Final Report Summary - SURFING (A Survey of Feedback in Nearby Galaxy Groups)
Project context and objectives
Most galaxies in the universe reside in small groups of a few to a few tens of members, while a small minority forms larger clusters that may contain hundreds or thousands of galaxies. The stars in these galaxies, while producing most of the optical light, often make up only a fraction of the baryon content of the group. High-temperature (10^6-10^8 K) X-ray-emitting haloes of plasma, in which the galaxies are embedded, is often the dominant baryonic component. Simulations of structure formation over the history of the universe produce reasonably accurate reproductions of the global distribution of galaxies and clusters, but fare poorly when we try to model these baryonic components. Simulations that consider only the effects of gravity predict too much gas cooling from the hot phase, which in turn leads to the production of too many stars. An additional heat source is required in the form of a feedback mechanism, which can maintain a stable balance between cooling and heating.
It is now widely accepted that active galactic nuclei (AGN) provide this heating in massive groups and clusters of galaxies. All galaxies host supermassive black holes in their cores, and when material is accreted by these black holes they produce powerful jets of relativistic particles. In groups and clusters, gas cooling from the hot intra-cluster medium will flow into the central galaxy and accrete onto the black hole, producing jets that heat the surrounding gas, suppressing further cooling. Both gas and AGN are strong X-ray sources, while the relativistic jets are most visible via their radio emission.
To date, most studies of this AGN feedback heating have focused on the most massive clusters with the most powerful AGN. This project focused on the more numerous and representative galaxy groups, where feedback should have the greatest impact on structure formation and galaxy evolution. Our goal was to understand how feedback has influenced the thermal history of galaxies and the intra-group medium, and thus most of the baryons in the universe. To approach this issue, we required a combination of X-ray, radio and (to a lesser extent) optical observations so as to be able to study the galaxies, hot gas and AGN in a large sample of groups.
Work performed
The project as originally envisioned was split into five main tasks:
- selection of a sample;
- acquisition and analysis of data in the X-ray;
- radio;
- other supporting wavebands;
- comparison of the results from galaxy clusters and numerical simulations.
A selection of a statistically complete sample was necessary to ensure that our results were unbiased and comparable with results from theoretical models. We therefore selected a sample of 53 galaxy groups in the nearby universe, which we refer to as CLoGS (the complete local-volume groups sample). Around 50 % had already been observed in the X-ray and about one-third had suitable radio data. The acquisition of new X-ray observations was the most critical requirement for the project, and we proposed to observe our sample by both major X-ray observatories; we negotiated with one of the Chandra instrument scientists for his support and an additional allocation of time. We were partially successful in obtaining 50ks of Chandra and 175ks of the X-ray Multi Mirror (XMM) Newton time, enough to cover the richer half of the sample. Unfortunately, each observatory operates on year-long cycles, and the fellowship ended before these observations could be completed.
We therefore restructured the project and placed greater emphasis on the analysis of a pilot sample of 18 groups for which good quality data was already available, and on individual studies of particularly informative systems. Work on the larger sample has continued in parallel. We were highly successful in acquiring new radio data, with 128 hours of observations assigned so far; we expect to cover the whole group sample by the end of 2012. Members of the project team, which includes 16 scientists in five countries, have begun analysis of the available optical, near-infrared and ultraviolet data, and two postgraduate students in the Birmingham group will carry out project work as major components of their PhD studies. The project has resulted in nine refereed publications, with more in preparation.
Main results
Our studies of individual groups have revealed a number of important results that challenge common assumptions about AGN feedback. One example is the question of the AGN duty cycle. The typical timescale of AGN activity is usually taken to be a 10 million-year period of activity, followed by 100 million years in which the system relaxes. We find differences of at least an order of magnitude in the duration of episodes of jet activity, with some AGN maintaining powerful jets over at least 150 million years. Observations of sources with evidence of multiple outbursts also show that the activity timescale can change; AGN need not maintain a stable pattern of activity. We find evidence of direct heating by supersonic shocks in a number of systems, as well as uplift and mixing of gas by the AGN jets. Although a number of examples of shocks in groups and clusters are now known, their contribution to feedback in low-mass systems is often ignored, since they are more difficult to detect than the giant cavities inflated by radio jets. We show that, when they are detected, they typically contain more energy than the cavities and probably play a critical role in heating many groups. Most importantly, we have studied two nearby systems that host galactic coronae, which are small-scale cooling regions made up of gas ejected from stars within a galaxy which nonetheless fuels powerful AGN. It appears that these systems can fuel the AGN for long periods, but are so small that the jets do not effectively heat the cooling gas, thus breaking the feedback cycle. Coronae may fuel up to half the active systems in cluster cores, but examples in clusters are too distant for detailed study. We focused on two nearby examples, providing detail that was impossible to achieve from cluster studies and allowing us to examine possible mechanisms for the formation and destruction of the corona.
