Cosmic Paleontology:
Uncovering the Origins of Fossil Galaxies

Källan Berglund

Brown University

UW Madison Astrophysics REU 2014

Contact Me

Abstract

This project investigates the formation and evolution of fossil galaxies, specifically the history of AGN activity as it relates to galaxy mergers. We used low-frequency radio data from the J-VLA radio telescope's new P-band recievers [300-350MHz] to observe fossil galaxy J171811.93+563956.1 (referenced as FG30) at a red-shift of z=0.114. FG30 was selected for observation because it has stron X-ray emission from the surrounding IGM, indicative of an active galactic nucleus (AGN). After cleaning and calibrating the data using CASA, I generated images mapping the intensity of radio emission, revealing that FG30 is nearly a point source and lacks any prominant AGN jets. Analysis of the SDSS optical spectrum of FG30 revealed strong evidence of shocks. We believe that past AGN activity heated the IGM to produce the strong X-ray emission, though the jets have been dormant for long enough that the intergalactic medium (IGM) filled in the regions previously cleared by jets. The density of new material is now causing strong shocks when hit by newly restarted jets. This implies the start of a new epoch of AGN activity for FG30, which was most likely caused by a recent galaxy merger. This overturns current beliefs that fossil galaxies have been dynamically quiescent for a significant period of time.

Background and Theory

Galaxies are collections of stars and solar systems held in the same reagon of space by their mutual gravitational attractions, usually orbiting around a central black hole. Galaxy groups are collections of galaxies with low velocity dispersion (moving slowly relative to each other) which are gravitationally bound to each other. The low velocity dispersion in galaxy groups permits galaxies to join into a new, larger galaxies when they pass through or near each other. This process is called a galaxy merger.

Fossil groups are dominated by a single, large elliptical galaxy which has the X-ray luminosity of an entire cluster. The other cluster members are primarily small dwarf galaxies. There are no mid-range galaxies (like our Milky Way). The current theory is that these large fossil galaxies are the result of sequential mergers of all the large and mid-range galaxies of the group.

The strong X-ray emission seen from the intergalactic medium (IGM) around fossil galaxies can be fed by active galactic nuclei sending jets of relativistic plasma out into space. This AGN activity can be spurred by major (and possibly minor) galaxy mergers. When charged particles, like the electrons in AGN jets, are accelerated in circles by magnetic fields, they emit electromagnetic radiation at signature radio wavelengths. We can detect this emission and graph the frequency v. amplitude to find the position of any "knee" where the graph curves more than in other places. The location of this "knee" tells us how recently the nucleus was fed new material. In the event of a galaxy merger, new gas, dust, and stars would fall into the black hole and fuel the AGN jets. Thus, we can determine from the synchrotron spectrum when in the galaxy's history it was in the process of merging with other galaxies.

The spectrum of a galaxy in optical wavelengths can tell us a lot about the behavior of a galaxy. If emission lines are broad, it could be a blazar (an AGN with jets oriented along out line of sight). If the spectrum is abundantly blue, there is a lot of star formation, and if it is red, there is very little star formation. Certain emission spikes can tell us about element abundances, like OII and H-$\alpha$ which can also indicate star formation.

Research

I have been processing large quantities of low-frequency radio data [300-350MHz] for J171811.93+563956.1, henceforth referred to as FG30. This data was gathered by the J-VLA, with new P-band receivers. After cleaning and calibrating very noisy data using Common Astronomy Software Applications (CASA), I generated images of FG30. The process of cleaning and calibrating the data consisted of making many types of plots in which radio frequency interferance stands out and can be cut from the dataset. The plots below represent the final checks that the worst noise has been removed and that the calibrations were applied sucessfully.

Plot of phase (deg) v. frequency (by channel), colored by spectral window and iterated over antenna. The continuous plots confirm the quality of the bandpass calibration.
Plot of amplitude v. time, colored by spectral window and iterated over antenna. The continuous plots confirm the quality of the gain solutions.
Phase v. amplitude of flux/bandpass calibrator, colored by corrolation. The circular blob confirms the quality of the calibrations.
Phase v. amplitude of phase calibrator, colored by corrolation. A circular blob would confirm the quality of the calibrations, so the triangular distribution is not understood. The quality of the calibrations have been confirmed via other methods, including flux comparisons wiht existing data. While this anomaly is remains unexplained, we are confident that it does not interfere with our results, and it is believed to be the result of an incompatibility of the new P-band recievers with existing dataprocessing tools such as CASA.
Intensity map of residual flux redistributed in the image-cleaning process. Coordinates of right ascention and declination wiht color representing flux in Jy/beam.

