Upload
bennett-davis
View
214
Download
0
Embed Size (px)
Citation preview
Powerful Outflows In Seyfert Galaxies: Wind- vs. Jet-Driven ModelsPowerful Outflows In Seyfert Galaxies: Wind- vs. Jet-Driven ModelsDrake University Department of Physics & AstronomyDrake University Department of Physics & Astronomy
Kory Kreimeyer, Deanna Berget, Julie Leifeld, Jordan Mirocha, Charles NelsonKory Kreimeyer, Deanna Berget, Julie Leifeld, Jordan Mirocha, Charles Nelson
While the black hole accretion disk model applies for all AGN, the unique While the black hole accretion disk model applies for all AGN, the unique appearance of Seyfert galaxies is due to the presence of two additional areas appearance of Seyfert galaxies is due to the presence of two additional areas within the nucleus. The innermost region surrounding the black hole is known as within the nucleus. The innermost region surrounding the black hole is known as the broad line region (BLR), from which we observe broad emission lines due to the broad line region (BLR), from which we observe broad emission lines due to the large velocity dispersion near the continuum source. This area is surrounded the large velocity dispersion near the continuum source. This area is surrounded by a dense torus of gas and dust, which obscures our view of the BLR if seen by a dense torus of gas and dust, which obscures our view of the BLR if seen edge on. Even if this is the case, the narrow line region (NLR) will still be edge on. Even if this is the case, the narrow line region (NLR) will still be visible. This region consists of gas clouds lying in the path of the nuclear visible. This region consists of gas clouds lying in the path of the nuclear outflows, perpendicular to the plane of the accretion disk and the BLR. In type I outflows, perpendicular to the plane of the accretion disk and the BLR. In type I Seyfert’s both of these regions are visible, while in type II’s only the NLR is Seyfert’s both of these regions are visible, while in type II’s only the NLR is observable. Type I and type II Seyfert galaxies are thought to be the same observable. Type I and type II Seyfert galaxies are thought to be the same phenomena, with the only difference being orientation with respect to earth’s line phenomena, with the only difference being orientation with respect to earth’s line of sight (Figure 1).of sight (Figure 1).
Background
The differences in the two models can be characterized in the following way. The jet models show line-splitting into blue The differences in the two models can be characterized in the following way. The jet models show line-splitting into blue and red shifted components at all points along the slit, since we are seeing emission from both approaching and receding and red shifted components at all points along the slit, since we are seeing emission from both approaching and receding gas at any given location. The wind models show blue shifts towards the top of the slit and red shifts toward the lower part gas at any given location. The wind models show blue shifts towards the top of the slit and red shifts toward the lower part of the slit, as expected for radial outflow in a cone slightly inclined to the line of sight. Although correspondence between of the slit, as expected for radial outflow in a cone slightly inclined to the line of sight. Although correspondence between the observed spectrum and our simple models is poor, the overall velocity gradient in the data is best matched by the wind the observed spectrum and our simple models is poor, the overall velocity gradient in the data is best matched by the wind models with constant velocity. However, there does seem to be some line-splitting in the data, albeit from fainter emitting models with constant velocity. However, there does seem to be some line-splitting in the data, albeit from fainter emitting regions. This indicates that though the wind may dominate, both wind and jet forces may be influencing the gas dynamics regions. This indicates that though the wind may dominate, both wind and jet forces may be influencing the gas dynamics of the NLR. More detailed modeling, involving varying the intensity, the influence of dust and filling factor, will reveal of the NLR. More detailed modeling, involving varying the intensity, the influence of dust and filling factor, will reveal more information about the ionized gas flow.more information about the ionized gas flow.
