Large Radio Sources and Episodic Nuclear Activity

A THESIS SUBMITTED TO THE KUMAUN UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN PHYSICS

By Sumana Nandi

Aryabhatta Research Institute of observational sciencES (ARIES) Manora Peak, Nainital 263 129, India

November 2012 i

DECLARATION

I hereby declare that the work presented in this thesis is a result of the investigation carried out by me at the Aryabhatta Research Institute of observational sciencES (ARIES), Nainital, under the joint supervision of Prof. H. C. Chandola (Department of Physics, Kumaun University, Nainital), Dr. Mahendra Singh (Aryabhatta Research Institute of observational sciencES, Nainital) and Prof. D. J. Saikia (National Centre for Radio As- trophysics, TIFR, Pune and Cotton College State University, Guwahati). This thesis has not been submitted for the award of any degree, diploma, associateship or fellowship of any University or Institute.

Place : Nainital Date : (Sumana Nandi) ii

CERTIFICATE FROM THE SUPERVISORS

This is to certify that

1. The synopsis of the thesis entitled “Large Radio Sources and Episodic Nuclear Activity” for the award of the degree of Doctor of Philosophy in Physics was ap- proved by the Kumaun University, Nainital (Letter no.-Res/43/Physics-2009, dated 14/12/2009).

2. This thesis embodies the work of Mrs Sumana Nandi herself.

3. Mrs Sumana Nandi worked under our joint supervisions for this thesis as a Research Fellow at the Aryabhatta Research Institute of observational sciencES (ARIES), Nainital. She has put in more than 200 days of attendance at ARIES, Nainital during this period.

4. This thesis has not been submitted before for the award of any degree, diploma, associateship or fellowship of any University or Institute.

Prof. H. C. Chandola Dr. M. Singh Prof. D. J. Saikia Department of Physics ARIES, Manora Peak NCRA-TIFR, Pune and Kumaun University Nainital Cotton College State University Nainital Guwahati iii

To my parents iv Acknowledgments

It has been a long journey and I am fortunate to have come in touch many nice people during this journey. It is a great pleasure to thank everyone who gave me the possibility to complete this thesis. First and foremost, I would like to express my deep and sincere gratitude to my supervisors, Prof. H. C. Chandola, Head of the Department of Physics, DSB Campus, Kumaun University, Dr. Mahendra Singh, Scientist E, ARIES, Nainital and Prof. D. J. Saikia, National Centre for Radio Astrophysics Tata Institute of Fundamental Research, Pune and presently Vice Chancellor, Cotton College State Uni- versity, Guwahati. I wish to express my warm and sincere thanks to Prof. H. C. Chandola for encouraging my research and strong support. I am deeply grateful to Dr. Mahendra Singh who has been supportive, friendly and has given me the freedom to pursue various projects without objection during the period. His patience, invaluable advices and suggestions have been a source of inspiration to me. I owe my sincere gratitude to Professor D. J. Saikia who gave me the opportunity to work at NCRA-TIFR. Sykes, your advice and guidance on both research as well as on my career have been invaluable. I convey my heartfelt gratitude to Prof. Ram Sagar, Director, ARIES, Nainital. He always encouraged me and has provided the necessary guidance at different stages. I am grateful to Marek Jamrozy, Ishwara Chandra, Hum Chand, Oleg V. Verkhodanov, Chiranjib Konar, Gopal Krishna, Subhasish da, Dipanjan da, Nimisha, Nissim, Jayaram, Sandeep Sirothia, Chandreyee di, Nirupam, Prasun for useful discussions and many helpful comments on various aspects this work. I am thankful to academic committee and all faculty members of ARIES for numerous helps and kind supports. I thank ARIES and NCRA library staff and computer sections staff for providing the necessary facilities for my work. I am also thankful to all faculty members and other staff of Department of Physics and Research Board of Kumaun University for their kind co-operation, advice and help. I want to thank all my funny and friendly office mates Ravi ( knowledgeable expert in AGN and Badminton), Krishna (very soft hearted guy), Jai (our ticket agent) and Akash (always helpful in radio astronomy discussion). Thanks to my next door office mates Brajesh (Bhaishab), Eswaraiah (a good singer), Subhash (always friendly and cheerful about everything), Raman, Pradeep, Narendra, Ram Kesh (most reliable person to consult if water problem comes in hostel), Tapaswini and Abhishek. I also thank my ju- niors Devesh (probably only one poet in ARIES), Haritma, Hema, Rajiv, Sumit (Rajani), v

Archana (always been like a little sister to me), Piyush, Vijay, Neha, Subhajeet and all 1st year students for their lively company. I would like to thank my seniors Raman, Arti, Manash, Neelam, Jessy, Chavi, Amitava da, Bharat, Sivraj. I also wish to thank Pradip da and Samaresh da for making me feel at home. Throughout, my long journey in ARIES I get opportunity to feel the beauty of hidden paradise. Nature has been a source of inspiration to me. I like to spend my leisure under the old deodar tree in front of my office and try to find out exciting and enchanting wonderful gifts of nature. Among the mountain ranges of Kumaun I liked the land’s end peak most. It changes all its beauty with the season and becomes most beautiful in full moon night of autumn. The red budas flowers, color full clouds of western sky, snow capped pine trees, chilled nights, dew-dripping green grass and beautiful birds of Manora Peak, I love them all. My special thanks to my friends at NCRA for the good times we spent. Aritra, Rahul, Prasun, Arti, Naren (the great cook), Vishal, Shweta, Alka, Ujjwal, Yojesh, Aruan, Abhishek, Aditya, Aditi, Breezy, Kiran, Viswesh, Sarvanan, Sambit, Maryam have con- tributed a lot to my stay at NCRA memorable. They are all a fun bunch with lots of enthusiasm. I would like to spatially thanks Aritra, Prasun, Chandreyee di, Nirupam, Arti for numerous help during the data analysis and scientific discussion. My time at NCRA was made enjoyable in large part due to my spacial friends Chandreyee di, Nirupam, Ar- itra, Bhaswati di and Jayanta da. Chandreyee di always taught me how to face challenges in our life with a smile. It is impossible to forget those pleasant evenings at NCRA hostel corridor or at hostel roof and listening ghost stories from Nirupam, singing Rabindra sangeet and enjoying black tea specially made by Chandreyee di. I must acknowledge the contribution of Projit Palit, Biupul da, Lalu da and my school teacher Ajit da, my professors Pritam Roy, Ranabir Dutta and Somnath Chakraborti. They all helped me to realized that Physics is my most favorite subject. Above all this thesis would not have been possible without the support and encouragement of my family: my dear parents Dr. Subodh Gopal Nandi and Arati Nandi, my beloved sweet twin sisters Sudarshana and Suranjana. Now I feel very happy when I think that I have fulfilled at least one wish of my parents through this work. I wish a bright carrier for my sisters. I must remember my grand mother and my didu’s blessings when I left Santiniketan 1st time for this study. Saikat, Ushree, Raja, Tumpa boudi, my uncles, aunts, Sanjib da, didi and Rit always make me feel happy through their chatting and phone calls. vi

I am grateful to my mother-in-law (Gouri Roy) and father-in-law (Ranjan Roy) for their all time supports. This page I can not close without acknowledge Rupak. He is the person whom I bothered too much during this work. Fortunately I was his office mate, and always got a good opportunity to discuss both my academic and personal problems with him at any time even when he is writing his important paper. I am really surprised to see his understanding and patience. Thank God, you have sent me Rupak like great friend during my research period and for my whole life. I am immensely indebted to the several institutes and organisations. I thank DST, Government of India for financial support vide Grant No. SR/S2/HEP-17/2005. The GMRT is a national facility operated by the National Centre for Radio Astrophysics of the Tata Institute of Fundamental Research. I thank the staff for help with the observations. The National Radio Astronomy Observatory is a facility of the National Science Founda- tion operated under co-operative agreement by Associated Universities Inc. I thank the VLA staff for easy access to the archival data base. This research has made use of the NASA/IPAC extragalactic database (NED) which is operated by the Jet Propulsion Labo- ratory, Caltech, under contract with the National Aeronautics and Space Administration. This research also has made use of Sloan Digital Sky Survey (SDSS) (The SDSS Web Site is http://www.sdss.org/) and Faint Images of Radio Sky at Twenty-centimeters (FIRST). I thank numerous contributors to the GNU/Linux group and Matteo Murgia for access to the SYNAGE software.

Sumana Nandi vii Preface

Powerful extragalactic radio sources exhibit a wide range of linear sizes, ranging from the compact steep-spectrum sources, which could be less than a few tens of parsecs and are conﬁned within the central regions of their parent galaxies, to the giant radio sources

(GRSs). The GRSs, which are amongst the largest objects in the Universe, are deﬁned

1 1 to be those which have a projected linear size 1 Mpc (H =71 km s− Mpc− , Ω =0.27, ≥ o m

Ωvac=0.73, Spergel et al., 2003). These sources are useful for studying the late stages of evolution of radio sources, constraining orientation-dependent uniﬁed schemes and probing the intergalactic medium at diﬀerent redshifts (e.g. Subrahmanyan & Saripalli 1993; Subrahmanyan, Saripalli, & Hunstead 1996; Mack et al. 1998; Ishwara-Chandra & Saikia 1999; Kaiser & Alexander 1999; Blundell, Rawlings, & Willott 1999 and references therein; Schoenmakers 1999; Schoenmakers et al. 2000c, 2001; Konar et al. 2008; Jamrozy et al. 2008).

In the standard model for high-luminosity radio sources, the energy emitted by the nucleus in the form of relativistic jets traverse outwards at close to the velocity of light initially through the host galaxy and latter through the intracluster or intergalatic medium. The jets dissipate their energy at the leading edges, forming regions of high brightness called ‘hotspots’. The relativistic particles ﬂowing out of the hotspots form the lobes and bridges of emission which are older as one goes away from the hotspots. The radio continuum spectra in the diﬀerent regions have information on the history of the various energy losses and gains of the radiating particles during the lifetime of the source. Multifrequency studies of these sources have yielded important information on their ages and possible reacceleration of particles. The importance of combining lowfrequency the Giant Metrewave Radio Telescope (GMRT) observations with higher frequency data from the Very Large Array (VLA) has been demonstrated in a number of recent studies (Konar viii et al. 2004, 2006, 2008; Jamrozy et al. 2008).

These large radio galaxies are believed to be powered by an active galactic nucleus (AGN) which consists of a supermassive black hole with an accretion disk. For all AGNs, an important and interesting issue is whether their active phase is episodic, and if so, the duration and the time scales for recurrence of such periods of activity. Episodic activity could be related to periodic feeding of a supermassive black hole which is responsible for the nuclear activity. In radio loud objects, a very striking example of episodic activity is when a new pair of radio lobes is seen closer to the nucleus in a radio loud AGN before the older and more distant radio lobes have faded. There are two dozen such well- established sources which are widely known as double-double radio galaxies (DDRGs; Schoenmakers et al. 2001; Saikia, Konar, & Kulkarni 2006). Here the newly formed jets are traversing through the cocoon of plasma formed by the earlier cycle of activity rather than the general interstellar medium of the host galaxy or the intergalactic medium. In addition to the DDRGs, relativistic plasma from an earlier cycle of activity may also form a halo of diﬀuse emission around the newly formed lobes or jets as seen in the nearby radio galaxy 4C29.30 (Jamrozy et al. 2007). Although most of the DDRGs tend to be associated with GRSs, evidence of episodic activity is seen in smaller sources as well.

In this thesis we propose to study a sample of large radio sources at a wide range of frequencies using both the GMRT and VLA to determine their spectra and spectral ages; and explore and examine evidence of episodic activity in radio loud AGN and understand possible causes for such episodic activity. We also propose to make low-frequency observations with the GMRT and other telescopes to identify possible new examples of activity in these as well as other sources. Since the plasma from the earlier cycle of activity is much older, possibly >108 yr, these features are likely to be more easily visible ∼ ix at lower frequencies. We plan to study these cases of episodic activity and possible time scales of their episodic activity.

In light of the objectives proposed for this thesis, the following main results are obtained.

1. In the first case we present a multifrequency observations of a FRII type large radio galaxy 3C46 (J0135+3754). This source is suitable for the classical spectral-ageing analysis due to its large angular extent which is covered by a significant number of resolution elements. Using both GMRT and VLA data we have determined the spectral breaks in consecutive strips along the lobes of 3C46 as well as estimated spectral ages for each strips. Like most of the radio galaxies the spectral age of 3C46 increases with distance from the hotspot towards the core regions. This confirms that the acceleration of particles occurs close to the hotspots. The estimated injection spectral index is 1.0 whereas the maximum spectral age is 15 Myr near the core. Through low frequency observation it is ∼ possible to detect diffuse radio emission formed by an earlier cycle of activity of central

AGN. This diﬀuse emission is expected to remain visible for nearly 108 yr and shows a steep radio spectrum due to radiative losses. In our study the low frequency GMRT images of 3C46 do not show any diﬀuse emission possibly due to an episodic activity. So, this result supports that the renewed activity of AGN is not a common phenomenon even in large radio sources. The detailed work on 3C46 is presented in Chapter 3 of this thesis

2. In order to understand the episodic activity of AGN and possible time scale in between two episodes a steep spectrum core-dominated radio galaxy 3C293 (J1352+3126) has been observed over a large frequency range. This object is somewhat unusual in the sense that the inner double is not well aligned with the outer doubles, unlike x other known DDRGs. The projected linear size of the outer double and inner double is 190 kpc and 4.2 kpc respectively. Though the inner double of 3C293 is small in linear ∼ ∼ size it contributes more than 70% of the total ﬂux density of the whole source. The high resolution observations can resolve the double lobe structure of the central steep spectrum core and shows that it is highly asymmetric in location. Though our observations can not resolve the central region but have determined some reliable physical parameters, like the spectra, injection spectral index, magnetic ﬁeld strengths and spectral age for northern lobe, central component and southern lobe. Our estimation shows that the interruption of jet activity of 3C293 to be <0.1 Myr, much smaller than the other known DDRGs. A ∼ detailed description of this work is given in Chapter 4.

