Neutral Hydrogen in the Interacting Pair NGC 4618/4625
This work has been submitted to AJ! (Bush & Wilcots 2003) In the process of preparing this work for publication, much of it has been improved upon or even redone. I've resigned myself to the fact that this website won't be truly updated until this semester is over, so look for updates by the end of January.
Introduction and Objectives
Over the years galaxy interactions and mergers have come to play a very important role in our understanding of galaxy evolution. In order to explore exactly how interactions and mergers contribute to evolution we need to answer several questions. How common are interactions? Over what timescales do interactions take place? How strongly can an interaction deform a galaxy?
Observations of neutral hydrogen (HI) in galaxies can play an important role in answering these questions. The pair of galaxies I am studying, NGC 4618 and 4625, look like unassociated small Magellanic spirals when viewed optically. But when their HI is observed, NGC 4618 shows tidal arms reaching to four times its optical radius, and NGC 4625 reveals a gas disk 9 times its optical radius! By studying the HI distributions, hopefully we can learn about how interactions have affected this galaxy pair. Along the way we hope to determine:
- Masses (neutral hydrogen and dynamical)
- Rotation curves
- Interaction history
NGC 4618/4625
Before getting too deep into the neutral hydrogen data, we should look at the work that has been done in the optical. Figure 1 is an optical image from the Digitized Sky Survey [1] .
Fig. 1 An optical image from the Digitized Sky Survey.
Notice that in the optical, the galaxies look totally unrelated. NGC 4625 is the galaxy in the upper left, and NGC 4618 is the galaxy in the lower right. Odewahn 1991 [2] does optical analysis of the galaxies and we will use many of his values for comparison. From his measured luminosity, assuming a mass to light ratio for stars of 1 M sun / 1 L sun , we adopt a stellar mass value of 2.37 x 10 9 M sun for NGC 4618 and 0.42 x 10 9 M sun for NGC 4625. He gives effective radii (the radius at which one half the galaxies' light has been emitted) of 1.5 kpc and 0.7 kpc for Ngc 4618 and 4625 respectively. We also use his distance estimates to adopt a distance of 6.0 Mpc.
Background and Methodology
In order to "see" neutral hydrogen we observe radio emission at a 21 cm wavelength. This is the wavelength of light emitted when the electron in a neutral hydrogen atom undergoes a "spin flip." However, if this light is emitted by HI with a certain velocity, the emission is Doppler shifted slightly away from exactly 21 cm, broadening the line. (Most introductory college astronomy texts will cover the 21 cm line and line broadening.) By observing the broadened line, and chopping it up into small wavelength windows within the line (called "channels"), we get what we call an data cube. There are three data axis, right ascension, declination and frequency, which is directly related to the velocity of the emitting gas. Also, if there is more gas in the part of the galaxy you are looking at (known as a higher column density of gas), there will be more hydrogen atoms undergoing a spin flip and therefore more emission. So the radio data gives us two important pieces of information: how the neutral hydrogen gas is distributed in the galaxy and how it is moving in that distribution.
Data
Our data was taken using the VLA in the C configuration. (Technical details: Our frequency coverage was 1.416418150 GHz to 1.419201353 GHz with a velocity resolution of 5.17 km/s. Our beam size was 374.3 arcsec, and we had a one sigma sensitivity of 2 x 10 18 cm -1 .) We reduced and calibrated the data using the Astronomical Image Processing System (AIPS). The details are not horribly exciting unless you are a radio astronomer, so I won't go into them here. For information on how radio data is taken, reduced and calibrated see An Introduction to Radio Astronomy [3] for introductory information and Synthesis Imaging in Radio Astronomy II [4] for more technical information. The virtual radio interferometer is a good way to get a feel for how radio data reduction works.
Once the data has been reduced, there are two ways of viewing the data cube. The first is to look at the channels, the hydrogen gas at a given velocity, in succession, creating a movie. Figure 2 is one section of the movie of my different channels in still frames.