Discussion
Our analysis of the gas cloud kinematics in NGC 1068 has provided evidence of powerful outflows within Our analysis of the gas cloud kinematics in NGC 1068 has provided evidence of powerful outflows within the galaxy’s nuclear regions. In addition to revealing the magnitude of nuclear outflows, the STIS spectra the galaxy’s nuclear regions. In addition to revealing the magnitude of nuclear outflows, the STIS spectra provide detailed information about the narrow line region’s velocity distribution. Our models attempt to provide detailed information about the narrow line region’s velocity distribution. Our models attempt to recreate these high velocity outflows by defining a perfectly conical or perfectly cylindrical velocity field, recreate these high velocity outflows by defining a perfectly conical or perfectly cylindrical velocity field, representing a powerful galactic wind or jet-gas interaction, respectively. Although wind models seem to be representing a powerful galactic wind or jet-gas interaction, respectively. Although wind models seem to be a better match to the data, there is evidence that both processes are at least partially responsible for the a better match to the data, there is evidence that both processes are at least partially responsible for the observed kinematics in the narrow line region. To test if one mechanism truly dominates, a detailed mapping observed kinematics in the narrow line region. To test if one mechanism truly dominates, a detailed mapping of the gas cloud velocity and brightness distribution in the region will be required. In collaboration with Dr. of the gas cloud velocity and brightness distribution in the region will be required. In collaboration with Dr. Tim Urness, we plan to use computer data visualization techniques to facilitate this analysis.Tim Urness, we plan to use computer data visualization techniques to facilitate this analysis.
Summary
ObservationsLong slit spectroscopy is the primary data collection technique for the study of Long slit spectroscopy is the primary data collection technique for the study of active galactic nuclei. Our spectra were taken with the Space Telescope Imaging active galactic nuclei. Our spectra were taken with the Space Telescope Imaging Spectrograph (STIS) which is one of the instruments onboard Hubble Space Spectrograph (STIS) which is one of the instruments onboard Hubble Space Telescope (HST). For NGC 1068, five slit positions were placed over the central Telescope (HST). For NGC 1068, five slit positions were placed over the central region of the galaxy, positioned strategically in order to capture as many features region of the galaxy, positioned strategically in order to capture as many features of the narrow line region as possible (Figure 2). Each slit preserves the vertical of the narrow line region as possible (Figure 2). Each slit preserves the vertical spatial dimension, while providing spectral data for each point along its length spatial dimension, while providing spectral data for each point along its length (Figure 3). Three prominent AGN emission lines lie within this spectral range: (Figure 3). Three prominent AGN emission lines lie within this spectral range: one hydrogen Balmer line, Hβ (4861Å), and two lines from doubly-ionized one hydrogen Balmer line, Hβ (4861Å), and two lines from doubly-ionized oxygen, [OIII] (4959 Å) and [OIII] (5007 Å). In the vertical dimension, each oxygen, [OIII] (4959 Å) and [OIII] (5007 Å). In the vertical dimension, each pixel corresponds to 0.05 arcseconds and so provides good spatial resolution. In pixel corresponds to 0.05 arcseconds and so provides good spatial resolution. In the end there are five data files for NGC 1068, each with spectral and positional the end there are five data files for NGC 1068, each with spectral and positional data of clouds in the galaxy’s narrow line region. data of clouds in the galaxy’s narrow line region.
NGC 1068 is a spiral NGC 1068 is a spiral galaxy located in the galaxy located in the constellation Cetus, and is constellation Cetus, and is the prototypical type II the prototypical type II Seyfert galaxy (narrow Seyfert galaxy (narrow emission lines only). It emission lines only). It lies at a distance of about lies at a distance of about 45 million light years, 45 million light years, which allows for high which allows for high resolution study of the resolution study of the central regions. At this central regions. At this distance, one arcsecond distance, one arcsecond corresponds to about 220 corresponds to about 220 light years.light years.