Two papers describing our pilot sample have been published and two more are in preparation. By using low-frequency radio observations from the Giant Metrewave Radio Telescope (in India), we were able to identify old fading radio galaxies that had gone undetected in earlier surveys. This provided a much clearer view of the range of jet activity in groups, and led to the discovery of several diffuse radio structures whose origins we are now attempting to determine. A comparison of the observed properties of AGN jets between our groups and previously observed clusters showed that while there are simple relationships between jet power and luminosity, a revision of the models describing these relations is needed. We have also found that although the power output of AGN jets is closely matched to the rate of radiative cooling in galaxy clusters, jets in groups are a factor of ~4 more powerful. This suggests either that AGN in groups are active for shorter periods, or that clusters are able to retain more of the energy released. In either case, this mass-dependent difference in feedback physics is clearly important, and is now a central goal of our research.
In terms of the original five tasks, only the first task is complete, owing to the delay in acquiring new X-ray observations. Although the remaining tasks could not be completed within the project lifespan, we have made important progress in every area, and expect to meet the project goals over the longer term. In terms of scientific progress, the project has been a great success, highlighting the importance of AGN feedback groups to the wider astronomical community and providing the basis for the CLoGS project, the first statistically complete multi-wavelength survey of galaxy groups in the local universe.
The project website describes our 53-group sample and our progress in acquiring and analysing the data. It is hosted on the Birmingham School of Physics and Astronomy website.
Project website:
http://www.sr.bham.ac.uk/~ejos/CLoGS.html
Most galaxies in the universe reside in small groups of a few to a few tens of members, while a small minority forms larger clusters that may contain hundreds or thousands of galaxies. The stars in these galaxies, while producing most of the optical light, often make up only a fraction of the baryon content of the group. High-temperature (10^6-10^8 K) X-ray-emitting haloes of plasma, in which the galaxies are embedded, is often the dominant baryonic component. Simulations of structure formation over the history of the universe produce reasonably accurate reproductions of the global distribution of galaxies and clusters, but fare poorly when we try to model these baryonic components. Simulations that consider only the effects of gravity predict too much gas cooling from the hot phase, which in turn leads to the production of too many stars. An additional heat source is required in the form of a feedback mechanism, which can maintain a stable balance between cooling and heating.
It is now widely accepted that active galactic nuclei (AGN) provide this heating in massive groups and clusters of galaxies. All galaxies host supermassive black holes in their cores, and when material is accreted by these black holes they produce powerful jets of relativistic particles. In groups and clusters, gas cooling from the hot intra-cluster medium will flow into the central galaxy and accrete onto the black hole, producing jets that heat the surrounding gas, suppressing further cooling. Both gas and AGN are strong X-ray sources, while the relativistic jets are most visible via their radio emission.
To date, most studies of this AGN feedback heating have focused on the most massive clusters with the most powerful AGN. This project focused on the more numerous and representative galaxy groups, where feedback should have the greatest impact on structure formation and galaxy evolution. Our goal was to understand how feedback has influenced the thermal history of galaxies and the intra-group medium, and thus most of the baryons in the universe. To approach this issue, we required a combination of X-ray, radio and (to a lesser extent) optical observations so as to be able to study the galaxies, hot gas and AGN in a large sample of groups.
Work performed
The project as originally envisioned was split into five main tasks:
- selection of a sample;
- acquisition and analysis of data in the X-ray;
- radio;
- other supporting wavebands;
- comparison of the results from galaxy clusters and numerical simulations.