3. To enlarge the sample of DDRGs, we have focused on the Faint Images at Radio Sky at Twenty centimeter (FIRST) survey, where Proctor (2011) has classified 242 sources as DDRGs. We have studied their radio structure in detail and try to identify the possible optical host positions of these sources. Our result shows that out of 242 samples only 23 sources are good examples of DDRGs. We also make a list of 63 candidate DDRGs which require a high resolution radio observations or deep optical observations to identify their optical host galaxy position. We have examined the symmetry parameters for these 23 candidates. The asymmetric nature of the inner double compared with outer ones are not significantly different in this case. The overall projected linear sizes of these new sources are significantly smaller than the known DDRGs. Till date there are over about a dozen good examples of DDRGs known in literature. The projected linear size of these DDRGs are quite large compared to our 23 new detected candidate. These relatively small size DDRGs will help to give a better perspective of the range of time scales involved in recurrent AGN activities. The detailed work on these DDRGs has been discussed in Chapter 5. xi

The organisation of thesis is as follows. Introduction is presented in Chapter 1. In

Chapter 2, I discuss about radio antenna, Interferometry principle and the interferometric observational techniques. The analysis techniques of the radio-interferometric data are also presented in detail. In Chapter 3, I present the radio study of a large radio galaxy 3C46. In this study I estimate the injection spectral index, magnetic ﬁeld and spectral age of 3C46. Chapter 4 discusses the interesting case of DDRG 3C293 whose overall projected linear size is smaller and misalignment angle is larger compared to the other DDRGs. To enlarge the sample of DDRGs I examine 242 probable DDRGs from Proctor (2011). A detailed study of this sample is presented in Chapter 5. Chapter 6 summarizes the thesis work, conclusions of my results and outline of my future work. In Appendix A, I have listed a catalogue of 320 sources which have been detected at 154MHz. The list of candidate DDRGs and non-DDRGs identiﬁed from Proctor (2011) is given in Appendix B. xii

List of Publications

Publications in refereed journals

1. A multifrequency study of the large radio galaxies 3C46 and 3C452 S. Nandi, A. Pirya, S.Pal, C. Konar, D. J. Saikia, M. Singh, 2010, Monthly Notices of the Royal Astronomical Society, 404, 433

2. Radio spectra of giant radio galaxies from RATAN-600 data Khabibullina M. L., Verkhodanov O. V., Singh M., Pirya A., Nandi S., Verkho- danova N. V., 2010, Astronomy Reports, 54, 571

3. A radio study of the double-double radio galaxy 3C293 S. A. Joshi, S. Nandi, D. J. Saikia, C. H. Ishwara-Chandra, C. Konar, 2011, Monthly Notices of the Royal Astronomical Society, 141, 1397

4. A second set of RATAN-600 observations of giant radio galaxies Khabibullina M. L., Verkhodanov O. V., Singh M., Pirya A., Nandi S., Verkho- danova N. V., 2011, Astronomy Reports, 55, 392

5. A study of giant radio galaxies at RATAN-600 M.Khabibullina, O.V. Verkhodanov, M. Singh, A. Pirya, S.Nandi, N.V. Verkho- danova., 2011 Astrophysical Bulletin, 66, 717

6. A low-frequency study of two asymmetric large radio galaxies A. Pirya, S. Nandi, D. J. Saikia, M. Singh, 2011, Bulletin of the Astronomical So- ciety of India, 35, 141

7. Rejuvenated radio galaxies J0041+3224 and J1835+6204: how long can the quiescent phase of nuclear activity last? xiii

Konar C., Hardcastle M. J., Jamrozy M., Croston J. H., Nandi S., 2012, Monthly Notices of the Royal Astronomical Society, 424, 1061

8. Double-double radio galaxies from the FIRST survey Nandi S., Saikia D. J., 2012, Bulletin of the Astronomical Society of India, 40, 121

9. Multifrequency study of Double-double radio galaxies from the FIRST survey Nandi S., Saikia D. J., Singh M., Chandola H. C. (to be submitted)

10. A tale of two candidate rejuvenated radio galaxies: 3C 258 and TXS 1548+274 D.J. Saikia, P. Thomasson, S. Sirothia,1 C. Konar, S. Nandi, A. Pirya, M. Singh, H.C. Chandola (to be submitted)

11. Multiwavelength perspective of a neglected type Ib event R. Roy et al. (including S. Nandi and H.C. Chandola, to be submitted soon)

Publications in conference proceedings

1. Giant radio galaxies: problems of understanding and problems for CMB ? Verkhodanov O.V., Khabibullina M.L., Singh M., Pirya A., Verkhodanova N.V., Nandi S., 2008, "Practical Cosmology, Proceedings of the International Confer- ence "Problems of Practical Cosmology", 247

2. Episodic Activity in Active Galactic Nuclei D. J. Saikia , M. Jamrozy, C. Konar, S. Nandi, 2010, Proceedings of the 25th Texas Symposium on Relativistic Astrophysics., 14

3. The double-double radio galaxy 3C293 S. A. Joshi, S. Nandi, D. J. Saikia, C. H. Ishwara-Chandra, C. Konar, 2011, Journal of Astrophysics and Astronomy, 32, 487 xiv

4. A Multifrequency Study of Five Large Radio Galaxies Pirya A., Nandi S., Saikia D. J., Konar C., Singh, M., 2011, Journal of Astrophysics and Astronomy, 32, 471

5. The Dynamics of Radio Galaxies and Double-Double Radio Galaxies Konar C., Jamrozy M., Hardcastle M. J., Croston J. H., Nandi S., Saikia D. J., Machalski J., 2012 Journal of Astrophysics and Astronomy, 32, 477 xv

NOTATIONS AND ABBREVIATIONS

The notations and abbreviations which have been used for this thesis are collected here for a for a quick reference. All these notations and abbreviations have been also explained on their ﬁrst appearance in the text.

Notations

Å Angstrom(unitofwavelength) α spectral index

αin j injection spectral index B magneticﬁeldstrength

BiC magnetic ﬁeld strength equivalent to the CMB radiation B(l,m) far ﬁeld antenna reception pattern

′, arcmin arc minute

′′, arcsec arc second cm centimeter D theapertureofthetelescope Dec., δ Declination

◦, deg Degree eV Electron Volt Fig. Figure GHz Giga Hertz Hz Hertz (unit of frequency)

H0 Hubble constant h,hr Hour hrs Hours J2000 epochofobservation km Kilo meter λ Wavelength I(l,m) skybrightnessdistribution Jy Jansky kpc Kilo parsec (unit of distance)

MBH Mass of black hole M Mass of the Sun ⊙ Mpc Mega parsec xvi m meter mm millimeter µm micro meter milliarcsec milli arcsecond min Minutes

νbr break frequency nT nanotesla

Ωm Matter density of the Universe

Ωvac Vacuum energy density pc Parsec (unit of distance)

P178 luminosity at 178 MHz Ref. References RA Right Ascension rms, σ root mean square s, sec Seconds Sect. Section sr steradian V(u,v,w) complexvisibilities W watt yr year z Redshift

Abbreviations

AGN Active Galactic Nuclei AIPS Astrophysical Image Processing System ARIES Aryabhatta Research Institute of observational SciencES ATCA Australia Telescope Compact Array BL Lac BL Lacertae BLR Broad Line Region BLRGs Broad Line Radio Galaxies xvii

CSS compact steep spectrum DDRGs Double-Double Radio Galaxies DSS-POSS II Digitized Sky Survey - The Second Palomar Sky Survey FFT fast Fourier transform FIRST Faint Images at Radio Sky at Twenty centimeter FITS Flexible Image Transport System FR I Fanaroﬀ Riley Class I FR II Fanaroﬀ Riley Class II FSRQ Flat Spectrum Radio Quasar FWHM Full Width at Half Maximum GBT Green Bank Telescope GHB GMRT hardware backend GMRT Giant Metrewave Radio Telescope GPS gigahertz peaked-spectrum GSB GMRT Software Backend GRSs Giant Radio Sources HYMORS Hybrid morphology radio sources ISM Interstellar medium IGM inter galactic medium LOFAR Low Frequency Array LSB Lower Side Band NED NASA Extragalactic Database NLR Narrow Line Region NLRGs Narrow Line Radio Galaxies NRAO National Radio Astronomical Observatories NVSS NRAO VLA Sky Survey OVV Optically Violently Variable QSO Quasi Stellar Object QUASARs QUAsi-StellAr-Radio Sources RFI Radio Frequency Interference SKA The Square Kilometre Array SMBH Super Massive Black Hole SMART Stretch Mesh Attached to Rope Trusses USB Upper Side Band UV Ultra-violet xviii

VLA Very Large Array VLBI Very Long Baseline Interferometry WATs wide-angle tailed sources WSRT Westerbrok Synthesis Radio Telescope Contents

1 Introduction 1 1.1 Activegalacticnuclei: ...... 2 1.1.1 ClassiﬁcationofAGNs: ...... 3 1.2 Radiogalaxies: ...... 6 1.3 Double-doubleradiogalaxies(DDRGs): ...... 9 1.4 Spectralageinganalysis: ...... 12 1.5 Issuesinvestigatedhere: ...... 15

2 Data Set and Analysis Procedure 17 2.1 RadioTelescopes: ...... 17 2.2 Interferometryprinciple: ...... 20 2.2.1 Radio-Interferometer: Giant Metrewave Radio Telescope(GMRT) 23 2.2.2 Radio-Interferometer: Very Large Array (VLA) ...... 24 2.3 Analysis techniques of the radio-interferometric data:...... 25 2.3.1 PreProcessing: ...... 27 2.3.2 Dataediting: ...... 29 2.3.3 Calibration: ...... 32 2.3.4 Bandpasscalibration: ...... 35 2.3.5 Imaging: ...... 39 2.3.6 TheCLEANAlgorithm ...... 41

3 A multifrequency study of a large radio galaxy 3C46 45 3.1 Introduction...... 45 3.2 Observationsandanalyses ...... 48

xix xx CONTENTS

3.3 Observationalresults ...... 49 3.4 Discussionandresults...... 51 3.4.1 Radiativelosses...... 51 3.4.2 Spectralageinganalysis ...... 54 3.4.3 Searchforepisodicactivity...... 58 3.5 Concludingremarks...... 59

4 A radio study of the double-double radio galaxy 3C293 61 4.1 Introduction...... 62 4.2 Observationsanddatareduction ...... 65 4.2.1 GMRTobservations ...... 65 4.3 Observationalresults ...... 66 4.3.1 Overallstructureof3C293 ...... 67 4.3.2 Spectra ...... 68 4.4 Discussion...... 70 4.4.1 Spectralages ...... 70 4.4.2 Recurrentactivitytimescale ...... 72 4.5 Concludingremarks...... 73

5 Double-double radio galaxies from the FIRST survey 75 5.1 Introduction...... 76 5.2 MethodologyofclassiﬁcationoftheDDRGs ...... 78 5.3 Resultsanddiscussions ...... 88 5.3.1 Linearsizes...... 88 5.3.2 Arm-lengthandbrightnessratios...... 89 5.4 Concludingremarks...... 90

6 Conclusion and discussion 91 6.1 Summaryofconclusions ...... 91 6.2 Futurework:...... 94

A Appendix A: Catalogue of 320 sources detected at 154MHz 99

B Appendix B: List of the candidate DDRGs and non-DDRGs 113 List of Figures

1.1 An AGN cartoon; The green arrows represent the type of AGN which can be seen from a certain viewing angle according to the uniﬁcation scheme. (Imagecredit:NASA) ...... 4 1.2 Modelofradiogalaxy...... 8 1.3 5 GHz image of the FRI radio galaxy NGC 4261. Credit: Teddy Cheung. 8 1.4 5 GHz image of the FRII radio galaxy Cygnus A. Image courtesy of NRAO/AUI...... 9 1.5 DDRGJ1158+2621Imagecredit: Owen&Ledlow(1997) ...... 10

2.1 WindowsintheEarth’satmosphere...... 18 2.2 Componentsofasingledishradiotelescope ...... 19 2.3 Very Large Array, an interferometric array. (Image courtesy of NRAO/AUI) 21 2.4 GMRT full array (left panel) and central array (right panel) conﬁguration. 24 2.5 An Amplitude vs UV distance plot for 3C286 before editing and calibration (left panel). A Phase vs UV distance plot for 3C236 before data editingandcalibration(rightpanel) ...... 29 2.6 An Amplitudevs UV distance plotfor 3C286 after editingand calibration (left panel) An Phase vs UV distance plot for 3C236 after data editing and calibration(rightpanel)...... 35 2.7 Vector averaged cross-power spectrum of several baselines for 3C286 without bandpass (upper panel); Vector averaged cross-power spectrum of several baselines for 3C286 with bandpass. (lower panel) ...... 40

xxi xxii LIST OF FIGURES

3.1 GMRT low-frequency images of 3C46 at 153, 240, 332 and 606 MHz, and VLA higher-frequency images at 1465, 4841 and 8460 MHz. In this ﬁgure as well as in all the other ﬁgures, the peak brightness and the contour levels are given below of each image. In all the images the restoring beamisindicatedbyanellipse...... 47

3.2 The spectra of the extended emission of 3C46 obtained after subtracting the core flux density at frequencies greater than 1400 MHz from the ∼ total flux density. Any contributions of the core flux density at lower frequencies are less than 1 per cent and have been neglected. The total ∼ flux densities are from Laing & Peacock (1980) and the measurements presented in this paper. The fits to the spectra obtained using the SYNAGE package(Murgiaetal.,1999)arealsoshown...... 50

3.3 The spectral-index image of 3C46 with the strips used for estimating the spectral ages along the lobes being marked by vertical lines and labelled. The central region corresponds to the ‘Central core’ region in Tables 3.3 − and Fig. 3.6. The images have been rotated so that they lie in the east- west direction. The spectral indices have been estimated between 332 and 5000 MHz. The grey scale bar indicates variations in spectral index ∼ from 0.4 to 2.0. Some of the values at the edges are spurious...... 51

3.4 Typical spectra of the strips for the western (upper panel) and eastern (lower panel) lobes of 3C46, with the ﬁts from the JP model as described inthetext...... 53

3.5 Spectra of the central regions of 3C46 subtraction of the core ﬂux density, withtheﬁtsfromtheJPmodelasdescribedinthetext...... 56

3.6 The spectral age is plotted against the distance from the core for 3C46 usingthemagneticﬁeldestimatedforeachstrip...... 57

4.1 The GMRT images of 3C293 at 154, 240 and 614 MHz, and the VLA image at 4860 MHz. All these images have been made with an angular 2 resolution of 23.2 21.6 arcsec along a PA of 79◦, which is shown as an × ellipseinthebottomleft-handcorner...... 64 LIST OF FIGURES xxiii

4.2 The integrated spectrum of 3C293 using the measurements from Laing & Peacock (1980), Kuehr et al. 1981 and our measurements. These are shown as filled squares, and filled circles and open circles respectively (upper left). The spectra of the north-western (upper right), central (lower left) and south-eastern (lower right) components respectively. For the central component, the core flux density has been subtracted for measurements above 1400MHz,asdescribedinthetext...... 69 ∼ 5.1 FIRST images of the DDRGs listed in Table 6.1. In all the Figures the + sign when shown denotes the position of the optical object...... 81 5.1 (continued) ...... 82 5.1 (continued) ...... 83 5.1 (continued) ...... 84 5.2 The ratio of the luminosities of the outer to the inner doubles vs the projected linear size of the inner double. The + signs represent the DDRGs identified from the FIRST survey, while the open circles represent those fromSaikiaetal.(2006)...... 85 5.3 Upper panel represents arm-length ratio of the outer double vs the armlength ratio of the inner double where as Lower panel represents flux density ratio of outer double vs flux density ratio of inner double. The + signs represent the DDRGs identified from the FIRST survey, while the open circles represent those from Saikia et al. (2006)...... 87

A.1 A WAT source associated with a galaxy at a redshift of 0.061 along with evidence of diﬀuse emission. The left panel shows the higher resolution GMRT image at 154 MHz with a restoring beam of 15 12.5 arcsec2 along × 2 PA 73◦, while in the middle panel the restoring beam is 41 39 arcsec × along PA 159◦. The right panel shows the NVSS image at 1400 MHz. Crossesmarkthepositionoftheopticalgalaxy...... 100 A.2 A known giant radio galaxy with an FRII-type structure (left panel) and an FRI-type galaxy resembling the morphology of 3C 31 (right panel). Crosses in the images mark the positions of the optical galaxies. The restoring beam for the giant radio galaxy, J1400+3019, is 41 39 arcsec2 × along a PA=159◦, while for the FRI galaxy, J1356+3126, it is 15 12.5 × 2 arcsec along PA=73◦...... 101 xxiv LIST OF FIGURES

B.2 Examples of sources which we have classified as non-DDRGs. The inner emission in J0032 0019 is likely to be due to backflow from the hot- − spots, while in the case of J0759+4051 and J1434+0441, the optical identification is co-incident with one of the components of the inner double which is likely to be the radio core. The other feature is possibly a knot inthejet...... 114 List of Tables

1.1 AGNclassiﬁcation ...... 5

2.1 SystemparametersofGMRT...... 24 2.2 SystemparametersofVLA...... 25

3.1 Observinglog...... 48 3.2 The observational parameters and observed properties ofthesources . . . 52 3.3 Estimates of break frequency and spectral age for 3C46 ...... 56

4.1 Observing log. Columns (1) and (2) show the name of the telescope, and the array configuration for the VLA observations; columns (3) and (4) show the frequency and bandwidth used in making the images; column (5): the primary beamwidth in arcmin; column (6): dates of the observations...... 66 4.2 The observational parameters and flux densities. Column (1): frequency of observations in MHz, with the letter G or V representing either GMRT or VLA observations; columns (2) (4): the major and minor axes of the − restoring beam in arcsec and its PA in degrees; column (5): the rms noise 1 in units of mJy beam− ; column (6): component designation, where Cent refers to the central source including the radio core; columns (7) and (8): thepeakandtotalfluxdensitiesofthesource...... 68

5.1 DDRGsfromtheFIRSTsurvey ...... 80 5.2 Some of the observed properties of the sample of DDRGs ...... 86

6.1 DDRGsfromtheFIRSTsurvey ...... 94

xxv xxvi LIST OF TABLES

A.1 Sources within 2.2◦ of the phase centre of the GMRT observations at 154

MHz (HPBW 3.1◦). All sources with a peak flux density 7-σ, where σ ∼ ≥ 1 is the primary beam corrected local rms noise in units of mJy beam− are 1 listed. The values of σ range from 8.6 mJy beam− near 3C293 to typical 1 values of 3 4 mJy beam− in regions without strong sources. There are − a total of 320 sources. All but three of the weaker sources (J1352+3059, J1352+3039 and J1357+3111) are seen in the NVSS images at 1400 MHz. Using 5 times the rms noise in the NVSS images yields spectral indices steeper than 1.2 for these three sources. Columns 1 and 2: ∼ the right ascension (h:m:s) and declination (d:m:s) in J2000 co-ordinates; columns 3 and 4: the peak and integrated flux densities in units of mJy 1 beam− and mJy respectively; column 5: the angular size of the source in arcsec where U denotes an unresolved source. The flux densities and source sizes have been estimated as described in Sirothia et al.(2009a). . 102

B.1 CandidateDDRGs ...... 115 B.2 Non-DDRGs...... 118 B.3 Candidate DDRGs from the FIRST survey classiﬁed as WATs...... 123 Chapter 1

Introduction

Galaxies, the basic building blocks of the Universe, are large gravitationally bound systems of stars, interstellar medium (ISM), stellar remnants, dark matter and can have spiral, elliptical and irregular morphologies. Nearly every nearby galaxy appears to have a supermassive black hole ( 106 to109 M ) at its center with the mass of the black holecorrelated ∼ ⊙ with some of the properties of the host galaxy such as the spheroid or bulge luminosity and mass. Amongst the galaxies harboring a super-massive black hole, a small fraction emit an extraordinary amount of radiation across the entire band of the electromagnetic spectrum. These galaxies are known as ‘active galaxies’ and their highly luminous nuclei as ‘active galactic nuclei’ (AGN). Approximately 10% of AGNs are luminous at radio wavelengths, referred to as radio loud objects. The central AGN accretes surrounding matter in form of an accretion disk and generates highly collimated jets. The jets contain highly energetic particles, and the radio emission observed from these jets is due to the synchrotron mechanism. These galaxies, are known as radio galaxies, and have lobes of radio emission caused by the radio jets as these travel outwards. The projected linear sizes of the radio galaxies can range from less than a few parsec for the young sources where the lobes are conﬁned in the central region of the host galaxy to over several Mpc for the oldest objects which could be several times 108 yr old. Sometimes these radio galaxies

1 2 1. Introduction are found to be accompanied by a new pair of radio lobes sharing the same central AGN.