NGC 1068
Fig. 1 – Seyfert Galaxy ModelFig. 1 – Seyfert Galaxy Model
Abstract
To analyze the HST data, we used the software package IRAF (Image Reduction and Analysis Facility), To analyze the HST data, we used the software package IRAF (Image Reduction and Analysis Facility), which is widely used for data analysis in astronomical research. The spectra consist of multiple image which is widely used for data analysis in astronomical research. The spectra consist of multiple image files, each providing spectral and positional data for one of the slits (Figure 3). Although the Hfiles, each providing spectral and positional data for one of the slits (Figure 3). Although the Hββ and 4959 and 4959 lines are weaker, they share the same velocity structure, seen as the varying horizontal displacement of the lines are weaker, they share the same velocity structure, seen as the varying horizontal displacement of the features at different positions along the slit. Rightward indicates a redward shift, or recessional velocity, features at different positions along the slit. Rightward indicates a redward shift, or recessional velocity, which can be observed below the center of the image for all three emission lines. Above the center, all which can be observed below the center of the image for all three emission lines. Above the center, all emission lines have changed to exhibit large blue shifts. The narrow vertical emission above the main emission lines have changed to exhibit large blue shifts. The narrow vertical emission above the main structure is not associated with the NLR but arises from gas in the disk of the host galaxy. Using the structure is not associated with the NLR but arises from gas in the disk of the host galaxy. Using the “splot” task in IRAF, we produced plots of wavelength vs. flux (not shown) for each position along the slit. “splot” task in IRAF, we produced plots of wavelength vs. flux (not shown) for each position along the slit. Gaussian fitting and statistical analysis of the [OIII] lines then allowed for the determination of gas cloud Gaussian fitting and statistical analysis of the [OIII] lines then allowed for the determination of gas cloud velocities at each point along the slit.velocities at each point along the slit.
Analysis Kinematical Modeling
We have developed a computer model to produce theoretical images in an identical form to We have developed a computer model to produce theoretical images in an identical form to the observed spectrum (Figure 3), but based on an idealized velocity field. As mentioned the observed spectrum (Figure 3), but based on an idealized velocity field. As mentioned previously, the two competing outflow theories are a wind-driven conical system and a jet-previously, the two competing outflow theories are a wind-driven conical system and a jet-driven cylindrical system, and these two mechanisms are the primary study of the model. driven cylindrical system, and these two mechanisms are the primary study of the model. By defining a perfectly conical or perfectly cylindrical outflowing velocity field, it is By defining a perfectly conical or perfectly cylindrical outflowing velocity field, it is possible to observe the associated features of each system. In addition, the model allows possible to observe the associated features of each system. In addition, the model allows for the variation of a number of parameters, including inclination toward observer, cone for the variation of a number of parameters, including inclination toward observer, cone opening angle, maximum outflow velocity, and velocity-distance relation. To create the opening angle, maximum outflow velocity, and velocity-distance relation. To create the pictures below, we used input parameters obtained from previous studies. The opening pictures below, we used input parameters obtained from previous studies. The opening angle used was 65º, from Evans et al. (1991). Also, our inclination angle is 15º toward angle used was 65º, from Evans et al. (1991). Also, our inclination angle is 15º toward Earth, as given by Cecil et al. (1990). Earth, as given by Cecil et al. (1990).
Two wind-driven conical outflow models are presented below. Figure 4 shows a Two wind-driven conical outflow models are presented below. Figure 4 shows a constant conical outflow, while Figure 5 depicts a velocity fall-off with increasing distance constant conical outflow, while Figure 5 depicts a velocity fall-off with increasing distance from the center. Similarly, the lower two pictures show jet-driven cylindrical outflow from the center. Similarly, the lower two pictures show jet-driven cylindrical outflow models. Figure 6 is a cylindrical constant velocity field, and Figure 7 involves a decreasing models. Figure 6 is a cylindrical constant velocity field, and Figure 7 involves a decreasing velocity with increasing distance from the jet axis.velocity with increasing distance from the jet axis.
Note that the model produces an image of only the brightest emission line ([OIII] Note that the model produces an image of only the brightest emission line ([OIII] 5007 Å). The other lines exhibit identical velocity structure and could be creating by 5007 Å). The other lines exhibit identical velocity structure and could be creating by simply shifting and scaling the 5007 Å line.simply shifting and scaling the 5007 Å line.