Fig. 2 This is a section of channels going from a velocity of around 620 km/s in the upper left to 590 km/s in the lower right. NGC 4625 is apparent in all the frames, while 4618 shows up around frame 33. Since the different frames show gas at a certain velocity, the fact that the gas appears to move across the page is simply a velocity gradient caused by rotation. (Gas rotating towards us has a lower velocity, gas rotating away from us has a higher velocity).
The other method is to add these frames together to create an image of the entire HI distribution. This is done by integrating the flux coming from the galaxy along the velocity axis. It is called the moment zero map because you integrate the intensity in each channel times the velocity to the 0 power (we'll come back to this later). Figure 3 contains the moment 0 map. Figure 4 shows contours of the moment 0 map in these galaxies overlaid on the optical image of the galaxies from the Digitized Sky Survey (figure 1).
Fig. 3 The total HI map.
Fig. 4 The total HI extent overlaid on the optical image from the Digitized Sky Survey.
There are several remarkable things about this image. As mentioned before, each of the galaxies have HI that extends far beyond the galaxy. Also, the arcs on NGC 4618 are tidal arms: a clear indication that these galaxies are interacting and affecting each other's evolution. Also, just to the upper right of NGC 4625 is a thin streamer of material that may be a bridge between the two galaxies, pulled off one of them in an interaction. That said, the disk of NGC 4625 looks strangely unperturbed. We would like to understand each of these features.
The data cube used so far was created to emphasize large scale structure of the HI (natural weighting of baselines). Next we wanted to look at smaller scale structure.This data cube was created and figure 5 shows the contours from this map overlaid on a WIYN image of NGC 4618. (For the radio astronomers out there: using only uniform weighting of baselines lowered our signal to noise too much, so we chose to make the map halfway between natural and uniform weighting, robustness 0.)
Fig. 5 NGC 4618 HI extent overlaid on the optical image from WIYN. The contours are at 4, 10, 15, 20, 25, 30, and 35 times the noise level, which was approximately 21 Jy*km/s.
Analysis
First we began to look more closely at the HI distributions. By measuring the radii at three sigma contours, we get a radius of 232 ± 5 arcsec for NGC 4625. For NGC 4518,we calculate a radius of 147 ± 5 arcsec for the major axis of disk, and tidal arms extending out to 282 ± 5 arcsec. We can also look at how intensity varies with radius. By looking at the intensity in small rings and graphing this as a function of radius, we get a qualitative map of the column density of HI as a function of radius. These maps are figures 6 and 7. By taking the total flux from the galaxy and correcting for the distance to the galaxy (something farther away will look dimmer, and therefore lead us to account for less flux) we can find the total HI mass of the galaxy. We calculated the total HI mass for 4618 to be (5.3 ± 0.6)x10 8 M sun and for 4625 we found (3.9 ± 0.6)x10 8 M sun. We also calculated the HI mass of the entire system, and found (9.4 ± 0.6)x10 8 M sun, which is about the sum of the other two galaxies. It is also interesting to know how much mass is in the tidal arms on NGC 4618. The west arm has about (5.7 ± 0.6)x10 7 M sun, and the east arm has about (4.1 ± 0.6)x 10 7 M sun. All of the mass calculations are to 4 sigma column density sensitivity.
Fig. 6 The intensity in a thin ring on the NGC 4618 as a function of radius.
Fig. 7 The intensity in a thin ring on the NGC 4625 as a function of radius.
As the next step in our analysis, we wanted to look at how the gas is moving. To do this, we created a velocity field. The velocity field is created by making a moment 1 map: integrating the intensity in each channel times the velocity to the first power over all the velocities. This effectively gives the velocity of the gas weighted by the amount of gas that is going that velocity. The moment 1 map is in figure 8.
Fig. 8 The velocity field.
The velocity gradient across the galaxies indicates that they are rotating. However, NGC 4625 appears to be rotating quite cleanly while NGC 4618 has been perturbed quite a bit. From this we would like to be able to derive the rotation curve of the galaxy: the velocity of the gas as a function of radius. AIPS will fit a "tilted ring model" to the velocity field to give us a rotation curve. A tilted ring model involves approximating the galaxy as a set of rings of gas at increasing radius, each with their own velocity, inclination, and position of the major axis. The rotation curves achieved through this method are figures 9 and 10.
