applied with 1 (only one additional parameter of BD_b ) and 343 (dof_0-1) as number of dofs for the numerator and denominator of the F-distribution, the calculated 3σ and 5σ confidence levels ${\rm{CL}}\sim \left(\frac{{\chi }^{2}-{\chi }_{0}^{2}}{1}\right)/\left(\tfrac{{\chi }^{2}}{343}\right)$ for strong broad Hβ can lead ${\chi }^{2}-{\chi }_{0}^{2}$ to be 5 and 16, respectively. Here, ${\chi }_{0}^{2}$ and dof₀ are for the best-fitting results shown in top left panel of Figure 2.

The dependence of ${\chi }^{2}-{\chi }_{0}^{2}$ on BD_b is shown in Figure 5, leading to BD_b ∼ 160 and BD_b ∼ 64 for the obscured broad Hβ with 3σ and 5σ confidence levels, which leads E(B − V) to be around 3.4 and 2.6, assuming the intrinsic flux ratio of broad Hα to broad Hβ to be 3.1. Therefore, considering serious obscurations of the central BLRs with E(B − V) ∼ 3.4 (E(B − V) ∼ 2.6), the reddening-corrected line luminosity of the broad Hα should be about 1605 (283) times higher than the value from the observed broad Hα, which leads the reddening-corrected virial BH mass to be about 58 (22) times higher than the value from properties of the observed broad Hα. The corrected virial BH masses are shown as solid/open five-point stars in purple in Figure 3, and are significantly larger than the values expected from the M_BH–σ relation.

Zoom In Zoom Out Reset image size

In order to show clearer results in Figure 3, the 89 quiescent galaxies from Savorgnan & Graham (2015), the 29 reverberation-mapped (RM) AGNs from Woo et al. (2015), and the 12 TDEs from Zhou et al. (2021) are considered to draw the linear correlation between stellar velocity dispersion and BH mass:

$$\begin{eqnarray}&&\begin{array}{l}\mathrm{log}\left(\displaystyle \frac{{M}_{\mathrm{BH}}}{{M}_{\odot }}\right)=(-2.89\pm 0.49)+(4.83\pm 0.22)\\ \,\,\,\times \,\mathrm{log}\left(\displaystyle \frac{{\sigma }_{\star }}{\mathrm{km}\,{{\rm{s}}}^{-1}}\right)\end{array}\end{eqnarray} \tag{ 6 }$$

through the least trimmed squares robust technique (Cappellari et al. 2013). Then the 3σ and 5σ confidence bands to the linear correlation are determined and shown in Figure 3. Therefore, the reddening-corrected viral BH mass should cause SDSS J1241+2602 to be an outlier with confidence levels higher than 5σ.

Before we end this section, another point is noted. As shown in Section 3, a large stellar velocity dispersion of ∼146 km s ⁻¹ can be estimated in SDSS J1241+2602, if not considering AGN continuum emissions that are apparently included in the SDSS spectrum. It is necessary to discuss whether the larger stellar velocity dispersions can affect our final results shown in Figure 3. Here, we consider the question via the following two points. On the one hand, the F-test technique can be applied to confirm a confidence level higher than 5σ for the AGN continuum emissions described by the fifth-order polynomial function included in model functions to describe the SDSS spectrum of SDSS J1241+2602, based on χ²/dof = 3272.56/3382 ∼ 0.96 and χ²/dof = 3972.38/3387 ∼ 1.17 for the best descriptions to the SDSS spectrum with and without considering AGN continuum emissions as determined by the SSP method. Therefore, the stellar velocity dispersion of ∼110 km s⁻¹ is preferred in SDSS J1241+2602, considering contributions of AGN continuum emissions to the SDSS spectrum. On the other hand, even without considering AGN continuum emissions, the properties of SDSS J1241+2602 with the larger stellar velocity dispersion of about 146 ± 3 km s ⁻¹ are also shown in Figure 3 as triangles in different colors, to again support the notion that the reddening-corrected viral BH mass can also lead SDSS J1241+2602 to be a unique outlier in the space of viral BH mass versus stellar velocity dispersion. Therefore, different stellar velocity dispersions with and without considering AGN continuum emissions have few effects on our final conclusions.

Considering the BH mass expected from the M_BH–σ relation (no effects from obscuration of the central BLRs), heavily obscured central BLRs should be disfavored in SDSS J1241+2602, indicating unobscured BLRs in the Type-1.9 AGN SDSS J1241+2602 with apparent broad Hα but no broad Hβ. Besides SDSS J1241+2602 discussed in this paper, H1320+551 is the other individual Type-1.9 AGN reported in the literature (Barcons et al. 2003) with unobscured BLRs. Unfortunately, there is no clear information on stellar velocity dispersion in H1320+551, leading to no further discussions on BH mass determined through different methods as discussed in this paper. But in the near future, it will be interesting to check the virial BH masses of a large sample of Type-1.9 AGNs, to test whether unobscured BLRs are common in them, and then to provide clues to the evolution of different types of AGN under the framework of the AGN unified model.

Based on the stellar velocity dispersion measured through the absorption features in the Type-1.9 AGN SDSS J1241+2602 with apparent broad Hα but no broad Hβ, the BH mass expected from the M_BH–σ relation is consistent with the virial BH mass through the observed broad Hα without considering any obscurations of the central BLRs. Meanwhile, if considering serious obscurations of central BLRs to explain the disappearance of broad Hβ in the Type-1.9 AGN SDSS J1241+2602, the reddening-corrected broad Hα line luminosity should lead to SDSS J1241+2602 having a recalculated reddening-corrected virial BH mass that makes it an outlier in the M_BH–σ space with confidence level higher than 5σ. Based on the properties of virial BH mass, unobscured central BLRs are favored in the Type-1.9 AGN SDSS J1241+2602. The results indicate that obscured/unobscured BLRs of Type-1.9 AGN should be first discussed when using the properties of these AGNs to test the AGN unified model.

Zhang gratefully acknowledges the anonymous referee for constructive comments and suggestions that greatly improved the paper. Zhang gratefully acknowledges the research funding support from GuangXi University and the kind funding support from NSFC-12173020 and NSFC-12373014. This research has made use of the data from the SDSS (https://www.sdss.org/) funded by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation and the U.S. Department of Energy Office of Science. The research has made use of the MPFIT package https://pages.physics.wisc.edu/~craigm/idl/cmpfit.html to solve the least-squares problem through the Levenberg–Marquardt technique, and of the LTS_LINEFIT package https://www-astro.physics.ox.ac.uk/~cappellari/software/ to do linear fitting through least trimmed squares robust technique.