The recurrent AGN activity provides two or more pairs of radio lobes and such objects are called double double radio galaxies (DDRGs). In this chapter we give a brief introduction to AGNs (Sect 1.1) and their classiﬁcation (Sect. 1.1.1). The radio galaxies and their properties which have been discussed in the literature are brieﬂy discussed in Sect.

1.2. We also introduce the sources having recurrent activity and their evolution process (Sect. 1.3). The spectral ageing analysis for large radio sources, the assumptions, synchrotron radiation loss mechanisms and three models which have been used for this study are brieﬂy discussed in Sect. 1.4. Finally issues investigated in this thesis are given in

Sect. 1.5.

1.1 Active galactic nuclei :

The hidden mysteries at the centers of galaxies have been revealed slowly with new generation telescopes. Recent studies probed that a supermassive black hole, the fundamental energy source for AGN, resides at the center of almost all galaxies (Saikia & Jamrozy, 2009). The mass accretion of the black hole and energy dissipation through viscous processes in the ‘accretion disk’ shows spectacular emissions throughout the electromagnetic spectrum from radio to γ-rays. The high density gaseous plasma cloud or the ‘broad line region (BLR)’ located at 0.001 - 0.1 pc from black hole (See Fig. 1.1) gives UV and op- ∼ tical emission. The spectra of BLR in UV and optical show broad emission lines ranging

1 from 1000 to 10000km s− . The low-density gas clouds or the ‘narrow line region (NLR)’ far ( 1pc to 100pc) from the black hole gives narrow emission lines of widths of 100 ∼ ∼ 1 to 1000 kms− . These emission lines are important tools for identiﬁcation of the nature of AGNs. An optically-thick dusty torus in between BLR and NLR region obscures our view of the AGN when the line of sight passes through it. This torus extends from few parsec to a few hundred parsec from the central black hole. Recent X-ray studies revealed 1.1 Active galactic nuclei : 3 that this absorbing medium is not combination of uniform gas and dust distribution but have some clumpiness in its structure (Bianchi et al., 2012). Along the poles of the torus or accretion disk collimated streams of plasma in the form of jets ﬂows out and forms kpc- to Mpc-scale structures (Begelman et al., 1984; Urry & Padovani, 1995).

1.1.1 Classiﬁcation of AGNs:

According to luminosity, spectra and appearance active galaxies are classified into the following types: Seyfert galaxies, which have a spiral host and fluctuations in brightness at their cores; radio galaxies, characterized by extended radio lobes on opposite sides of the central AGN; quasars, (‘quasi-stellar objects’ or QSOs), which are very compact, most luminous class of AGN and have usually broad emission lines in optical spectra; BL Lac- ertae objects, which shows continuum variability at all wavelengths, have week (or do not have) emission lines. These further have been subclassified based on their relative flux density in radio to optical bands and their optical spectrum.

Radio loud and radio quiet AGN: On the basis of radio loudness the AGNs are classified in two main groups: 1) radio-loud and 2) radio quiet. For radio-loud AGN the ratio of radio flux at 6cm (5GHz) to the optical B-band (4400Å) flux is 10 (Keller- ≥ mann et al., 1989) whereas for radio-quiet this ratio is <10. Generally, the host galaxies of radio-loud AGNs are elliptical, while radio-quiet AGNs are hosted by spiral galaxies (Antonucci, 1993). However, QSO host galaxies could be either elliptical or spiral, and are often found to have close companions.

Type 1 and Type 2 AGNs: According to the spectroscopic properties of optical wavebands, AGNs are broadly classiﬁed in two groups: 1) Type 1 AGNs and 2) Type 2 AGNs. The Type 1 AGN shows broad as well as narrow emission lines generated from the 4 1. Introduction

Figure 1.1: An AGN cartoon; The green arrows represent the type of AGN which can be seen from a certain viewing angle according to the uniﬁcation scheme. (Image credit: NASA) 1.1 Active galactic nuclei : 5

Table 1.1: AGN classiﬁcation

Type Type 2 Type 1 Type 0 narrow line broad line unusual Radio loud NLRG BLRG BL Lac OVV FSRQ Radio quiet Seyfert 2 Seyfert 1 QUASAR The angle to the line of sight is decreasing from left to right.

BLR and NLR regions, while Type 2 AGN shows only narrow emission lines produced by NLR regions. However most of these objects may be intrinsically similar, with the observed diﬀerence being due to orientation eﬀects.

Orientation based unified models of AGNs: Orientation plays a major role of classification of AGNs. The basic idea of unified model is that Type 1 and Type 2 AGNs are intrinsically the same class of object but appears as different objects because of the orientation effect (Bianchi et al., 2012). AGN properties strongly depend on the viewing angle of the observer. When viewed from the large angle >45◦ from jet axis the central portion of AGN is entirely hidden by the obscuring dusty torus, only the NLR can be seen. Therefore, only narrow emission lines can be seen and the AGN appears as Type 2. This

Type 2 AGNs include Seyfert 2 galaxies at low luminosities and narrow line radio galaxies (NLRGs). As the line of sight gets closer to the jet axis the AGN becomes visible and the BLR can be seen directly and the object is classiﬁed as Type 1 AGN. Seyfert 1 galaxies, QUAsi-StellAr-Radio Sources (QUASARs) and broad line radio galaxies (BLRGs) are

Type 1 AGN. If the line of sight falls within the beaming cone of the relativistic jet near

0◦, the object is classified as BL Lacertae (BL Lac) objects, optically violently variable (OVV) quasars and flat spectrum radio quasars (FSRQ). BL Lac objects, OVV quasars and FSRQ are classified as Type 0 AGN. The class of AGNs are listed in table 1.1. 6 1. Introduction

1.2 Radio galaxies:

The pioneering radio interferometry by Jennison & Das Gupta (1953) revealed Cygnus A as a double-lobed radio source. Since this discovery, studies of such exciting powerful radio galaxies have been made for a large number of such objects. These are powerful radio sources, a subclass of the general AGN zoo, with total radio luminosity 1041 to 1046 erg ∼ 1 s− . It is widely believed that large radio galaxies represent the late stage of evolution of radio sources which have evolved from the compact powerful gigahertz peaked-spectrum (GPS; peak near 1 GHz; linear size < 1 kpc) and compact steep spectrum (CSS; peak near 100MHz; linear size < 15 kpc) radio sources. Radio galaxies consist of two oppositely located lobes of magnetic ﬁeld and synchrotron emitting relativistic plasma. These lobes are connected by relativistic jets ﬂowing out in opposite directions from the AGN. Most of the radio galaxies are associated with an elliptical host galaxy (there are some exceptional cases of spiral-hosts, e.g. Hota et al. 2011) and have bright emission lines which help in easy measurement of their cosmological redshift.

The widely accepted basic model (see Fig. 1.2) of formation and evolution of a radio galaxy was developed by Blandford & Rees (1974) and by Scheuer (1974). The main four components of radio galaxy are: a central core, radio jets, hotspots and lobes. The radio core associated with the nucleus of the host galaxy is compact, has a flat spectrum and its linear size is less than 0.1 pc and is unresolved on the VLBI scale. Ac- ∼ cording to the model, the relativistic fluid in the form of jet flows outwards in opposite directions of the central component or nucleus of the host galaxy. The jet, collimated streams of plasma, propagates through the inter-stellar medium (ISM) and later through the inter-galactic medium (IGM) and forms a cavity or cocoon. Observationally it has been seen that these jets may be one sided or two sided. One sided jets are quite common for radio quasars. Due to the effect of environment, the jets may not be smooth and some knots can be seen in the jet structure. The typical spectral index of these jets are 0.6. ∼ 1.2 Radio galaxies: 7

Some sources do not show jets but bridges of emissions in between core and lobes. The hotspots form at a certain point where the flow of jet is terminated. The linear size of a hotspot is usually less than 1 kpc, and its typical spectral index is in the range 0.5<α<1, ∼ ∼ ∼ flatter than the spectral index of the lobes. The electrons get accelerated at the hotspots and flow out to form the extended lobe emission. The cocoon expands into the external medium and creates a bow shock. A thin boundary of shocked interstellar or intergalactic medium forms just behind this bow shock and separates cocoon from external medium. Usually the linear size of a large source varies from hundreds of kpc up to few Mpc (e.g. largest radio source J1420 0545; linear size 4690 kpc; (Machalski et al., 2008)) and − shows steep radio spectra (α>1). ∼

According to the morphology, radio galaxies are divided into two groups Fa- naroff & Riley (1974) Class I (FRI) and Class II (FRII). For low-luminous (P 0.5 178 ≤ × 25 1 1 10 W Hz− sr− ) FRI radio galaxies, the outflowing jets are more dissipative in nature and this results in bright jets near the core and the edges dark (see Fig. 1.3) without any prominent hot-spots at the outer edges. Whereas for the high-luminosity (P > 0.5 178 × 25 1 1 10 W Hz− sr− ) FRII radio galaxies, jets flow without significant dissipation and create bright hotspots at the edges (see Fig. 1.4 ). The sub-classes of FRI are: narrow angle tails, wide angle tails, fat doubles, twin jet sources, small twin jets etc. The large-scale jets of FRII radio galaxies are more asymmetric on opposite sides of the parent galaxy compared with FRI objects, and this is believed to be due to their higher bulk speed. The apparent asymmetry of the jets is largely due to the effects of relativistic beaming. Some objects show both the FRI and FRII radio morphologies, these are called ‘HYbrid MOrphology Radio Sources’, or HYMORS (Gopal-Krishna & Wiita, 2000). 8 1. Introduction

Figure 1.2: Model of radio galaxy.

Figure 1.3: 5 GHz image of the FRI radio galaxy NGC 4261. Credit: Teddy Cheung. 1.3 Double-doubleradiogalaxies(DDRGs): 9

Figure 1.4: 5 GHz image of the FRII radio galaxy Cygnus A. Image courtesy of NRAO/AUI.

1.3 Double-double radio galaxies (DDRGs):

One of the important aspects of observations of radio galaxies is to understand the duration of the active phase of AGN and its episodic nature. The central AGN activity may not be continuous during the whole lifetime of the radio source. One AGN can go through two or more episodes of activity. The time scale of such episodic activity is of the order of 106 to 108 yr for a large sample of radio galaxies. Sometimes radio observations of extended radio sources show a new pair of radio lobes, well separated from pre-existing relics of earlier cycle of AGN activity (e.g., Saikia et al., 2006) or well embedded inside the old faded lobes (e.g., Machalski et al., 2010). These sources have been christened as ‘double-double’ radio galaxies (DDRGs) by Schoenmakers et al. (2000b). In this context it is relevant to note that the inner double can be surrounded by old diﬀuse relic emission which may not have exactly classical double like structure (e.g., 4C29.30; Jamrozy et al., 2007). In DDRGs the newly-formed jets propagate outwards through the pre-existing cocoon rather than the IGM or ISM. In most of the DDRGs (see Fig. 1.5) the inner doubles 10 1. Introduction

Figure 1.5: DDRG J1158+2621 Image credit: Owen & Ledlow (1997) are well aligned with the outer doubles. The linear size of the inner double varies from few pc to kpc range where as the linear size of the outer double varies from few hundred kpc to Mpc range. Till date a few dozen good examples of recurrent activities of AGNs have been identiﬁed. For a detail description of these sources see Saikia & Jamrozy (2009).

It is almost certain now that the formation of DDRGs are due to interruption of central jet production mechanism. Most of the time the outer lobes of the DDRGs do not show any hotspots where as prominent hotspots can be seen in the newly formed young doubles. Absence of hotspots in the outer lobes implies that the jet activity due to earlier cycle has been interrupted. Restarting of new jet activity creates hotspots in the young lobes. The evolution process of such unusual sources can be probed through their structural and spectral information of extended emission. Study of DDRGs is not only required to understand the recurrence and time scales of AGN activities, but also 1.3 Double-doubleradiogalaxies(DDRGs): 11 provides information about the jet production process and its interaction with external environment (Kaiser et al., 2000; Saikia & Jamrozy, 2009). Studies of these sources provide information of several unanswered questions about the exact mechanism of jet disruption, accretion history of the central black hole, mechanism for precession of ejection axis.

The synchrotron emission from old faded ‘fossil radio lobes’ (Jones & Preston, 2001) can have very steep-spectrum. So, to identify new samples of DDRGs, a low- frequency radio survey is needed. Few high frequency as well as low frequency investigations done for wide ﬁelds contains several active galaxies or clusters of galaxies (e.g. Jones & Preston, 2001; Sirothia, Saikia, Ishwara-Chandra, & Kantharia, 2009b). But no striking example was found even in deep low-frequencies done by Sirothia et al. (2009b).

This suggests that such events are rare and their detection is very diﬃcult not only at high frequencies but also at low-frequencies.

There are several scenarios about the restarting jet activity of the DDRGs, but all these are not well understood. The extensively discussed model of formation of DDRGs is the merger of super massive black hole. The interruption and restarting of the jet formation in DDRGs may triggered by a galaxy merger (Liu et al. 2003). The presence of binary black hole systems within the nuclei of active galaxies supports this scenario (Gopal-Krishna et al. 2012) strongly. A recent study by Kozieł-Wierzbowska et al. (2012), represents that the radio galaxy associated with optical host CGCG 292 057 shows a − recurrent activity in the form of a double double morphology and it is identiﬁed with merging galaxies. Accretion disk instability (Schoenmakers et al., 2000b) is also a strong scenario for this mechanism, but still unclear. One of the strongest scenarios is accumulation of large gas cloud in the center of the galaxy. The fuel or large gas cloud may interrupt the jet activity and restart new episodes (Saripalli & Mack 2007). To ﬁnd out whether the acquisition of fuel may be responsible for recurrent activity, CO and HI line observations have been done in the central regions of DDRGs (Saripalli & Mack 2007; 12 1. Introduction

Saikia et al. 2007). Saripalli & Mack (2007) looked for CO gas in the central regions of nine recurrent radio galaxies, but except for 3C293 which is known to be rich in both atomic and molecular gas, there were no new detections of CO gas in their sample of DDRGs. Saikia, Gupta, & Konar (2007) reported the detection of H gas in the central region of the DDRG J1247+6723, and suggested that H is seen more frequently with complex line proﬁles towards DDRGs compared with CSS and GPS objects where H is seen in about a third of the radio galaxies. Subsequent observations with both the Giant Metrewave Radio Telescope (GMRT) and the Arecibo telescope (e.g. Chandola, Saikia, & Gupta 2010; Salter et al., 2010) seem to support this suggestion although it is important to investigate this systematically for a large sample of sources. Saikia & Jamrozy (2009) also suggest more samples are needed to conﬁrm possible relationship between CO and H absorption and rejuvenation.