A selection of a statistically complete sample was necessary to ensure that our results were unbiased and comparable with results from theoretical models. We therefore selected a sample of 53 galaxy groups in the nearby universe, which we refer to as CLoGS (the complete local-volume groups sample). Around 50 % had already been observed in the X-ray and about one-third had suitable radio data. The acquisition of new X-ray observations was the most critical requirement for the project, and we proposed to observe our sample by both major X-ray observatories; we negotiated with one of the Chandra instrument scientists for his support and an additional allocation of time. We were partially successful in obtaining 50ks of Chandra and 175ks of the X-ray Multi Mirror (XMM) Newton time, enough to cover the richer half of the sample. Unfortunately, each observatory operates on year-long cycles, and the fellowship ended before these observations could be completed.
We therefore restructured the project and placed greater emphasis on the analysis of a pilot sample of 18 groups for which good quality data was already available, and on individual studies of particularly informative systems. Work on the larger sample has continued in parallel. We were highly successful in acquiring new radio data, with 128 hours of observations assigned so far; we expect to cover the whole group sample by the end of 2012. Members of the project team, which includes 16 scientists in five countries, have begun analysis of the available optical, near-infrared and ultraviolet data, and two postgraduate students in the Birmingham group will carry out project work as major components of their PhD studies. The project has resulted in nine refereed publications, with more in preparation.
Main results
Our studies of individual groups have revealed a number of important results that challenge common assumptions about AGN feedback. One example is the question of the AGN duty cycle. The typical timescale of AGN activity is usually taken to be a 10 million-year period of activity, followed by 100 million years in which the system relaxes. We find differences of at least an order of magnitude in the duration of episodes of jet activity, with some AGN maintaining powerful jets over at least 150 million years. Observations of sources with evidence of multiple outbursts also show that the activity timescale can change; AGN need not maintain a stable pattern of activity. We find evidence of direct heating by supersonic shocks in a number of systems, as well as uplift and mixing of gas by the AGN jets. Although a number of examples of shocks in groups and clusters are now known, their contribution to feedback in low-mass systems is often ignored, since they are more difficult to detect than the giant cavities inflated by radio jets. We show that, when they are detected, they typically contain more energy than the cavities and probably play a critical role in heating many groups. Most importantly, we have studied two nearby systems that host galactic coronae, which are small-scale cooling regions made up of gas ejected from stars within a galaxy which nonetheless fuels powerful AGN. It appears that these systems can fuel the AGN for long periods, but are so small that the jets do not effectively heat the cooling gas, thus breaking the feedback cycle. Coronae may fuel up to half the active systems in cluster cores, but examples in clusters are too distant for detailed study. We focused on two nearby examples, providing detail that was impossible to achieve from cluster studies and allowing us to examine possible mechanisms for the formation and destruction of the corona.
Two papers describing our pilot sample have been published and two more are in preparation. By using low-frequency radio observations from the Giant Metrewave Radio Telescope (in India), we were able to identify old fading radio galaxies that had gone undetected in earlier surveys. This provided a much clearer view of the range of jet activity in groups, and led to the discovery of several diffuse radio structures whose origins we are now attempting to determine. A comparison of the observed properties of AGN jets between our groups and previously observed clusters showed that while there are simple relationships between jet power and luminosity, a revision of the models describing these relations is needed. We have also found that although the power output of AGN jets is closely matched to the rate of radiative cooling in galaxy clusters, jets in groups are a factor of ~4 more powerful. This suggests either that AGN in groups are active for shorter periods, or that clusters are able to retain more of the energy released. In either case, this mass-dependent difference in feedback physics is clearly important, and is now a central goal of our research.
In terms of the original five tasks, only the first task is complete, owing to the delay in acquiring new X-ray observations. Although the remaining tasks could not be completed within the project lifespan, we have made important progress in every area, and expect to meet the project goals over the longer term. In terms of scientific progress, the project has been a great success, highlighting the importance of AGN feedback groups to the wider astronomical community and providing the basis for the CLoGS project, the first statistically complete multi-wavelength survey of galaxy groups in the local universe.
The project website describes our 53-group sample and our progress in acquiring and analysing the data. It is hosted on the Birmingham School of Physics and Astronomy website.
Project website:
http://www.sr.bham.ac.uk/~ejos/CLoGS.html