Results and Analysis

From our radio-frequency images, we find FG30 to not have much structure. It is nearly a point-source and has no visible AGN jets. It is possiible that the jets are small and therefore do not show up in our image, but the X-ray emission lead us to expect strong jets heating the IGM. Another possibility is that the jets are oriented along our line of sight, making FG30 a blazar, but if this were the case, we ought to have observed a very different radio flux value than other observations at different times, which we did not. Our flux values were consistent wiht previous, though lower-resolution, studies.

Intensity map of FG30 in low-frequency radio emission. Coordinates of right ascention and declination wiht color representing flux in Jy/beam.
Intensity map of FG30 in low-frequency radio emission. Coordinates of right ascention and declination wiht color representing flux in Jy/beam. White line contour map of X-ray intensity of FG30.

Our low-frequency radio readings are supported by previously observed higher-frequency radio data. The higher-frequency observations are more diffuse because they are lower-resolution, but they have the same nearly-point-source flux distribution as we see in the new lower-frequency data.

Intensity map of FG30 in low-frequency radio emission. Coordinates of right ascention and declination wiht color representing flux in Jy/beam. White line contour map of higher-frequency radio emission intensity of FG30 (lower resolution).

Optical observations made by SDSS also reveal the same round, elliptical shape of FG30, and the spectrum in optical wavelengths reveals a great deal about the galaxy. The fact that the emission lines are not extremely broad provides additional evidence against FG30 being a blazar. We considered the option that the radio emission concentrated at the centerof the galaxy could be thermal emission from new star formation, but the optical spectrum would be more blue if that were the case, and it is more red (as is typical of fossil galaxies, leading to the inference that they do not have much star formation). A key insight from the optiscal spectrum is that there is strong evidce of shocks, which would come from AGN jets hitting dense IGM material. This, combined with the lack of apparent jets in the radio image, leads us to believe that the jets may have just turned on (perhaps after a period of inactivity since the last AGN epoch). If the nucleus has recently been activated to the extent of producing shocks from jets, the most likely cause is a recent merger.

SDSS optical image of FG30 (also pictured at top of page).
SDSS optical spectrum of FG30. Amplitude v. frequency with emission spikes labeled.

Conclusions

Mapping the intensity of radio emission reveals that FG30 is nearly a point source and lacks any prominant AGN jets. Analysis of the SDSS optical spectrum of FG30 revealed strong evidence of shocks. We believe that past AGN activity heated the IGM to produce the strong X-ray emission, though the jets have been dormant for long enough that the intergalactic medium (IGM) filled in the regions previously cleared by jets. The density of new material is now causing strong shocks when hit by newly restarted jets. This implies the start of a new epoch of AGN activity for FG30, which was most likely caused by a recent galaxy merger. This overturns current beliefs that fossil galaxies have been dynamically quiescent for a significant period of time.

Further investigation of the presence of shock evidence in the optical spectrum is underway.

The next steps are to determine remaining low-frequency radio luminosity values for FG30 from the P-band image flux and add to the graph of frequency v. wavelength, in order to determine the aging of synchrotron emission. This will be done if there appears to be a "knee" in the plot.

All of these results will be compared to an analysis of star-formation history, assessed by taking spectra of sample regions in the galaxy and graphing the spectrum as though it is a single star. If there are multiple distinct stellar populations, their age could shed light on the merger history of the galaxy. This is because star formation can be triggered by mergers, just as AGN activity can be stimulated form the infall of new material.

In the future, we hope to conduct these same observations on many more fossil galaxies and compare the resulting flux to data taken at many other wavelengths.

Acknowledgments

Thank you to the UW Madison Astrophysics REU Program and to the National Science Foundation for providing this wonderful opportunity. I wish to express my gratitude to Dr. Eric Wilcots and Anna Williams for their assistance on this project, and to Dr. Eric Hooper for keeping the REU program running smoothly. I would also like to thank my fellow REU participants for helping to form such a wonderful community during this program, and lasting friendships for the future.

Related Links