1.4 Spectral ageing analysis:

The radio continuum synchrotron spectra in the different regions of the extended radio emission are related to the energy gains and losses of the radiating electrons, diffusion of electrons and injection of electrons. In classical double radio galaxies the lobes represent backflow from the hotspots where they were last accelerated. The resultant radio spectra for these sources steepen from hotspots towards the cores. These spectral gradients along the lobes indicate that the particles near to the core are older than those closer to the hotspots (e.g. Myers & Spangler 1985; Alexander & Leahy 1987). So the spectral index variation gives important information on evolution process of radiating particles in radio galaxies. The theory of spectral ageing analysis is well described by a number of authors (Kardashev 1962, Pacholczyk 1970). 1.4 Spectral ageing analysis: 13

Assumptions of spectral ageing analysis: The basic assumptions for spectral ageing analysis are mentioned below: 1) The magnetic ﬁeld strength of the source (B) is constant throughout the source and equal to the equipartition ﬁeld.

γ 2) The spectrum of initial electron injection is N(E)dE = N0E− dE. The index of the power law is constant over time. 3)There is no in situ reacceleration within these lobes and no signiﬁcant mixing of particles between two adjacent strips.

Synchrotron radiation loss mechanism: It is well known that radiation from a power-law energy distribution of electrons gives a standard power-law continuum syn-

γ α chrotron spectrum. If N(E) E− , then observed ﬂux density, S(ν) ν− , where γ, the ∝ ∝ slope of the powerlaw distribution of the energy of the radiating electrons is related to the spectral index α by α=(γ 1)/2. It is often observed that continuum spectra curve − downwards at higher frequencies. This is because the rate of loss of energy is proportional to the square of the energy, dE/dt B2E2, while the loss timescale, E/(dE/dt) is ∝ inversely proportional to the total energy. Thus the high energy electrons radiate faster than lower-energy ones. Initial power law having injection spectral index αin j will steepen at a particular frequency (νbr) which is called ‘break frequency’. The (νbr) is related to the synchrotron age through the equation:

1/2 B 1/2 τ = 50.3 ν (1 + z) − [Myr], (1.1) spec 2 2 br B + BiC { }

2 Here BiC=0.318(1+z) is the magnetic ﬁeld strength equivalent to the cosmic microwave background radiation. The magnetic ﬁeld strength B and BiC are expressed in units of nT, 14 1. Introduction the spectral break frequency in GHz.

For spectral ageing analysis both low-frequency and high-frequency observations have several important roles. The low-frequency measurements of injection index gives best reliable value since energy loss rate is very less in this regime. Compact radio sources also show spectral turnover at low frequencies because of synchrotron self absorption. Through low frequency data we can easily probe whether there is any low- energy cutoﬀ for the sources (Carilli et al., 1991, Hardcastle, 2009). On the other hand high frequency steepening due to synchrotron radiation loss varies in a wide range. The spectrum also depends on particular position on the source. So, for spectral ageing analysis similar-resolution observations over a broad range of frequencies are required.

Three models for synchrotron ageing: The radio spectrum produced by en- semble of electrons evolve with time due to synchrotron radiation loss. The spectral index variation of these sources plays an important role for testing the models of evolution of relativistic particles. Therefore we consider three models for the evolution of the energy spectrum of the electrons with time (i) Kardashev-Pacholczyk model (KP; Karda- shev 1962; Pacholczyk 1970), (ii) Jaﬀe-Perola model (JP; Jaﬀe & Perola 1973) and (iii) Continuous Injection model (CI; Kardashev 1962; Pacholczyk 1970).

For all three models, we consider a constant magnetic field and no expansion effect; the emission spectrum from electron distribution initially follows a power law with an injection spectral index αin j. The spectrum will steepen at break frequency where this initial power law is no longer maintained. This break frequency relates the age of the source through the above mentioned relation. Since the whole spectral shape cannot be derived analytically, we computed it numerically through these models. By fitting the 1.5 Issues investigated here: 15 spectral data to the numerically computed spectrum (KP, JP, CI) we obtain the break frequency for the source. For this we used a fitting algorithm SYNAGE (Murgia et al., 1999), where the break frequency, injection spectral index and the normalization are kept as free parameters.

The JP and KP models consider one-time injection of the power-law distribution of electrons and no further particle injection. The KP model considers the pitch-angle of the electrons are constant throughout their lifetime whereas in the JP model the pitch angles are scattered and isotropized on time scales much smaller than the radiative life times of the electrons. Due to rapid isotropization sharper break in the energy spectrum can be seen in JP model. The CI model considers a continuous injection of a power law distribution of fresh electrons. Basically CI model developed for unresolved sources (Carilli et al., 1991; Murgia et al., 1999). The JP and KP models work ﬁne for a discrete element of plasma which has been injected in a time scale small compared to its life time.

1.5 Issues investigated here:

The thesis consists of a study of spectral ageing analysis of a few large radio sources, including those which show signs of episodic activity. We also explore and identify new candidates in the latter category, especially of smaller sizes than those which have been studied so far. Since the evolutionary stage of a radio source can be probed through their structural and spectral information of extended emission, I have made high-quality radio images over a large frequency range for several sources with a range of overall projected linear sizes. Such a study provides information about their spectral ages, injection spectral indices, and explores their consistency with models of radio sources. The existence of DDRGs is evidence that the jet activity in AGN is not continuous during the whole life- 16 1. Introduction time of a source. One of the important issues concerning these DDRGs is the duty-period of AGN. Through the multi-frequency observations of DDRGs, we determine the spectral ages of the diﬀerent components and thereby constrain the time scales of episodic activity. From the previous discussions and reviews it has been found that the time scale of episodic activity and the average projected linear size is large for most of the DDRGs. In this study I examine whether large sizes and large time scales are required for all episodic candidates. For this purpose I used low-frequency as well as high-frequency radio data from diﬀerent telescopes. The low-frequency radio observations have been performed using the Giant Metrewave Radio Telescope (GMRT) whereas for the high frequencies, the Very Large Array (VLA) archival data have been used. The following chapters of the thesis explore the above research themes. Chapter 2

Data Set and Analysis Procedure

My research interests broadly have been spectral ageing analysis of large radio sources, their episodic activities and to explore new double double candidates. For this purpose low frequency as well as high frequency radio data has been acquired from different telescopes. Low frequency (150 MHz, 240 MHz, 325MHz, 610MHz) radio observations have been performed using the Giant Metrewave Radio Telescope (GMRT) whereas for the high frequencies (1.4 GHz, 4 GHz, 8GHz), the Very Large Array (VLA) archival data have been used. In this chapter we briefly describe the configurations of radio telescopes (Sect. 2.1), the principle of aperture synthesis (Sect. 2.2) and the data reduction techniques that has been adopted for analysis (Sect. 2.3).

2.1 Radio Telescopes:

There are two dominant windows (see Fig. 2.1), radio ( 1cm to 10m ) and optical ∼ ( 0.366µm to 4.5µm) wavebands, where earth’s atmosphere is transparent and suitable ∼ for ground based astronomy. It is also true that the width of the radio window truncated due to earth ionospheric activity at 10m and 1cm due to water vapor contained in ∼ ∼ 17 18 2. Data Set and Analysis Procedure

Figure 2.1: Windows in the Earth’s atmosphere. air. Radio emission from large radio sources is produced via non-thermal synchrotron mechanism, in which relativistic electrons get accelerated in strong magnetic ﬁelds. Such phenomena occur in very powerful sources like supernova remnants, radio galaxies etc.

It was the year 1931, when people started to explore the so called “invisible universe”. Karl G. Jansky (1905-1950) came forward to build up an antenna to ﬁnd out the noise producing radio sources in communication systems. Existence of some extrater- restrial radio sources were found during these observations and through extensive study Jansky concluded that the signal is originating from our own galaxy, Milky Way. This discovery conﬁrms that the radio signal emitted by celestial objects can be received on the earth surface with a new kind of “eye” - the radio telescopes.

The basic components of radio telescope is: a large radio antenna and a radio receiver. The most common radio antennas are parabolic reﬂector or dish. A single dish radio telescope consists of a parabolic reﬂector which converges the incoming radio signal 2.1 Radio Telescopes: 19

Figure 2.2: Components of a single dish radio telescope

at its focal point where “feed” is placed. The electric current generated in feed is further ampliﬁed, ﬁltered and recorded (see Fig. 2.2). Some of the important single dish radio telescopes are Arecibo single dish (diameter of the dish is 305 meters) , The 64 meter radio telescope at Parkes Observatory, RATAN-600 located near Nizhny Arkhyz, Russia and 100 metre Green Bank Telescope (GBT). 20 2. Data Set and Analysis Procedure

The angular resolution of a telescope is given by diﬀraction limited Rayleigh cri- terion θ λ/D where λ is the observed wave length and D is the aperture of the telescope. ∼ In small wavelength regime, like optical, telescope with even small aperture can provide a good resolution. For example , in optical V band (λ 5450 Å), our eye, which is also a ∼ telescope of aperture roughly 6mm can resolve two sources having angular separation ∼ ∼ 19 aresec. So to achieve similar resolution in radio wavelengths a gigantic dish is needed. This is not practical. Interferometry is a technique which combines several radio antennas to synthesize the aperture of a large single antenna. Some of the major interferometer telescopes are Giant Metrewave Radio Telescope (GMRT), Vary Large Array (VLA), Low

Frequency Array (LOFAR), Australia Telescope Compact Array (ATCA) and Westerbrok Synthesis Radio Telescope (WSRT). The Square Kilometer Array (SKA), which is aimed to have a large frequency coverage, better resolution and collecting area of a square kilometer is under construction. The resolution of a single dish at low frequencies is not enough to resolve the extended emission of large radio sources. High resolution interferometric observations as well as sensitive low frequency observations are appropriate for our samples.

2.2 Interferometry principle:

Imaging of any celestial object in radio wave lengths is based on the principal of aperture synthesis. Instead of a large telescope radio astronomers follow a special technique called interferometry. Here a very large area is covered by several small antennas which act coherently like a big telescope. Interferometric arrays are collection of two element interferometers. The observed quantities in a radio-interferometric observations are the complex visibilities V(u, v, w) and is related to the sky brightness distribution I(l,m) by the relation: 2.2 Interferometry principle: 21

Figure 2.3: Very Large Array, an interferometric array. (Image courtesy of NRAO/AUI) 22 2. Data Set and Analysis Procedure

2πι(ul+vm+w( √1 l2 m2)) dldm V(u, v, w) = I(l, m)B(l, m)e− − − , (2.1) Z Z √1 l2 m2 − − Here (l,m) refers to the direction cosines in the sky and B(l,m) is the far field antenna reception pattern. For a radio interferometer u, v, w are referred as ‘baselines’ and are the projected antenna separation vectors at a plane perpendicular to the line of sight of observation in terms of the observing wavelength. The above relation is known as the van-Cittert Zernike theorem. For small field of view (l2 + m2 << 1) and small values of w the visibilities V(u, v) can be approximated as the Fourier transform of the source brightness distribution I(l,m). In the approximation of a constant sky, i.e, when the sky brightness distribution I(l,m) is not a function of time, then image can be reconstructed from the visibilities recorded at a finite number of (u,v,w) points. Usually an radio-interferometer consists of several antennas (say N antennas) with each pair record-

N ing one Fourier component. Hence at any given time instant only C2 Fourier components are recorded. Since the projected separation of a pair of antenna along the plane perpendicular to the line of sight to the observed direction changes with earth rotation, the number of discrete samples of visibilities observed in an observation with a moderate time span covers a considerable volume of the (u, v, w) space and hence the sky brightness distribution can be reconstructed. This method of synthesizing a large aperture telescope even with a ﬁnite number of antennas using earth rotation is known as “earth rotation aperture synthesis” and is widely used by all the radio-interferometers (Ref : Low Fre- quency Radio Astronomy, Course Notes from a School, NCRA, 1999. eds by Jayaram N. Chengalur, Yashawant Gupta, K.S. Dwarakanath, May 2003.).

Imaging at low frequency using interferometer provides information about the extended emission of large radio sources. GMRT plays a signiﬁcant role for studying large radio sources because of its high resolution as well as sensitivity to the large scale 2.2 Interferometry principle: 23 emission. The diﬀuse emission from the plasma (greater than 108 yr old) of the earlier ∼ cycle of activity can be detected more easily via low frequency observations because of their steep spectrum. GMRT low-frequency observations make it possible to explore new candidates with episodic activity. We used GMRT observational data for low frequency and available archival VLA data for high frequency imaging. These two interferometers have been described below.

2.2.1 Radio-Interferometer: Giant Metrewave Radio Telescope (GMRT)

It is an interferometric array of thirty alt-azimuth mount fully steerable parabolic dishes. GMRT (Swarup et al., 1991) is located about 80 km north of Pune, India (longitude=

74◦.05 E, latitude = 19◦.092 N ). It is an interferometric array of thirty alt-azimuth mount- able parabolic dishes. Each antenna is 45 m in diameter and fully steerable. Parabolic reflecting surface of the GMRT antennas are constructed utilizing the concept of Stretch Mesh Attached to Rope Trusses (SMART), which helped keep the antenna weight and wind loading considerably low. GMRT operates at a range of frequencies from 50 MHz to 1450 MHz. The hybrid array configuration of GMRT (see Fig. 2.4) fulfill both the requirements of high angular resolution as well as detection of diffuse extended radio emission at lowfrequencies. There are 14 antennas in the central square region distributed randomly. These central square antennas provides a large number of short baselines which helps to image large extended emission. The remaining 16 antennas are distributed in three Y shaped “East” “West” and “South”arms. The maximum baseline length produced by arm antennas is nearly 25 km. For high resolution image these arm antennas play a significant role. The system parameters of GMRT are given in table 2.1. 24 2. Data Set and Analysis Procedure

Figure 2.4: GMRT full array (left panel) and central array (right panel) conﬁguration.

2.2.2 Radio-Interferometer: Very Large Array (VLA)

We used VLA (Thompson et al., 1980) high frequency archival data extensively. It is situated at the plains of San Agustin, west of Socorro, New Mexico (latitude=

34◦04′43.497′′N, longitude= 107◦37′03.819′′W). Each antenna is 25 m in diameter and the weight of each solid dish is 230 tons (see Fig. 2.3). Location of these antennas can be changed to arrange them in four different array configurations. The maximum antenna separations are of 36 km, 10 km, 3.6 km and 1 km for A array, B array, C array and D array respectively. These configurations change in every four months or so. The important parameters are presented in the table 2.2.

Table 2.1: System parameters of GMRT Frequencies(MHz) 50 153 233 327 610 1420 Primary Beam (degrees) planned 3.8 2.5 1.8 0.9 0.4 Full array (arcsec) planned 20 13 9 5 2 CentralSquare(arcmin) planned 7.0 4.5 3.2 1.7 0.7 SystemTemperature(K) planned 482 177 108 92 76 Gain of an antenna(K/Jy) planned 0.33 0.33 0.32 0.32 0.22 rms noise(µ Jy) planned 46 17 10 9 13 2.3 Analysis techniques of the radio-interferometric data: 25

Table 2.2: System parameters of VLA Band 4 Band P Band L Band C Band X Band U Band K Band Q Band Freq(GHz) 0.07-0.074 0.30-0.34 1.34-1.73 4.5-5.0 8.0-8.8 14.4-15.4 22-24 40-50 Wavelength(cm) 400 90 20 6 3.6 2 1.3 0.7 Primarybeam(arcmin) 600 150 30 9 5.4 3 2 1 Highestresolution(arcsec) 24.0 6.0 1.4 0.4 0.24 0.14 0.08 0.05 SystemTemp 1000-10,000.K 150-180.K 37-75.K 44.K 34.K 110.K 50-190.K 90-140.K

2.3 Analysis techniques of the radio-interferometric data:

Astronomical data are usually archived in a universally accepted common format, namely the “Flexible Image Transport System” or FITS. However, because of different technical reasons the direct recorded data for different interferometer may vary. During the tenure of this thesis several observations were performed using the GMRT which records the raw visibility in the observatory specific long term accumulation or lta format. Observatory standard software namely listscan and gvfits are used to transform the recorded data in the

FITS format. GMRT hardware backend (GHB) was the main backend till end of 2010. The Upper Side Band (USB) correlator and Lower Side Band (LSB) correlator of GHB produce two separate ﬁles with extension .lta and .ltb respectively. Each ﬁle contains 128 spectral channel. Now GMRT started a new GMRT Software Backend (GSB) system.

This produces single lta ﬁle and can have 256 or 512 channels.

