A Study of the Soft X-Ray Emission Lines in NGC 4151. I. Kinematic Properties of the Plasma Wind

NGC 4151 (z = 0.003262) is a type 1.5 Seyfert active galactic nucleus (AGN) with a black hole (BH) mass of M_BH = 3.59 × 10⁷ M_⊙ (Bentz & Katz 2015). While NGC 4151 is highly variable in the hard X-ray band (e.g., Beuchert et al. 2017), the soft X-ray spectrum is very similar to that of NGC 1068 with multiple strong emission features on top of an almost negligible continuum (e.g., Grafton-Waters et al. 2021). This extreme contrast between soft and hard X-ray bands makes NGC 4151 an ideal target to study the properties of the narrow line region (NLR). The NLR is part of the outflowing wind, composed of separate plasma regions, that produces the emission lines observed in the X-ray spectra (see Figure 1). The aim of this ORBYTS research-with-schools public engagement was to assess whether the NLR properties changed between observations spanning a 15 yr period, by measuring the outflow velocities and distances of each region. The code can be found at Grafton-Waters (2021).

Zoom In Zoom Out Reset image size

NGC 4151 was observed 27 times by XMM-Newton during six separate epochs between 2000 and 2015. Here we analyzed the reflection grating spectrometer (RGS; den Herder et al. 2001) data from each observation. The data were reduced using the RGSPROC command in the SAS software v 17.0.0. ³ Any large background counts in CCD9 were removed before combining RGS1 and RGS2 data with RGSCOMBINE.

We modeled the thirteen strongest emission lines (labeled in Figure 1) that were present in all X-ray spectra with a simple Gaussian model in Python, defined as

$$\begin{eqnarray}&&G(x)=\displaystyle \frac{A}{\sqrt{2\pi {\sigma }^{2}}}\exp \left[-\displaystyle \frac{{\left(x-\mu \right)}^{2}}{2{\sigma }^{2}}\right],\end{eqnarray} \tag{ 1 }$$

where A is the amplitude, σ is the standard deviation, and μ is the mean value (the emission line center). To measure the properties of each emission line we defined a wavelength range (λ_R ) to ensure we were only fitting one line at a time. This is shown in the insert in Figure 1(a). This Gaussian model and the X-ray spectrum were then fed into the LMFIT interface in Python (Newville et al. 2014), which returned the parameter values (μ, σ, and A) and errors.

To estimate the outflow velocity (v_out) of the plasma regions, we used the redshift (z) equation, given by

$$\begin{eqnarray}&&z=\displaystyle \frac{{v}_{\mathrm{out}}}{c}=\displaystyle \frac{{\lambda }_{\mathrm{obs}}-{\lambda }_{\mathrm{rest}}}{{\lambda }_{\mathrm{rest}}},\end{eqnarray} \tag{ 2 }$$

where c is the speed of light, and λ_obs and λ_rest are the observed and rest frame wavelengths, respectively. The velocity shift was measured for each emission line individually. The observed wavelength (μ) was measured from our modeling and the rest wavelengths were obtained from the SPEX line list. ⁴ We note that it is not the emission line that has the velocity, but instead it is a property of the line-emitting plasma region. If many of the emission lines had similar values they were likely to originate from the same plasma region within the outflowing wind. To estimate the distances (R) from the BH we assumed that v_out was greater than or equal to the escape velocity (v_esc) of the BH, given by

$$\begin{eqnarray}&&R=\displaystyle \frac{2{{GM}}_{\mathrm{BH}}}{{v}_{\mathrm{esc}}^{2}},\end{eqnarray} \tag{ 3 }$$

where G is the gravitational constant, and M_BH is the BH mass.

After modeling each emission line, and obtaining the velocities and distances from each observation, we compared results. This allowed us to evaluate whether the wind properties changed over time.

Figure 1(a) shows the 2000 RGS spectrum with the Gaussian models fitted to the strongest emission lines; these lines are labeled. The insert in Figure 1(a) displays the O vii forbidden line at 22.15 Å with the Gaussian model fitted on top. From the spectral modeling of all the observations, we compare the results for the velocities and distances in 2000 (left) and 2015 (right) in Figures 1(b)–(c) and (d)–(e), respectively. The black dashed lines in Figure 1 indicate the plasma boundaries for v_out and R, suggesting that there are at least two to three plasma regions. However, we cannot rule out the possibility of more. The results for 2000 and 2015 are consistent with the velocities and distance from all observations.

Our aim was to see whether the outflowing wind properties changed over the course of 15 yr. However, based on this investigation, we found no evidence of this. One may expect that from 2000 to 2015, the distances and velocities would have increased and decreased, respectively, as the plasma regions move away from the BH. However, assuming the plasma is traveling at 500 km s⁻¹, the distance traveled in a 15 yr time period would only be 0.008 pc, which is insignificant compared to the distances that we measured (i.e., >1 pc). One explanation could be that we are viewing the plasma perpendicular to the outflow direction, meaning we are unable to observe the true outflow velocity, since the Doppler shifting of lines would only represent a component of its true outflow velocity. Alternatively, the plasma has been relatively static over the 15 yr, which is reasonable given the large distances from the BH. In that case, 15 yr may simply not be enough time for us to observe any significant changes with regards to the outflowing wind in NGC 4151.

We did have some problems when modeling the data, which could have affected the final results. For example, the line center (μ) depended on λ_R which we set when modeling the lines. This meant that if λ_R was too large, then multiple lines would be fitted together by the model, or if λ_R was too small we would lose information about the feature. This was a particular problem if λ_R for an emission line differed between observations, especially if the S/N ratios were poor. As a result, a slightly different velocity was obtained each time.

From our RGS analysis on NGC 4151, we found that there are at least two to three plasma regions that produce the emission features in the outflowing wind. However, the velocities and distances are consistent throughout the observations. This implies that either the plasma regions have been relatively static over the 15 yr observation period, or because we are seeing the outflowing wind perpendicular to the flow direction, we are not measuring the velocities in the direction of motion.

This work was undertaken with the ORBYTS Research-with-schools public engagement project, partnering scientists with schools to support students' involvement in space research. S.G.W. acknowledges the support of a PhD studentship awarded by the UK Science & Technology Facilities Council (STFC).