Fig. 9 Tilted ring model rotation curve for NGC 4618. The top graph is the rotation curve, with error bars given by the standard deviation in the fit. The bottom two curves show the fitted position of the major axis and inclination for each ring.
Fig. 10 Tilted ring model rotation curve for NGC 4618. In order for the program to fit NGC 4625, we were forced to fit the two sides of the galaxy separately. The receding side is shown in dots, and the approaching side is shown in open circles. A reasonable fit could not be obtained for the receding side at 30 km/s. The top graph is the rotation curve, with error bars given by the standard deviation in the fit. The bottom two curves show the fitted position of the major axis and inclination for each ring.
Notice while the velocity field is much more behaved for NGC 4625, its rotation curve does not look smooth like the rotation curve for NGC 4618. In fact, we were forced to fit the receding and approaching sides of NGC 4625 separately to get a reasonable fit at all. This is evidence that although NGC 4625 looks like a fairly unperturbed system, it has been affected by the interaction with NGC 4618 as well.
We also looked at the velocity of the gas as a function of radius by effectively slicing along the major axis of the galaxy and looking at the velocity of the gas there. These curves, called position velocity diagrams, can be more useful than the fits for analyzing the qualitative movement of the gas, but the velocities given only reflect the part of the velocity of the gas. So the magnitude of the velocity of the gas would need to be adjusted for the inclination of the galaxy before we can use it quantitatively. Figures 11 and 12 show these curves.
Fig. 11 Position velocity diagram for NGC 4618. The two sides of the galaxy have been plotted together, where the open circles are for the approaching side of the galaxy, and the closed circles are for the receding side.Error bars are given by the channel width.
Fig. 12 Position velocity diagram for NGC 4625. The two sides of the galaxy have been plotted together, where the open circles are for the approaching side of the galaxy, and the closed circles are for the receding side. Error bars are given by the channel width.
Notice that the position velocity diagram for NGC 4525 looks much more smooth than its rotation curve. We are still trying to understand how to reconcile those plots. Also, while the position velocity diagram for 4618 does not include the tidal arms at all (they had to be excluded for the clarity of the curve), the last two points on the receding side of NGC 4625's curve might belong to the bridge feature.
Now that we have rotation curves, we can use the velocity of the gas to tell us something about the mass distribution (total this time, not HI mass) that is causing the gas to rotate at that velocity. Right now we are writing a program that uses a generic halo model to derive this total dynamical mass. Results will be posted soon...
Interaction History
The next step is to use all this information to see what constraints we can put on the interaction history of these galaxies... This will be posted as it happens. For now, Dr. Chris Mihos' Galaxy Collision simulation applet is an excellent way (not to mention a lot of fun) to begin to understand how galaxies interact.
Conclusions
Right now we have calculated a lot of information about the morphology and kinematics of the gas in these galaxies, but are still in the process of deciding what this tells us about the galaxies. So for now, this section will remain conspicuously blank...
References
[1] The Digitized Sky Survey was produced at the Space Telescope Science Institute under U.S. Government grant NAG W-2166. The images of these surveys are based on photographic data obtained using the Oschin Schmidt Telescope on Palomar Mountain and the UK Schmidt Telescope. The plates were processed into the present compressed digital form with the permission of these institutions.
[2] Odewahn, S. C. 1991, AJ, 101, 829.
[3] Burke and Graham-Smith. An Introduction to Radio Astronomy. (Cambridge University Press, © 1997)
[4] Synthesis Imaging in Radio Astronomy. A.S.P. Conference Series Volume 180. © 1999
Links
Sites that are somehow associated with me, good astrophysics resources, etc....
UW's Astronomy Department
CWRU's Astronomy Department
CWRU's Physics and Astronomy Club
NED (Extragalactic database)
NASA Astrophysics Data Service