In principle the spatial correlation function or the visibility, V(u,v), measured across several baselines using an interferometer can be inverse Fourier transformed to reconstruct the sky brightness distribution I(l,m). However, the measured visibilities are affected by man made interferences and scaled by the antenna gains. Because of these a series of analysis techniques has to be applied to them before they are ready to be used as scientific data. These procedures can broadly be classified in three different groups, here we give a very short description of them here. We shall discuss the details of algorithm used to perform these operations and the implementation of those later. 26 2. Data Set and Analysis Procedure

Flagging: Any unwanted signal, mainly man made, received by the radio antennas are known as the Radio Frequency Interference or RFI signal. RFI’s are often several orders of magnitude stronger than the astronomical signal and hence they corrupt the visibility data completely. However, these signals carry a diﬀerent statistical property compared to the astronomical signal and hence can be identiﬁed. The process of identifying the RFI’s from the recorded data and tagging them as bad data is termed as FLAGGING. This must be done at the very initial stage of the radio interferometric data analysis.

Calibration: As mentioned earlier, the recorded visibility across a baseline in an interferometer is the spatial correlation at these baselines multiplied by the complex gain of the telescope. Hence, the raw visibilities has to be calibrated for the telescope gains and then can be used for scientiﬁc purposes. The process of determining the interferometer gains and using them to obtain the spatial correlation function is known as Calibration.

Deconvolution and Imaging: In order to reconstruct the sky brightness distribution or the ‘image’ of the observed sky, it is required to have a continuous measurement of the visibilities in the (u, v) plane. However, because of the sparse distribution of the antennas, for most of the radio interferometers visibilities are measured only at some discrete points. This effect of incomplete coverage of the (u, v) plane in terms of visibility measurements introduce a nonlinear affect on the image. Different algorithms are applied to deconvolve the effect of the incomplete (u, v) coverage of the interferometer from the image. This final step is Deconvolution and Imaging. There are several implementation of the image processing algorithms required to perform the above steps in terms of different softwares. We have used the Astrophysical Image Processing System (AIPS) developed by the National Radio Astronomical Observatories (NRAO) to perform all the data analysis required for this thesis. AIPS implements different algorithms to perform the above three basic steps and also many others required for scientific analysis required 2.3 Analysis techniques of the radio-interferometric data: 27 subsequently. In AIPS terminology, different implementations of a given step in data analysis are called as “task”. The parameter inputs required for a task are called “verbs” and “adverbs”. In this section we shall discuss the algorithms used to perform flagging, calibration and deconvolution. In parallel to the algorithms the AIPS tasks that implements these steps, the verbs and adverb inputs are given in the text along with them. We also give a list of frequently used utility AIPS tasks, verbs and adverbs.

2.3.1 Pre Processing :

The FITS data are loaded in AIPS in a multi source visibility data base by using the task FITLD. If one starts AIPS from the same directory where data is stored, the ‘environment variable’ which tell AIPS where to look for data is set to present working directory (PWD). If the data or other files which are located in a different directory then before starting AIPS one can have to define an environment variable where the FITS file is stored. A disk area /home/user can be defined as a environment variable in the following way

export T=/home/user (for bash) setenv T=/home/user (for csh)

A typical common set up for all task

>task ‘task name’ >inp (to check the input parameters of a particular task)

>go (runthetask)

In the task ‘FITLD’, following settings are used : 28 2. Data Set and Analysis Procedure

> task ‘FITLD’ > infile ‘PWD:TEST.FITS’ (the fits file ‘TEST.FITS’ kept in AIPS directory) > DOUVCOMP -1 > inp

> go

The uv data is indexed by the task INDXR for the fast access of data set. This task generates one calibration table (CL) and one index table (NX). The CL table is like a template of all antenna gains for diﬀerent calibration tasks. The NX table is for navigating in the data base. The task was run using following settings:

> task ‘INDXR’ > getname 1 (catalogueno. 1istheuvdataﬁle)

> inﬁle ‘’ > go

To get the list of the observed sources task LISTR is used. The output of LISTR is then saved in a ﬁle for future use. The inputs parameters for LISTR are:

> task ‘LISTR’ > getname 1

> OPTYPE ‘SCAN’ > DOCRT -1 > OUTPRINT ‘PWD:source.list’ (Creates a ‘source.list’ ﬁle) > go 2.3 Analysis techniques of the radio-interferometric data: 29

Figure 2.5: An Amplitude vs UV distance plot for 3C286 before editing and calibration (left panel). A Phase vs UV distance plot for 3C236 before data editing and calibration (right panel)

2.3.2 Data editing :

Radio astronomical measurements particularly at low frequencies are affected by RFI. A good quality image is possible only when the data is fully free from RFI. The process of editing the radio data is commonly known as “flagging”. It is a very important part of data reduction. The bad data has to be identified carefully, sometimes one can loose good data because of wrong identification. There are many ways to identify poor quality data.

Mainly UVPLT, TVFLG, VPLOT, tasks are used for identification of the poor data and UVFLG is used for flagging. The defects of data set comes in different form - antenna or baseline based, time based and channel based. Generally we start data editing on the flux and phase calibrators which are point like sources. The visibility amplitude of an unre- 30 2. Data Set and Analysis Procedure solved source or point source remains constant as a function of baseline length. By the task UVPLT we plot visibility amplitude vs UV distance and find out inconsistent data points. Through task VPLOT we can find out the particular antenna or baseline which is responsible for those discrepant points. We generally choose a representative channel and identified the bad data using above mentioned methods and put these information in a FLAG file which is given as an input file to the task UVFLG. If there exists any dead antenna or bad baseline for both flux and phase calibrator through out the whole observation then it has to be removed from entire data set by the task TVFLG as well. Most of the time first few minute data of each scan is of poor quality. These data has to be removed from overall scan using task QUACK.

The inputs of all the tasks used for data editing are given below:

> task ‘QUACK’ > getname 1 > OPCODE ‘BEG’ > APARM (2)= 0.5

> go

> task ‘UVPLT’ > getname 1

> SOURCES ‘name of any source in ﬁts ﬁle’ 2.3 Analysis techniques of the radio-interferometric data: 31

> BCHAN 60 (any good channel)

> ECHAN 60 > DOCALIB -1 > BPARM 0 (amplitudevs uvdistance plotsofall baselinesfor the channel 60) > DOTV 1

> go

> task ‘VPLOT’ > getname 1

> SOURCES ‘name of any source in ﬁts ﬁle’ > BCHAN 60 > ECHAN 60 > DOCALIB -1

> BPARM (3)=-1 (ﬁx the scale of amplitude) > DOTV 1 > go

> task ‘TVFLG’ > getname 1 > SOURCES ‘name of any source in ﬁts ﬁle’ > BCHAN 60

> ECHAN 60 > DPARM (6)=16.9 or 8 (depends on integration time) > go

> task ‘UVFLG’ 32 2. Data Set and Analysis Procedure

> getname 1

> SOURCES ‘name of source’ (no need to put any name if mentioned in the flag file ) > BCHAN 60 (no need to specify any channel if mentioned in the flag file ) > ECHAN 60 > STOKES ‘ ’ (no need to specifyif mentionedin the flag file )

> OPCODE ‘FLAG’ > go

2.3.3 Calibration :

Through calibration, observed visibilities are corrected to the true visibilities. At the time of observation we observe flux calibrator ( 15 minute) in the beginning and at the end ∼ of the observation and the target source observations are interspersed with observations of the secondary phase calibrator. The desirable flux density calibrators have to be unresolved so that they will appear same in all baselines. Furthermore flux density has to be known and be constant with time which helps to find out any systematic atmospheric or instrumental errors (Fassnacht & Taylor, 2001). Inhomogeneous and dynamic earth ionosphere is responsible for rapid change of refractive index with time and position for radio waves. The ionosphere adds an differential phase rotation to the complex visibilities. To calibrate this atmospheric phase fluctuations we need a secondary calibrator which is near ( 20deg) to the target source and whose true visibility is known. In calibration process ≤ observed visibility is to be divided by the true visibility to get the gain factor. This is applied as a correction to the visibilities of the target source. To get the gain values in between calibrator observations interpolation needs to be done. However if a flux calibrator is near by our target source then secondary phase calibrator is not needed.

The calibration process starts with setting the flux density of the flux calibrator 2.3 Analysis techniques of the radio-interferometric data: 33 using the task SETJY. This task follows standard Baars et al. (1977) formulae and takes frequency information from the header to determine the flux densities of primary calibrators. The task CALIB is used for the selected single channel to determine the antenna based complex gains. This creates a solution (SN) table which contains the complex gain solutions. The SN table can be plotted by the task SNPLT to check the gain solutions. If the solutions are not reasonable then one can repeat the same process starting from the flagging. The flux density of the phase calibrator was computed by the task GETJY. This creates a SU table and corrects the amplitude in the SN table. The flux density of the phase calibrator are cross checked with value given by VLA calibrator list. If it differs too much then more flagging is needed for the phase calibrator. The task CLCAL is used to interpolate the gains of the target with time. This creates CL table version two which contains solutions derived from SN table by linear vector interpolation. The amplitude and phase deviations of all the flux and phase calibrator has to be checked applying CL table in UVPLT. If there is any significant deviation then one has to identify the bad data and flag them. Deleting the SN tables and CL table 2 same calibration process has to be repeated again. An Amplitude vs UV distance and phase vs UV distance plot before and after editing and calibration of the flux calibrator 3C286 has been shown in figure 2.5 and

2.6 respectively.

The inputs of all the tasks used for calibration are given below:

> task ‘SETJY’

> getname 1 > OPTYPE ‘CALC’ > SOURCE ‘name of ﬂux calibrator’ > go 34 2. Data Set and Analysis Procedure

> task ‘CALIB’ > getname 1 > CALSOUR ‘name of ﬂux calibrator and phase calibrator’ > BCHAN 60

> ECHAN 60 > UVRANGE if there exists any uv range (check it from VLA calibrator list) > DOCALIB -1 > REFANT give an good antenna

> SOLINT 1 > SOLMODE A&P > go

> task ‘GETJY’ > getname 1 > SOURCES ‘name of phase calibrator’

> CALSOUR ‘name of the ﬂux calibrator’ > go

> task ‘CLCAL’

> getname 1 > SOURCES ‘’ > CALSOUR ‘name of ﬂux calibrator and phase calibrator’ > OPCODE ‘CALI’

> REFANT give an good antenna 2.3 Analysis techniques of the radio-interferometric data: 35

Figure 2.6: An Amplitude vs UV distance plot for 3C286 after editing and calibration (left panel) An Phase vs UV distance plot for 3C236 after data editing and calibration (right panel)

> go

2.3.4 Bandpass calibration :

The complex gain of an antenna can also be different across different frequency channels. This differential complex gain arises because of the bandpass filters in the telescope and hence has to be calibrated out. The method that determines the bandpass shape and cor- rect the observed visibilities based on that is termed as bandpass calibration. To achieve this we follow the following procedure in our analysis. We observe a source with relatively higher flux density at the centre of the field of view (phase center) which is known to have almost constant flux across the bandwidth of observation. This is referred to as the bandpass calibrator for the particular observation. Usually the flux calibrator can be 36 2. Data Set and Analysis Procedure used as the bandpass calibrator if it is known to be free from any absorption or emission across the observed bandwidth. We then determine the complex gain using the usual flux and phase calibration techniques discussed above for any one of the good spectral channels. The spectral dependency of the complex gain or the bandpass is then determined with respect to this calibrated channel by interpolating/polynomial fitting. In practice, in order to increase the signal to noise ratio, a few frequency channels are averaged together. However, since the baseline values at which the visibilities are recorded varies weekly with frequency, averaging too many channels to determining the bandpass solution may introduce unwanted error, usually this is called bandwidth smearing. We determine the appropriate number of channels to average by considering a trade off between maximiz- ing the signal to noise and minimizing the bandwidth smearing. Note that in the entire bandpass calibration procedure we use only those channels which have no RFI. We apply the calibration to the visibility data and visually inspect the gain corrected visibilities to further investigate any remaining RFIs. The entire pipeline of calibration is repeated until there is no obvious RFI present in the data. The calibrated and RFI flagged data is then used for the further analysis. The calibration procedures is described in detail in "Inter- ferometry and synthesis in radio astronomy by By Anthony Richard Thompson, James

M. Moran, George Warner Swenson"(Thompson et al., 1986), interested reader may consider looking at this reference. The task BPASS is used for band pass calibration. BPASS solves both the amplitude and the phase variations across the band and gives bandpass averaged table (BP). We use flux calibrator as a bandpass calibrator and checked bandpass corrected band shapes using task POSSM. Generally the RFI from any channel is flagged using by the task UVFLG, SPFLG, FLGIT. Once the data is free from RFI then the multi source file is splitted in to a single target source file by the task SPLIT. To increase signal- to-noise ratio (snr) channel averaging is done with the same task. The splited data has to be checked by UVPLT, VPLOT, IBLEAD. If any bad data is found, it should be removed 2.3 Analysis techniques of the radio-interferometric data: 37 from the data set. A vector averaged cross-power spectrum of several baselines for 3C286 without and with bandpass has been demonstrated in figure 2.7. Now the calibrated, band- passed, RFI free averaged data is ready for imaging. The inputs of the task used for band pass calibration are given below:

> task ‘BPASS’ > getname 1 > CALSOUR ‘ﬂux calibrator’ > DOCALIB 1

> DOBAND -1 > ICHANSEL 606001 > BPVER -1 > REFANT giveagoodantenna

> go

One has to remove all the RFIs applying BP table on each channel. For this we use the tasks UVFLG and SPFLG. The detail of SPFLG is given below:

> task ‘SPFLG’ > getname 1 > SOUR ‘source name’

> STOKES ‘ ’ > BCHAN 5 > ECHAN 110 > ANTENNAS 0

> BASELINE 0 38 2. Data Set and Analysis Procedure

> DOCALIB 1

> DOBAND 1 > BPVER 0 > DPARM (6)=16.9 or 8 (depends on integration time) > go

> task ‘FLGIT’ > getname 1 > OUTFGVER -1

> SOURCES ‘’ > DOCALIB 1 > DOBAND 1 > BPVER 0

> APARM (1)= (set a maximum allowed amplitude) > APARM (2)= (6-8 times the rms noise for a single channel) > APARM (3)= (maximum allowed residual) > APARM (4)= (5 sigma of RMS)

> APARM (5)= (1.5 times of RMS) > ICHANSEL 20 100 (mentionthegoodchannels) > go

It creates a new UV ﬁle with extension .FLGIT. One should check data by UVPLT whether it has ﬂagged too much data. If it is satisfactory then one can proceed channel averaging with this data set.

The inputs of the task used for split & channel averaging are given below: 2.3 Analysis techniques of the radio-interferometric data: 39

>task ‘SPLIT’

> getname 1 (or the UV data created by FLGIT) > SOURCES ‘name of the target source’ >BCHAN 29(startingwithagoodchannel) >ECHAN 224

>DOCALIB 1 >FLAGVER 0 >DOBAND 1 >BPVER thebestbandpasstable

>APARM 1 >NCHAV 14 (depends on how many channels you want to average) >CHINC 14 >ICHANSEL 0

>go

2.3.5 Imaging :

The visibilities measured by all antennas are all discrete points in uv- plane. Dirty beam is the Fourier transform of uv sampling function or uv tracks during our observation. Dirty map is convolution between real brightness distribution of the source and Dirty beam. The

Dirty beam and the Dirty map comes from Fourier transformation of sampling function and visibilities by fast Fourier transform (FFT) algorithm. So the visibilities is related to the sky brightness distribution as

V ⇋ I,

This equation is valid for continuous uv sampling and ⇋ denotes the Fourier Transform. 40 2. Data Set and Analysis Procedure

30 10 -10 30 C00: - C08: 1 - 8 C00: - C09: 1 - 9

IF 1(RR) IF 1(RR) 3010 10 -10 30 C00: - C10: 1 - 10 C00: - C11: 1 - 11

IF 1(RR) IF 1(RR) 3010 10 -10 30 C00: - C12: 1 - 12 C00: - C13: 1 - 13

IF 1(RR) IF 1(RR) 10 0 20 40 60 80 100 120 0 20 40 60 80 100 120 Channels Channels Lower frame: Ampl Jy Top frame: Phas deg

30 10 -10 30 C00: - C08: 1 - 8 C00: - C09: 1 - 9

IF 1(RR) IF 1(RR) 3010 10 -10 30 C00: - C10: 1 - 10 C00: - C11: 1 - 11

IF 1(RR) IF 1(RR) 3010 10 -10 30 C00: - C12: 1 - 12 C00: - C13: 1 - 13

IF 1(RR) IF 1(RR) 10 0 20 40 60 80 100 120 0 20 40 60 80 100 120 Channels Channels Lower frame: Ampl Jy Top frame: Phas deg

Figure 2.7: Vector averaged cross-power spectrum of several baselines for 3C286 without bandpass (upper panel); Vector averaged cross-power spectrum of several baselines for 3C286 with bandpass. (lower panel) 2.3 Analysis techniques of the radio-interferometric data: 41

Since the interferometric array contains a ﬁnite number of antennas so the uv plane is sampled at discrete uv points. Now the above equation can be written as

V.S ⇋ I DB = DI, ∗

Here DB is the Dirty beam and DI is Dirty map. ‘*’ represents the convolution operator.

