Lisa's Webpage

The AGN-Starburst Connection in Warm Infrared Galaxies

Most of this thesis is either published or in press. See either the Publications Section, or download the Gzipped Postscript file here (3.2 Mb)

Abstract

This thesis presents an investigation into the connection between starburst activity and Active Galactic Nuclei (AGN) in warm infrared galaxies. Previous work has found that some infrared galaxies contain Active Galactic Nuclei, while in others, the infrared source is intense star formation. It is unknown whether there is an evolutionary connection between the starburst and AGN activity in these galaxies.

In order to study this issue, a large optical spectroscopic survey of 285 IRAS galaxies was carried out in order to obtain accurate optical classification. Some infrared galaxies, particularly those with high infrared luminosities, are known to contain AGN which are obscured to the optical by large quantities of dust. To determine the fraction of galaxies in our sample which contain obscured AGN, the Parkes-Tidbinbilla interferometer was used to search for compact radio cores in a subsample of ~60 objects. Of these, 37% of the starburst galaxies and 90% of the AGN type galaxies contain compact radio cores. A bimodal core luminosity distribution is observed, implying that the source of the compact radio emission in many of the starburst galaxies is from complexes of compact radio supernovae rather than from AGN. Deep VLBI imaging and X-ray spectroscopy observations are planned to investigate this further.

Theoretical starburst and AGN modelling was carried out and involves the use of stellar population synthesis models (Starburst99 and Pegase) to produce the ionizing radiation field which is then utilised by our shock and photoionization modelling code MAPPINGs vIII. These models were used to define a new classification scheme for the optical diagnostic diagrams. This new scheme has been used to classify the galaxies in our sample into AGN, starburst and LINER types and compares favourably against the previous semi-empirical scheme of Veilleux & Osterbrock (1987), producing only 6% ambiguity in classifications between the different diagnostic diagrams compared with 16% ambiguity using the traditional method. A large fraction (70%) of the galaxies in our sample are classified optically as starburst, 17% are Sy2, 4% are Sy1, and 0.4% are LINERs.

Many galaxies in our sample lie on a mixing line between starburst and AGN, and most likely contain both phenomena. Theoretical models were used to define mixing lines on the diagnostic diagrams to determine the AGN contribution to the optical emission for these galaxies. The optical AGN contribution agrees well with that determined from infrared studies, and it is concluded that dust extinction is relatively unimportant for the majority of the objects in our sample.

To test whether an evolutionary scenario exists for the galaxies in our sample, a subsample of galaxies was selected which are currently undergoing interactions, as traced by tidal features (tails, plumes and bridges). For these galaxies, the AGN contribution was compared with three indicators of merger progress; morphological interaction class, age of the old stellar population, and projected separation. There is a distinct lack of optical AGN activity in the more evolved mergers, in apparent contradiaction to the standard merger scenario. This may tentatively be explained in terms of a combination of increased dust extinction and circumnuclear star formation as gas and dust is funnelled towards the centre of the merger in the later stages.


	The AGN-Starburst Connection in Warm Infrared Galaxies
	Most of this thesis is either published or in press. See either the Publications Section, or download the Gzipped Postscript file here (3.2 Mb)
	Abstract
	This thesis presents an investigation into the connection between starburst activity and Active Galactic Nuclei (AGN) in warm infrared galaxies. Previous work has found that some infrared galaxies contain Active Galactic Nuclei, while in others, the infrared source is intense star formation. It is unknown whether there is an evolutionary connection between the starburst and AGN activity in these galaxies. In order to study this issue, a large optical spectroscopic survey of 285 IRAS galaxies was carried out in order to obtain accurate optical classification. Some infrared galaxies, particularly those with high infrared luminosities, are known to contain AGN which are obscured to the optical by large quantities of dust. To determine the fraction of galaxies in our sample which contain obscured AGN, the Parkes-Tidbinbilla interferometer was used to search for compact radio cores in a subsample of ~60 objects. Of these, 37% of the starburst galaxies and 90% of the AGN type galaxies contain compact radio cores. A bimodal core luminosity distribution is observed, implying that the source of the compact radio emission in many of the starburst galaxies is from complexes of compact radio supernovae rather than from AGN. Deep VLBI imaging and X-ray spectroscopy observations are planned to investigate this further. Theoretical starburst and AGN modelling was carried out and involves the use of stellar population synthesis models (Starburst99 and Pegase) to produce the ionizing radiation field which is then utilised by our shock and photoionization modelling code MAPPINGs vIII. These models were used to define a new classification scheme for the optical diagnostic diagrams. This new scheme has been used to classify the galaxies in our sample into AGN, starburst and LINER types and compares favourably against the previous semi-empirical scheme of Veilleux & Osterbrock (1987), producing only 6% ambiguity in classifications between the different diagnostic diagrams compared with 16% ambiguity using the traditional method. A large fraction (70%) of the galaxies in our sample are classified optically as starburst, 17% are Sy2, 4% are Sy1, and 0.4% are LINERs. Many galaxies in our sample lie on a mixing line between starburst and AGN, and most likely contain both phenomena. Theoretical models were used to define mixing lines on the diagnostic diagrams to determine the AGN contribution to the optical emission for these galaxies. The optical AGN contribution agrees well with that determined from infrared studies, and it is concluded that dust extinction is relatively unimportant for the majority of the objects in our sample. To test whether an evolutionary scenario exists for the galaxies in our sample, a subsample of galaxies was selected which are currently undergoing interactions, as traced by tidal features (tails, plumes and bridges). For these galaxies, the AGN contribution was compared with three indicators of merger progress; morphological interaction class, age of the old stellar population, and projected separation. There is a distinct lack of optical AGN activity in the more evolved mergers, in apparent contradiaction to the standard merger scenario. This may tentatively be explained in terms of a combination of increased dust extinction and circumnuclear star formation as gas and dust is funnelled towards the centre of the merger in the later stages.