Seyfert galaxies are active galaxies characterized by luminous outflows from a Seyfert galaxies are active galaxies characterized by luminous outflows from a central super massive black hole. The outflow mechanism has not been fully central super massive black hole. The outflow mechanism has not been fully determined, but the two primary competing systems are a wind-driven conical determined, but the two primary competing systems are a wind-driven conical outflow and a jet-driven cylindrical outflow. We are developing models with a outflow and a jet-driven cylindrical outflow. We are developing models with a variety of parameters, including inclination, cone opening angle, etc. By varying variety of parameters, including inclination, cone opening angle, etc. By varying these parameters, we can add detail to the model to more completely match the these parameters, we can add detail to the model to more completely match the observed velocities. We compare the models to the Hubble Space Telescope observed velocities. We compare the models to the Hubble Space Telescope long-slit spectra of the Seyfert galaxy NGC 1068 in an effort to more fully long-slit spectra of the Seyfert galaxy NGC 1068 in an effort to more fully understand the actual outflow mechanism of these galaxies.understand the actual outflow mechanism of these galaxies.
Understanding The ModelThe model begins by laying down a 3-dimensional grid of volume elements called “voxels.” At each of these points in The model begins by laying down a 3-dimensional grid of volume elements called “voxels.” At each of these points in space, it determines the magnitude and direction of the velocity vector caused by the particular choice of velocity field, space, it determines the magnitude and direction of the velocity vector caused by the particular choice of velocity field, either conical or cylindrical. It also assigns an intensity and a velocity spread to this point. The velocity vector then is either conical or cylindrical. It also assigns an intensity and a velocity spread to this point. The velocity vector then is inclined to the line of sight, and the final radial velocity is used as the center of a velocity-intensity Gaussian distribution. inclined to the line of sight, and the final radial velocity is used as the center of a velocity-intensity Gaussian distribution. Each voxel has a corresponding Gaussian distribution, but to simulate the actual observations, all depth positions along the Each voxel has a corresponding Gaussian distribution, but to simulate the actual observations, all depth positions along the same line have to be added together. This allows for multiple components of velocity within one line, something seen same line have to be added together. This allows for multiple components of velocity within one line, something seen frequently in the actual data.frequently in the actual data.
It is important to note that the model does not allow for a partially filled NLR. Any point that lies within the desired It is important to note that the model does not allow for a partially filled NLR. Any point that lies within the desired spatial region is given an intensity; however, we know the NLR is not completely filled with gas and does contain regions spatial region is given an intensity; however, we know the NLR is not completely filled with gas and does contain regions of empty (and therefore dark) space. Obscuration by dust clouds within the emitting region produces a similar effect. This of empty (and therefore dark) space. Obscuration by dust clouds within the emitting region produces a similar effect. This was an acceptable omission at this time, since our main interest is in determining what observations was an acceptable omission at this time, since our main interest is in determining what observations couldcould be explained by a be explained by a conical or cylindrical velocity field.conical or cylindrical velocity field.
Fig. 4 – Wind model with constant Fig. 4 – Wind model with constant velocitiesvelocities
Fig. 6 – Jet model with constant velocitiesFig. 6 – Jet model with constant velocities Fig. 7 – Jet model with velocities decreasing Fig. 7 – Jet model with velocities decreasing as ras r-1/2-1/2
Fig. 5 – Wind model with velocities Fig. 5 – Wind model with velocities decreasing as rdecreasing as r-1/2-1/2
Fig. 2 – Narrow Line Region of NGC 1068 shown Fig. 2 – Narrow Line Region of NGC 1068 shown in [OIII].in [OIII].
Fig. 3 – STIS spectrum: from left to right, Hβ (4861Å), Fig. 3 – STIS spectrum: from left to right, Hβ (4861Å), [OIII] (4959 Å), [OIII] (5007 Å)[OIII] (4959 Å), [OIII] (5007 Å)