The uv sampling function is given by

S (u, v) = δ(u uk, v vk), Xk − −

Here uk and vk are the actual (u, v) points measured by the telescope. The sampling function acts like a delta function

S (u, v) = 1

for all measured (u,v) points.

S (u, v) = 0

every where else.

Now the Dirty image is the convolution of true brightness distribution with Dirty beam. This Dirty image is not a satisfactory ﬁnal image. The true image can be achieved by deconvolution.

2.3.6 The CLEAN Algorithm

One of the widely used deconvolution algorithm is CLEAN (Hogbom 1974). The main assumption of this algorithm is that the sky is composed of point sources and most of 42 2. Data Set and Analysis Procedure the part is empty. At the ﬁrst stage it searches for the peak (greatest absolute intensity) and its position in the dirty image. It adds a fraction g (loop gain) of this point source to the clean component list. Then it subtracts the fraction g times dirty beam from the dirty image. If the residuals are not noise like then it again goes to the ﬁrst step. accumulated point source model is convolved with a Gaussian synthesized beam. At the end it adds the residuals of the dirty image to the ‘CLEAN’ image. Though CLEAN works well for point sources, one would be careful to use this for devolving extended sources.

So far we have considered a small field of view with small value of the w term where one can approximate the van-Cittert Zernike theorem as a two dimensional spatial Fourier transform. At low frequencies, this approximation does not hold true and a direct Fourier transform of the data can not be used. In the case of wide field imaging non-coplanar baselines and spherical curvature of the sky causes the apparent shift in source position. In AIPS we reduce this effect by w-term ( the term in the above equation w √1 l2 m2) correction method. In AIPS task imagr does a 3D inversion using poly- − − hedron imaging algorithm. Here the total field of view is divided into a two dimensional grid of facets. A small part of sky centered around each facet is imaged by shifting the phase center of the visibility to the center of the facet. Then for each facet normal 2D imaging is applied. The size of facets should not be big otherwise distortion can take place at the edge of the facets.

Phase calibrators are observed at regular time intervals. But in between this time gap the atmospheric fluctuations gives an inaccurate phase solutions. This introduces phase error in the visibilities, effect of which the image is not perfect. These spatial or temporal fluctuations of atmosphere can be eliminated with the help of self calibration. In this method first image created by above mentioned methods can be used as a model for its own calibration. The calibration and imaging cycles repeated for several times to refine the calibration process and new input model. Generally phase errors are more com- 2.3 Analysis techniques of the radio-interferometric data: 43 mon than amplitude errors, so we perform four or five rounds self-calibration with phases and one final amplitude self calibration for further improvement of the image quality. The whole process of imaging and deconvolution is done by task IMAGR. The self-calibrated cleaned set of facet images were put together to get the total field of view by task FLATN. Using task PBCOR we get primary beam corrected final image. For more detail information about radio interferometry one can follow Perley et al. (1989). The details of all tasks for 3D imaging and self calibration are given below: Imaging: > task ‘IMAGR’

> getname 2 (catalogueno. 2isthesplitteduvdata) > SOURCES ‘name of the target source > DOCALIB -1 > DOBAND -1

> BPVER -1 > BCHAN 1 > ECHAN 14 (splitted ﬁle contains 14 channel, each of which had 14 original channel) > CHANNEL 0

> NCHAV 14 (the number of channels to be averaged together) > CHINC 0 > OUTNAME ‘test’ > CELLSIZE 2 (the pixel separation in arc second)

> IMSIZE 1024 (thesizeoftheﬁledinpixels) > NFIELD 25 (number of facets) > DO3DIMAG -1 > NITER 500000(themaximumnumber of iteration)

> BOXFILE ‘PWD:TEST.BOX’ (a box ﬁle created by SETFC) 44 2. Data Set and Analysis Procedure

> OVERLAP 2

> DOTV 1 > go The CELLSIZE, IMSIZE depends on the working frequency and for low resolution map one have to use UVTAPER. A boxﬁle can be created using task ‘SETFC’. Here 5X5 grid of facets has been created in ’TEST.BOX’ ﬁle.

Self calibration: > task ‘CALIB’

> getname 2 > CALSOUR ‘’ > BCHAN 1 > ECHAN 0

> DOCALIB -1 > DOBAND -1 > IN2NAME ‘Cleaned map name produced by 1st IMAGR, TEST’ > OUTNAME ‘calib-1’

> SOLMODE ‘P’ (‘P’, phase only; ‘A&P’, amplitude and phase) > go This self calibrated output ﬁle ‘calib-1’ has been used as an input of Task ‘IMAGR’. After some loop of IMAGR and phase self calibration one time amplitude and phase calibration also done. Chapter 3

A multifrequency study of a large radio galaxy 3C46

We present a multifrequency study of a large radio galaxy 3C46 using the Gi- ant Metrewave Radio Telescope (GMRT) and the Very Large Array (VLA). We estimate its spectral age and examine any evidence of diffuse extended emission at low radio frequencies due to an earlier cycle of activity. We find that this source is consistent with a straight spectrum with injection spectral index of 1.0. Our results show the spectral age ∼ increases with distance from the hotspots and the maximum spectral age for the oldest relativistic plasma close to core is nearly 15 Myr. We do not find any evidence of extended ∼ emission due to an earlier cycle of activity.

3.1 Introduction

The linear size of classical double radio galaxy can vary from parsec to megaparsec scale, the largest of which are the giant radio sources (GRSs), deﬁned to be those which have a

1 1 projected linear size 1 Mpc (H =71 km s− Mpc− , Ω =0.27, Ω =0.73, Spergel et al., ≥ o m vac Result presented in this Chapter is published in S. Nandi et al. 2010, MNRAS, 404, 433.

45 46 3.Amultifrequencystudyofalargeradiogalaxy3C46

2003), are useful for studying the late stages of evolution of radio sources and possible episodic activity in these objects, constraining orientation-dependent uniﬁed schemes and probing the intergalactic medium at diﬀerent redshifts (e.g. Subrahmanyan & Saripalli 1993; Subrahmanyan, Saripalli, & Hunstead 1996; Mack et al. 1998; Blundell, Rawl- ings, & Willott 1999 and references therein; Ishwara-Chandra & Saikia 1999; Kaiser

& Alexander 1999; Schoenmakers et al., 2001, 2000c; Singal, Konar, & Saikia 2004). Multifrequency study of large radio sources provides useful insights in understanding the diﬀerent physical processes and stages of evolution. In addition, large radio sources are useful for studying the eﬀects of electron energy loss in the lobe plasma due to inverse-

Compton scattering with the Cosmic Microwave Background Radiation (CMBR) photons at different redshifts (e.g. Konar et al. 2004), making independent estimates of the magnetic field from the inverse-Compton scattered X-ray flux density from the lobes (e.g. Croston et al. 2005; Konar et al. 2009) and spectral as well as dynamical ageing analysis to understand the evolution of the sources (e.g. Konar et al., 2006, 2008; Jamrozy, Konar, Machalski, & Saikia 2008; Machalski, Jamrozy, & Saikia 2009).

We have selected a large radio galaxy 3C46 to make detailed low-frequency radio images with the Giant Metrewave Radio Telescope (GMRT) as well as we used archival Very Large Array (VLA) data to make higher-frequency images. The objectives of this study was to look for diﬀuse emission at low frequencies from an earlier cycle of activity and estimate the spectral ages of the lobes from data over a large frequency range. Com- bining the low-frequency available data with high-frequency data gives the most reliable estimates of the injection spectral indices (αinj), and also spectral ages from the break frequency. For examining emission from an earlier cycle of activity, it is relevant to note that evidence of episodic activity is seen usually in large radio sources (e.g. Schoenmakers et al. 2000c; Saikia et al. 2006, and references therein) but not in the small sources even at low radio frequencies (Sirothia et al., 2009b). 3.1 Introduction 47

3C46 GMRT 153 MHz 3C46 GMRT 240 MHz 37 55 30 37 55 15

00 00

54 45

54 30 30

15 00 00 DECLINATION (J2000) DECLINATION (J2000) 53 30 53 45

00 15

00 01 35 40 35 30 25 20 01 35 38 36 34 32 30 28 26 24 22 20 RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) Cont peak flux = 2.3246E+00 Jy/beam Cont peak flux = 7.7060E-01 Jy/beam Levs = 3.000E-02 * (-1, 1, 2, 4, 8, 16, 32, 64) Jy/beam Levs = 2.793E-02 * (-1, 1, 2, 4, 16) Jy/beam

3C46 GMRT 332 MHz 3C46 GMRT 606 MHz 37 55 15

37 55 00 00

54 45 54 45

30 30

15 15

00 00 DECLINATION (J2000) DECLINATION (J2000) 53 45 53 45

30 30 15 15 01 35 38 36 34 32 30 28 26 24 22 01 35 36 34 32 30 28 26 24 22 RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) Cont peak flux = 7.3355E-01 Jy/beam Cont peak flux = 2.7672E-01 Jy/beam Levs = 3.000E-03 * (-1, 1, 2, 4, 8, 16, 32, 64, Levs = 1.812E-03 * (-1, 1, 2, 4, 8, 16, 32, 64, 128) Jy/beam 128) Jy/beam

3C46 VLA 1465 MHz 3C46 VLA 4841 MHz 3C46 VLA 8460 MHz 37 55 15 37 55 15 37 39 45

00 00 30

54 45 54 45 15 30 30

00 15 15

00 38 45 00 DECLINATION (B1950) DECLINATION (J2000) DECLINATION (J2000) 53 45 30 53 45

30 30 15

15 15 00

01 35 38 36 34 32 30 28 26 24 22 01 35 38 36 34 32 30 28 26 24 22 01 32 42 40 38 36 34 32 30 28 RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) RIGHT ASCENSION (B1950) Cont peak flux = 2.2564E-01 Jy/beam Cont peak flux = 7.5571E-02 Jy/beam Cont peak flux = 2.9726E-02 Jy/beam Levs = 2.361E-03 * (-1, 1, 2, 4, 16, 32, 64) Jy/beam Levs = 8.000E-04 * (-1, 1, 2, 4, 16, 32, 64) Jy/beam Levs = 1.600E-04 * (-1, 1, 2, 4, 16, 32, 64, 128) Jy/beam

Figure 3.1: GMRT low-frequency images of 3C46 at 153, 240, 332 and 606 MHz, and VLA higher-frequency images at 1465, 4841 and 8460 MHz. In this ﬁgure as well as in all the other ﬁgures, the peak brightness and the contour levels are given below of each image. In all the images the restoring beam is indicated by an ellipse. 48 3.Amultifrequencystudyofalargeradiogalaxy3C46

Table 3.1: Observing log Source Teles- Array Obs. Phase Obs. cope Conf. Freq. Calib. Date MHz (1) (2) (3) (4) (5) (6) 3C46 GMRT 153 3C48 2007 Dec 07 3C46 GMRT 240 3C48 2007 Jun 09 3C46 GMRT 332 3C48 2008 Feb 23 3C46 GMRT 606 3C48 2007 Jun 09 3C46 VLAa BnC 1465 2254+247 2000 Mar 13 3C46 VLAa D 4841 2250+143 2000 Jul 24 3C46 VLAa D 8460 2251+158 1998 Jan 24 a archival data from the VLA

The radio galaxy 3C46 (J0135+3754), which has a very close companion galaxy (de Vries et al., 1998), is at a redshift of 0.4373 (Smith & Spinrad, 1980) and has a largest angular size of 150 arcsec which corresponds to 846 kpc, and a total radio luminosity of ∼ 1 log P1.4GHz (W Hz− ) = 27.01 (Konar et al., 2004). VLA B- and C-array images at L-band (Gregorini et al. 1988; Vigotti et al. 1989) and the D-array image at C-band (Konar et al., 2004) show the extended lobes of emission with an edge-brightened structure. de Koﬀ et al. (2000) report evidence of a dust lane lying across the nucleus with the radio axis being nearly perpendicular to it.

3.2 Observations and analyses

Both the GMRT and the VLA observations were made in the standard fashion, with each target source observations interspersed with observations of the phase calibrator. The primary flux density and bandpass calibrator was 3C48 at the different frequencies, with all flux densities being on the scale of Baars et al. (1977). The total observing time on the source is about a few hours for the GMRT observations while for the VLA observations the time on source ranges from a few minutes to 10 minutes. The low-frequency GMRT ∼ data were sometimes significantly affected by radio frequency interference, and these data 3.3 Observational results 49 were flagged. All the data were analysed in the standard fashion using the NRAO AIPS package. For the GMRT observations, besides flagging bad data, the steps followed include gain calibration of one spectral channel data, bandpass calibration and channel averaging to obtain the continuum data base. These were then imaged and cleaned using multiple facets for the different low-frequency GMRT observations (a detail description of data reduction technique is given in Sect. 2.3). All the data were self calibrated to produce the final images, which were then corrected for the gain of the primary beam.

The observing log for both the GMRT and the VLA observations is given in Table 3.1 which is arranged as follows. Column 1 and 2 show the name of the source and the telescope; column 3 gives the array conﬁguration for the VLA observations; column

4 shows the frequency of the observations in MHz, column 5 lists the phase calibrators used for the diﬀerent observations, while column 6 lists the dates of the observations.

3.3 Observational results

The GMRT and VLA images of 3C46 are presented in Fig. 3.1. The observational parameters and some of the observed properties are presented in Table 3.2, which is arranged as follows. Column 1: Name of the source; column 2: frequency of observations in units of MHz, with the letter G or V representing either GMRT or VLA observations; columns

3–5: the major and minor axes of the restoring beam in arcsec and its position angle (PA)

1 in degrees; column 6: the rms noise in units of mJy beam− ; column 7: the integrated flux density of the source in mJy. We examined the change in flux density by specifying different areas around the source and found the difference to be within a few per cent. The flux densities at different frequencies have been estimated over similar areas. Columns 8, 11 and 14: component designation, where W, E and C denote the western, eastern and core components respectively; columns 9 and 10, 12 and 13, and 15 and 16:

1 the peak and total ﬂux densities of each of the components in units of mJy beam− and 50 3.Amultifrequencystudyofalargeradiogalaxy3C46

1000

100

1 10 100 1000

Figure 3.2: The spectra of the extended emission of 3C46 obtained after subtracting the core flux density at frequencies greater than 1400 MHz from the total flux density. Any contributions of the core flux density at lower∼ frequencies are less than 1 per cent and have been neglected. The total flux densities are from Laing & Peacock∼ (1980) and the measurements presented in this paper. The fits to the spectra obtained using the SYNAGE package (Murgia et al., 1999) are also shown. 3.4 Discussion and results 51

Figure 3.3: The spectral-index image of 3C46 with the strips used for estimating the spectral ages along the lobes being marked by vertical lines and labelled. The central region corresponds to the ‘Central core’ region in Tables 3.3 and Fig. 3.6. The images have been rotated so that they lie− in the east-west direction. The spectral indices have been estimated between 332 and 5000 MHz. The grey scale bar indicates variations in spectral index from 0.4 to 2.0. Some∼ of the values at the edges are spurious. mJy respectively. The superscript g indicates that the flux densities have been estimated from a two-dimensional Gaussian fit to the core component. The spectra of the extended emission after subtracting the core flux density at frequencies larger than 1400 MHz are ∼ shown in Fig. 3.2 along with the fits to the data using the SYNAGE package (Murgia et al., 1999). The total flux densities are from Laing & Peacock (1980) and our measurements.

3.4 Discussion and results

3.4.1 Radiative losses

In the high-luminosity FRII radio sources, as the jets of relativistic plasma traverse outwards initially through the interstellar medium of the host galaxy and later through the intracluster and intergalactic medium, they dissipate their energy at their leading edges. This gives rise to the intense regions of emission called ‘hotspots’. The relativistic particles ﬂow out from the hotspots to form the extended lobes of radio emission, so that 52 3.Amultifrequencystudyofalargeradiogalaxy3C46

Table 3.2: The observational parameters and observed properties of the sources

Source Freq. Beamsize rms SI Cp Sp St Cp Sp St Cp Sp St MHz ′′ ′′ ◦ mJy mJy mJy mJy mJy mJy mJy mJy 1 1 1 1 beam− beam− beam− beam− (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) 3C46 G153 28.5 16.0 65 6.9 10880 W 2325 6029 C E 1852 4899 G240 13.2 8.1 48 4.3 6806 W 771 3821 C E 721 3099 G332 10.3 9.1 69 0.54 5957 W 725 3172 C E 734 2689 G606 5.2 3.8 48 0.23 2051 W 157 1537 C E 277 1402 V1465 13.8 12.7 12 0.61 1202 W 253 640 C E 197 590 V4841 13.9 11.6 166 0.15 355 W 76 185 Cg 1.2 1.1 E 60 172 V8460 7.7 6.9 32 0.04 183 W 25 88 Cg 1.6 1.8 E 30 94 the radiating particles closest to the hotspots are the youngest while those farthest from it are the oldest. The radio continuum spectra in different parts of an extended radio source contain information about the various energy losses and gains of the radiating particles during the lifetime of the source. If there is no significant reacceleration within these lobes and no significant mixing of particles, there should be a spectral gradient across the radio source. The hotspots where the particles are being accelerated should have the flat- test spectral index, reflecting the injection spectral index αinj, while the spectrum should steepen with increasing distance from the hotspot. Since the high-energy particles lose energy more rapidly, the steepening in the spectrum would be seen more clearly at high frequencies. This trend has been reported in several studies and used to estimate the radiative ages and expansion velocities in the powerful 3CR sources (e.g. Myers & Spangler

1985; Alexander & Leahy 1987; Leahy, Muxlow, & Stephens 1989; Carilli et al. 1991; Liu et al. 1992), in the low-luminosity and medium-luminosity radio galaxies (e.g. Klein et al. 1995; Parma et al. 1999), giant radio sources (Konar et al. 2008; Jamrozy et al. 2008 and references therein) and compact steep-spectrum sources (Murgia et al., 1999).

However, there are several caveats in the interpretation which one needs to bear in mind. These include details of the backflow of the lobe material, difficulties in disentangling the different energy losses of the radiating particles and variations of the local magnetic field (e.g. Wiita & Gopal-Krishna 1990; Rudnick, Katz-Stone, & Anderson 1994; Eilek

& Arendt 1996; Jones, Ryu, & Engel 1999; Blundell & Rawlings 2000.

Nevertheless, these provide useful inputs towards understanding the diﬀerent 3.4 Discussion and results 53

1000 1000

1000

100

100 10

3C46 strip W1 3C46 strip W3 3C46 strip W4

100 1000 10 100 1000 10 100 1000

1000

100

1 10 3c46 strip E3 3C46 strip E6 3C46 strip E7 1 100 1000 10 100 1000 10 100 1000

Figure 3.4: Typical spectra of the strips for the western (upper panel) and eastern (lower panel) lobes of 3C46, with the ﬁts from the JP model as described in the text. 54 3.Amultifrequencystudyofalargeradiogalaxy3C46 physical processes which play a role in the evolution of these radio sources. The large radio sources are particularly suitable for the classical spectral-ageing analysis due to their large angular extent which can be covered by a signiﬁcant number of resolution elements. Combining low-frequency information from the GMRT along with high-frequency ones from the VLA is likely to yield the most reliable estimates of the spectral break frequency, as has been demonstrated in a number of recent studies (Jamrozy et al. 2008 and references therein; Konar et al. 2008).

3.4.2 Spectral ageing analysis

The observed spectra have been fitted using the Jaffe & Perola (Jaffe & Perola 1973, JP) and the Kardashev-Pacholczyk (Kardashev 1962; Pacholczyk 1970; KP) models using the SYNAGE package (Murgia et al., 1999). As reported by Jamrozy et al. (2008), there is no significant difference between these two models over the frequency range of our observations, and the JP model tends to give a better fit to the different strips in the lobes than with the continuous injection (Kardashev 1962; CI) model. The CI model sometimes gives a somewhat better fit in the area of a prominent hotspot, but since with the resolution of our observations the flux density of the hotspots are contaminated by lobe emission it does not make a significant difference.

Assuming that (i) the magnetic ﬁeld strength in a given strip is constant throughout the energy-loss process, (ii) the particles injected into the lobe have a constant power- law energy spectrum with an index γ, and (iii) the time-scale of isotropization of the pitch angles of the particles is short compared with their radiative lifetime, the spectral age,

τspec, is given by

1/2 B 1/2 τ = 50.3 ν (1 + z) − [Myr], (3.1) spec 2 2 br B + BiC { } 2 where BiC=0.318(1+z) is the magnetic ﬁeld strength equivalent to the CMBR. Here B, 3.4 Discussion and results 55

the magnetic ﬁeld strength of the lobes, and BiC are expressed in units of nT, νbr is the spectral break frequency in GHz above which the radio spectrum steepens from the initial power-law spectrum given by α =(γ 1)/2. Alexander & Leahy (1987) and Alexander inj − (1987) have suggested that the eﬀects of expansion losses may be neglected.

To estimate the values of αinj for the source, we fit the JP model to the flux densities of the entire lobes, treating αinj as a free parameter as well as the total flux density measurements of the source which go to lower frequencies (Laing & Peacock 1980) than our measurements. These yield injection spectral indices of 1.00 (Fig. 3.2). Having estimated the αinj value, the total-intensity images have been convolved to a common resolution of 14 arcsec to be consistent with the lowest-resolution image for all images at

240 MHz and above. The 150-MHz image has not been used for this analysis since their ∼ resolution is coarser by another factor of two, and we wish to have at least 10 resolution ∼ elements along the axis of the source. Each lobe is then splitted into a number of strips as shown in Fig. 3.3, separated approximately by the resolution element along the axis of the source, and also ensuring that the core component lies between two vertical lines and can be subtracted reliably. The extreme strips are centred at the peaks of brightness on the convolved maps. 3C46 has prominent bridges of emission, and we have also fitted the spectrum to the central region of the source after subtracting the flux density of the radio core. Using the SYNAGE software we determine the best fit to the spectrum in each strip from 240 to 8000 MHz using the JP model, and derive the relevant value of ν . A few ∼ br examples of the fits to the different strips of both the lobes in 3C46 are presented in Figs.

3.4 , while the ﬁts to the central regions of the source, which as expected show the lowest values of νbr are presented in Fig. 3.5.

In order to estimate the spectral age, we have to estimate the magnetic-ﬁeld strength. We have estimated the magnetic ﬁeld strength by integrating the spectrum from a frequency corresponding to a minimum Lorentz factor, γ 10, for the relativis- min ∼ 56 3.Amultifrequencystudyofalargeradiogalaxy3C46

-1

3C46 Central-core

0 2 4

Figure 3.5: Spectra of the central regions of 3C46 subtraction of the core ﬂux density, with the ﬁts from the JP model as described in the text.

Table 3.3: Estimates of break frequency and spectral age for 3C46 2 Strip Dist. νbr χred B τspec (kpc) (GHz) nT (Myr) Western lobe αinj = 1.0 W1 316 > 100 5.42 1.64 < 1.7 W2 237 > 100 1.78 1.35 < 2.2 +6.5 +4.1 W3 158 36.4 28.3 1.71 1.35 3.6 0.4 −+0.7 +−1.7 W4 79 10.6 3.9 0.34 1.34 6.7 0.2 − −

Eastern lobe αinj = 1.0 E1 519 > 100 10.2 1.50 < 1.9 E2 440 > 100 11.4 1.36 < 2.2 E3 361 > 100 5.1 1.20 < 2.4 E4 282 > 100 1.0 1.12 < 2.6 >+6.6 +3.3 E5 203 36.3 25.8 0.9 1.26 3.9 0.3 >−+0.6 +−1.0 E6 124 9.7 2.2 0.2 1.30 7.3 0.2 −>+0.4 −+0.2 E7 56 7.0 0.3 4.1 1.22 9.1 0.3 − −

Central-core αinj = 1.0 +0.05 +2.6 2.5 0.7 44 1.28 14.5 0.1 − − 3.4 Discussion and results 57

3C46

10 age (Myr)

0 -600 -400 -200 0 200 400 distance (kpc) Figure 3.6: The spectral age is plotted against the distance from the core for 3C46 using the magnetic field estimated for each strip. tic electrons to an upper limit of 100 GHz, which corresponds to a Lorentz factor ranging from a few times 104 to 105 depending on the estimated magnetic field strength (see Hard- castle et al. 2004; Croston et al. 2005; Konar et al. 2008, 2009). It has also been assumed that the filling factors of the lobes are unity, and the energetically dominant particles are the radiating particles only, neglecting the contribution of the protons. We have assumed a cylindrical geometry for the entire lobe, and have estimated the magnetic field for the entire lobe as well as for individual strips of emission each separated by approximately a beamwidth along the long axis of each source. The magnetic field strengths for each strip are listed in Table 3.3. The magnetic field strengths for the western and eastern lobes of 3C46 are 1.66 and 1.53 nT respectively. The equipartition magnetic fields are usually within a factor of 2 of those estimated from inverse-Compton scattering of the radiating electrons by the microwave background radiation (e.g. Croston et al. 2005 Konar et al.

2009).

The results of our spectral ageing analysis are tabulated in Tables 3.3, which are 58 3.Amultifrequencystudyofalargeradiogalaxy3C46 arranged as follows. Column 1: identification of the strip, column 2: projected distance of the centre of the strip from the radio core in units of kpc, column 3: break frequency of the spectrum of the strip according to the JP model in units of GHz, column 4: reduced χ2 value of the fit, column 5: magnetic-field strength in units of nT, and column 6: spectral age of particles in the strip. The strips close to the hotspots are consistent with having straight spectra and the SYNAGE fits yield spectral breaks at frequencies larger than several hundred GHz. Our observations show no spectral break till 10 GHz. In Table 3.3 we ∼ have listed the values corresponding to a break frequency of 100 GHz for these strips. Observations at millimetre wavelengths are required to determine reliably the spectral breaks in these regions. The spectral age increases with distance from the hotspot, with the maximum spectral ages estimated for 3C46 being 15 Myr in the regions closest to ∼ the core (Fig. 3.6).

3.4.3 Search for episodic activity

The outer diffuse lobes from an earlier cycle of activity in sources with episodic activity are expected to have a steep spectra due to radiative and adiabatic losses. This has been observationally demonstrated in some cases such as J1453+3308 (Konar et al., 2006) and 4C29.30 (Jamrozy et al., 2008). Therefore, ideally one should be able to detect these features more easily at low frequencies. Our low-frequency images show the prominent bridges of emission but no diffuse features that could be attributed to an earlier cycle of activity. In the case of 3C46, a diffuse component of say 20 mJy at 153 MHz and ∼ a spectral index of 1 would have a surface brightness similar to the rms noise in the

8460-MHz image. The value would increase by a factor of 2 and 7 for spectral indices ∼ 1 of 1.2 and 1.5 respectively. With the rms noise value of 6.9 mJy beam− at 153 MHz, diﬀuse emission not seen at the highest frequency could have been just about detected in 3C46. While it is important to make more sensitive images at the lowest frequencies, the 3.5 Concluding remarks 59 non-detection of extended emission due to an earlier cycle of activity in this source, is consistent with the trend that such objects are rare even amongst large radio sources (cf. Schoenmakers et al. 2000c; Saikia et al. 2006).

3.5 Concluding remarks

The maximum spectral age determined for 3C46 is 15 Myr, which is similar to the val- ∼ ues of Jamrozy et al. 2008 obtained for a sample of 10 large radio galaxies by combining GMRT and VLA data. Their values range from 6 to 36 Myr with a median value of ∼ 20 Myr using the classic equipartition magnetic fields. These estimates are significantly ∼ older than those of smaller sources (e.g. Leahy et al., 1989; Liu et al., 1992), and broadly consistent with the tendency for spectral age to increase with the projected linear size (Jamrozy et al., 2008 and references therein). The injection spectral index is 1.0, compared with the values ranging from ∼ 0.55 to 0.88 with a median value of 0.6 for the sample of Jamrozy et al. (2008). Our ∼ estimate for 3C46, which have prominent hotspots, is consistent with the higher values in the sample of Jamrozy et al. (2008), and studies of smaller FRII sources studied by Leahy et al. (1989); Liu et al. (1992). Our estimate of the injection spectral index appear steeper than theoretically expected values for a strong, non-relativistic shock in a Newtonian fluid where αinj = 0.5 (Bell, 1978a,b; Blandford & Ostriker, 1978), or for different scenarios involving relativistic shocks where αinj vary in the range of 0.35 to 0.65 (Drury & Voelk, 1981; Axford, Leer, & McKenzie, 1982; Kirk & Schneider, 1987; Heavens, 1989;). High- resolution observations of these sources at even lower frequencies with future instruments should help in determining injection spectra more reliably. 60 3.Amultifrequencystudyofalargeradiogalaxy3C46 Chapter 4

A radio study of the double-double radio galaxy 3C293

We present radio continuum observations at frequencies ranging from 150 to ∼ 5000 MHz of the misaligned double-double radio galaxy, DDRG, 3C293 (J1352+3126) using the Giant Metrewave Radio Telescope (GMRT) and the Very Large Array (VLA). The spectra of the outer lobes and the central source are consistent with being straight, indicating spectral ages of <17 23 Myr for the outer lobes, and <0.1 Myr for the central ∼ − ∼ source. The north-western lobe has a prominent hotspot suggesting that the interruption of jet activity is <0.1 Myr, consistent with the age of the inner double. The time-scale ∼ of interruption of jet activity appears signiﬁcantly smaller than observed in most other

DDRGs which are often associated with giant radio sources. These observations suggest that there is a wide range of time-scales of interruption of jet activity in active galaxies.

Result presented in this Chapter is published in S. A. Joshi, S. Nandi et al. 2011, MNRAS , 141, 1397.

61 62 4.Aradiostudyofthedouble-doubleradiogalaxy3C293

4.1 Introduction

An important aspect in our understanding of active galactic nuclei (AGN) is whether their nuclear activity is episodic. If so, it is important to determine the duration of their active phases, and understand the implications of such episodic activity. The extended radio emission in radio galaxies and quasars contains an imprint of the history of nuclear jet activity. The structure and spectra of their lobes of radio emission provide us with an opportunity of studying the time scales of episodic nuclear activity. A striking example of episodic nuclear activity is when new pairs of radio lobes are seen closer to the nucleus before the ‘old’ and more distant radio lobes have faded (e.g. Subrahmanyan, Saripalli, & Hunstead 1996; Lara et al. 1999). The ones with two pairs of radio lobes are usually referred to as ‘double-double’ radio galaxies (DDRGs; Schoenmakers et al. 2000b); while an example of one with three pairs of radio lobes has also been reported (Brocksopp et al. 2007). Presently, close to about two dozen good cases of AGN with episodic nuclear activity have been identiﬁed from either radio and/or X-ray observations (Saikia & Jamrozy

2009, for a review).

Although in most DDRGs, the outer double appears reasonably well aligned with the inner one (e.g. Saikia, Konar, & Kulkarni 2006), 3C293 is a striking example of a DDRG where the inner double is misaligned from the outer one, the misalignment angle being 35◦. The radio galaxy 3C293 is at a redshift of 0.0450 (Fouque et al. 1992) so that ∼ 1 1 1 arcsec corresponds to 0.874 kpc in a Universe with H0=71kms− Mpc− , Ωm=0.27 and

Ωvac=0.73. The optical host galaxy, VV5-33-12, is peculiar with compact knots and multiple dust lanes ( van Breugel et al. 1984; Martel et al. 1999; de Koﬀ et al. 2000; Capetti et al. 2000 ), and appears to be a merger remnant. It has a small companion galaxy situated 37 arcsec ( 30 kpc) towards the south-west ( Heckman et al. 1985; Evans et al. ∼ ∼ 1999; Beswick et al. 2004 ). Hubble Space Telescope (HST) observations have detected an optical/infrared (IR) jet within the central kiloparsec emitting synchrotron emission at 4.1 Introduction 63 these wavlengths ( Leahy, Sparks, & Jackson 1999; Floyd et al. 2006 ). Emonts et al.

(2005) have reported evidence of jet-induced outﬂow of warm gas in 3C293. Broad neutral hydrogen (H) absorption was observed by Baan & Haschick (1981) using the Arecibo telescope, and has since been studied with a wide range of angular resolutions revealing the complex gas distribution seen in absorption towards the diﬀerent radio components of the central source ( Shostak et al. 1983; Haschick & Baan 1985; Beswick et al. 2002, 2004; Morganti et al. 2003; Beswick et al. 2004 ). Evans et al. (2005) have found CO(1-0) gas in both emission and absorption within the central few kiloparsecs. The CO emission appears to be largely distributed in an asymmetric disc rotating about the AGN.

The large and small scale structure of 3C293 have been imaged by a number of authors. The outer double-lobed structure has a projected linear size of 190 kpc, and re- ∼ sembles an FRII radio galaxy. However, the two lobes are highly asymmetric in intensity. The north-western component has a hotspot which is brighter than the peak of emission at the outer extremeties of the southern lobe by a factor of 10 (e.g. Bridle, Fomalont, ∼ & Cornwell 1981; Beswick et al. 2004 ). The prominent central source has a steep radio spectrum, and when observed with high angular resolution resembles a compact double- lobed source with multiple components and a flat-spectrum radio core ( Akujor et al. 1996; Beswick et al. 2004 and references therein; Giovannini et al. 2005 ). The projected linear separation of the two prominent peaks on opposite sides of the nucleus is 1.7 kpc, ∼ and has been interpreted to represent a more recent cycle of AGN activity, reminiscent of 3C236. However, 3C293 has a relatively small overall projected linear size compared with most DDRGs which are over 1 Mpc in size. Also, the inner double of 3C293 has ∼ diffuse extended emission beyond the two prominent peaks; the total extent of the inner source including the extended emission is 4.2 kpc. Akujor et al. (1996) find the dif- ∼ fuse extended emission to have significantly steeper spectra than the inner components and have considered the possibility that this might represent yet another cycle of activity. 64 4.Aradiostudyofthedouble-doubleradiogalaxy3C293

3C293 GMRT 240 MHZ 3C293 GMRT 154 MHZ 31 29 00 31 30

28 30 29 00

27 30 28

26 30 27

00 DECLINATION (J2000) DECLINATION (J2000) 26 25 30

00 25

24 30

24 00 13 52 30 25 20 15 10 05 RIGHT ASCENSION (J2000) 13 52 30 25 20 15 10 05 Cont peak flux = 1.5174E+01 JY/BEAM RIGHT ASCENSION (J2000) Levs = 4.700E-02 * (-1, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512) Cont peak flux = 1.0831E+01 JY/BEAM Levs = 3.700E-02 * (-1, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512)

3C293 GMRT 614 MHZ 3C293 VLA 4860 MHZ 31 29 00 31 29 00 28 30 28 30 00 00

27 30 27 30

00 00

26 30 26 30 DECLINATION (J2000) DECLINATION (J2000) 00 00

25 30 25 30

00 00

24 30 24 30

13 52 30 25 20 15 10 13 52 30 25 20 15 10 RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) Cont peak flux = 6.4224E+00 JY/BEAM Cont peak flux = 1.5558E+00 JY/BEAM Levs = 1.060E-02 * (-1, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512) Levs = 1.800E-03 * (-1, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512)

Figure 4.1: The GMRT images of 3C293 at 154, 240 and 614 MHz, and the VLA image at 4860 MHz. All these images have been made with an angular resolution of 23.2 21.6 2 × arcsec along a PA of 79◦, which is shown as an ellipse in the bottom left-hand corner. 4.2 Observations and data reduction 65

Similar resolution data over a larger frequency range would be useful to further explore this possibility.

In this chapter, we present the results of low-frequency observations made with the Giant Metrewave Radio Telescope (GMRT) at 154, 240 and 614 MHz, as well as Very Large Array (VLA) observations at 4860 MHz made from archival data (see Fig. 4.1). These observations do not resolve the diﬀerent components of the central source seen in the high-resolution images. We have determined the spectra of the outer and inner lobes over a large frequency range, estimated their spectral ages and discuss the constraints on time-scale of episodic activity. The observations and data reduction are described in Section 4.2. The observational results, including the radio maps and spectra are presented in Section 4.3. The results are presented in Section 4.4, while the concluding remarks are given in Section 4.5.

4.2 Observations and data reduction

The analysis presented in this chapter is based on radio observations made with the GMRT and VLA archival data as well. The observing log for both the GMRT and VLA observations is listed in Table 4.1.

4.2.1 GMRT observations

The GMRT observations were made in the standard manner, with each observations of the target-source interspersed with observations of the phase calibrator. 3C286 was observed as the ﬂux density and bandpass calibrator, and all ﬂux densities are on the Baars et al.

(1977) scale using the latest VLA values. At each frequency the source was observed in a full-synthesis run of approximately 9 hours including calibration overheads. The data were calibrated and reduced in the standard way using the NRAO AIPS package. Several rounds of self calibration were done to improve the quality of the images. 66 4.Aradiostudyofthedouble-doubleradiogalaxy3C293

Table 4.1: Observing log. Columns (1) and (2) show the name of the telescope, and the array conﬁguration for the VLA observations; columns (3) and (4) show the frequency and bandwidth used in making the images; column (5): the primary beamwidth in arcmin; column (6): dates of the observations.

Teles- Array Obs. Band- Primary Obs. Date cope Conf. Freq. width beam (MHz) (MHz) (arcmin) (1) (2) (3) (4) (5) (6) GMRT 154 6 186 2009Jan27 GMRT 240 6 114 2008Dec22 GMRT 614 16 43 2008Dec22 VLAa C 4860 50 9 1984Oct04 a: VLA archival data.

4.3 Observational results

The observational results on 3C293 are described here. A catalogue of sources detected at 154 MHz within 2.2◦ of the phase centre, i.e. the position of 3C293, was made

(HPBW 3.1◦). Those with angular sizes larger than 45 arcsec were compared with the ∼ NRAO VLA Sky Survey (NVSS; Condon et al. 1998) images to examine evidence of any steep-spectrum relic emission not detected in the NVSS images at 1400 MHz. No such relic emission was seen in the 154-MHz images, consistent with earlier studies (e.g. Sirothia et al. 2009b), while a possible relic and a couple of extended sources, all seen in the NVSS images, are described brieﬂy in the Appendix A. The catalogue of sources

2 detected at 154 MHz with an angular resolution of 15 12.5 arcsec along PA 73◦ is pre- × sented in the Appendix A. A total of 320 sources, including 3C293, which have a peak ﬂux density 7-σ have been listed. The value of σ, the primary beam corrected local rms ≥ 1 1 noise, varies from 8.6 mJy beam− near 3C293 to typical values of 3 4 mJy beam− in re- − gions without a strong source. All but three of the weaker sources, namely, J1352+3059, J1352+3039 and J1357+3111, are seen in the NVSS images at 1400 MHz. Using 5 times the rms noise in the NVSS images yields spectral indices steeper than 1.2 for these three ∼ 4.3 Observational results 67 sources.

4.3.1 Overall structure of 3C293

The observational parameters and the flux densities estimated from these images are presented in Table 4.2. These have been estimated from the images made by tapering and weighting the data and convolving the images to match the resolution of the GMRT image at 154 MHz (see Fig. 4.1). The flux densitites have been estimated by specifying a similar area around the components at the different frequencies. We have examined the variation in flux density while varying the size of the area around the component. Considering these fluctuations as well as calibration errors, the error in the flux density has been estimated to be approximately 15 per cent at 150 and 240 MHz, 7 per cent at 610 MHz and 5 per cent at 4860 MHz.

The source is quite asymmetric in both brightness and location of the outer components. The ratio of separations from the core of the brighter north-western component to the weaker south-eastern one is 0.6. The peak brightness of the north-western com- ∼ ponent is larger by a factor of >10, while the ratio of the total flux densities lies between ∼ a factor of 5 and 7, depending on the frequency of observations (see Table 4.2). The ∼ central source is also highly asymmetric in location, with the ratio of separations from the core of the eastern component to the western one being 3. However, the flux densities ∼ are more symmetric with the peak brightness ratio being within a factor of 2, and the ∼ ratio of total flux densities varying between 1.2 and 1.6 (cf. Akujor et al. 1996; Beswick ∼ et al. 2004 ). Such asymmetries are often seen in compact steep-spectrum radio sources (cf. Saikia et al. 1995), and the asymmetries observed in 3C293 are likely to be due to a combination of the effects of both orientation and relativistic motion (cf. Jeyakumar et al. 68 4.Aradiostudyofthedouble-doubleradiogalaxy3C293

Table 4.2: The observational parameters and ﬂux densities. Column (1): frequency of observations in MHz, with the letter G or V representing either GMRT or VLA observations; columns (2) (4): the major and minor axes of the restoring beam in arcsec and its − 1 PA in degrees; column (5): the rms noise in units of mJy beam− ; column (6): component designation, where Cent refers to the central source including the radio core; columns (7) and (8): the peak and total ﬂux densities of the source.

Freq. Beam size rms Cmp. Sp St (MHz) (′′) (′′) (◦) (mJy) (mJy) (mJy) /b) /b) (1) (2) (3) (4) (5) (6) (7) (8) G154 23.2 21.6 79 8.6 NW 2198 5491 Cent 15216 15570 SE 215 1022 Tot 15216 22288 G240 23.2 21.6 79 11.7 NW 1401 3244 Cent 10825 11041 SE 127 577 Tot 10825 15142 G614 23.2 21.6 79 3.0 NW 754 1509 Cent 6421 6446 SE 61 300 Tot 6421 8282 V4860 23.2 21.6 79 0.67 NW 178 305 Cent 1556 1553 SE 10 42 Tot 1556 1909

2005 ).

4.3.2 Spectra

The integrated spectrum of 3C293 using the data from Laing & Peacock (1980) and Kuehr et al. (1981), as well as from our measurements is shown in Fig. 4.2 All the ﬂux densities are consistent with the scale of Baars et al. (1977). Excluding the points below 50 MHz which have large uncertainties and the measurement at 178 MHz which is signiﬁcantly lower than expected from the measurements at 154 and 160 MHz, the spectral index, α

α (S ν− ), of the source is 0.71 0.01. Using our measurements alone yields the same ∝ ± 4.3 Observational results 69

1e+06 100000

integrated spectrum north-western component

100000

10000

10000 Flux density (mJy) Flux density (mJy) 1000

1000

100 100 10 100 1000 10000 10 100 1000 10000 Frequency (MHz) Frequency (MHz) 100000 10000

central component south-eastern component

10000 1000 Flux density (mJy) Flux density (mJy) 1000 100

100 10 10 100 1000 10000 10 100 1000 10000 Frequency (MHz) Frequency (MHz) Figure 4.2: The integrated spectrum of 3C293 using the measurements from Laing & Peacock (1980), Kuehr et al. 1981 and our measurements. These are shown as filled squares, and filled circles and open circles respectively (upper left). The spectra of the north-western (upper right), central (lower left) and south-eastern (lower right) components respectively. For the central component, the core flux density has been subtracted for measurements above 1400 MHz, as described in the text. ∼ 70 4.Aradiostudyofthedouble-doubleradiogalaxy3C293 value for the spectral index.

In Fig. 4.2 we also present the spectra of the north-western, central and south- eastern components using our measurements. For the central component, we have also used the measurements made by Bridle et al. (1981) after subtracting the flux density of the flat-spectrum core component at frequencies of 1400 MHz and above using the ∼ values listed by Bridle et al. (1981), Akujor et al. (1996) and Beswick et al. (2004). The core is relatively weak, even at 15000 MHz, and has made no significant difference ∼ to the spectra. The flux densities of the extended lobes have not been listed by Bridle et al. (1981). The flux densities of all the three components are consistent with straight spectra, the spectral indices being 0.72 0.02 for the central component, and 0.80 0.02 ± ± and 0.91 0.03 for the north-western and south-eastern components respectively. The ± spectral index of the central component, which contributes over 70 per cent of the total flux density of the source, is consistent with the integrated spectral index.

4.4 Discussion

4.4.1 Spectral ages

The spectra of the outer components using the total ﬂux densities between 150 and 4860

MHz and that of the inner double using measurements between 150 and 15000 MHz are consistent with a single power-law. We have fitted the spectrum of the inner double after subtracting the core flux density for the Jaffe & Perola (1973,JP), Kardashev-Pacholczyk (KP, Kardashev 1962; Pacholczyk 1970 ) and the continuous injection (CI, Pacholczyk

1970 ) models using the SYNAGE package (Murgia et al., 1999). The break frequency obtained from these ﬁts are rather large (>7 105 GHz) and have huge uncertainties because ∼ × the spectra are practically straight. We have adopted conservative lower limits to the break frequency to be the highest frequency of our observations. 4.4 Discussion 71

In order to estimate the spectral ages, we have to estimate the magnetic field strength, which was done in a similar way to that of 3C46 and 3C452 (Nandi et al., 2010). The magnetic field strength has been estimated by integrating the spectrum from a frequency corresponding to a minimum Lorentz factor, γ 10 for the relativistic electrons min ∼ to an upper limit of 100 GHz, which corresponds to a Lorentz factor ranging from a few times 104 to 105 depending on the estimated magnetic field strength (see Hardcastle et al. 2004; Croston et al. 2005; Konar et al. 2008, 2009). It has also been assumed that the filling factors of the lobes are unity, and the energetically dominant particles are the radiating particles only, neglecting the contribution of the protons. A cylindrical geometry has been assumed for both the lobes, and the central source, and their sizes have been estimated more reliably from the availabe higher-resolution images (e.g. Bridle et al. 1981; Akujor et al. 1996). The deconvolved size of the central source estimated from our observations is consistent with that estimated from the higher-resloution image of Akujor et al. (1996).

The equipartition magnetic field estimate for the inner double is 16.92 1.67 nT, ± indicating that for a conservative break frequency >16 GHz, the inferred spectral age is ∼ <0.18 Myr. For the extended lobes, where reliable measurements of the total flux density ∼ are available up to 5 GHz, the magnetic field strengths are 1.12 0.11 and 0.88 0.09 ∼ ± ± nT for the north-western and south-eastern lobes respectively, while the corresponding spectral ages are <16.9 and 23.0 Myr respectively for a break frequency >5 GHz. For a ∼ ∼ break frequency of >100 GHz, as was adopted by Nandi et al. (2010), the spectral ages ∼ are <0.07, 3.77 and 5.13 Myr respectively for the central source, and the north-western ∼ and south-eastern components respectively. Machalski et al. (2007) have examined the dynamical ages of FRII radio sources and find that these agree with the spectral ages for objects less than 10 Myr, while for a sample of giant radio galaxies the ratio of dynamical age to the spectral age of the lobes lies between 1 and 5 (Machalski, Jamrozy, & Saikia ∼ 2009). Lower limits to the dynamical age may be estimated by assuming the velocity of 72 4.Aradiostudyofthedouble-doubleradiogalaxy3C293

advancement to be c, and an inclination of the source axis to the line of sight to be 45◦. ∼ This yields ages of >0.3 Myr for the outer north-western hot-spot, and >0.006 Myr for ∼ ∼ the more distant eastern hot-spot of the central source in 3C293.

4.4.2 Recurrent activity time scale

In addition to the age estimates, one can also constrain the time scale of interruption of jet activity from the presence of a hot-spot in the north-western lobe of 3C293. Hot-spots in the outer lobes have been seen in other sources with evidence of episodic jet activity, although the outer doubles are often diﬀuse as for example in J1453+3308 (Schoenmakers et al. 2000b; Konar et al. 2006). Examples of sources with hot-spots are the northern lobe of B1834+620 (Schoenmakers et al. 2000a), and the western outer lobe of 4C02.27

(Jamrozy, Saikia, & Konar 2009). The hot-spots have typical sizes of <10 kpc (e.g. ∼ Jeyakumar & Saikia 2000), and are expected to fade in 104 to 105 yr after the energy ∼ supply has been cut oﬀ (e.g. Kaiser, Schoenmakers, & Röttgering 2000). This is usually smaller than the time it takes for the material to reach the hot-spots from the radio core, and hence it is reasonable to assume that the hot-spot fades soon after the last jet material has passed through it.

For 3C293, the fraction of emission from the core, fc, at an emitted frequency of 8 GHz, which is often used as a statistical indicator of the orientation of the jet axis to the line of sight, is 0.017. The corresponding median value of f for galaxies is ∼ c 0.002 while for quasars it is 0.05 (e.g. Saikia & Kulkarni 1994), suggesting that the ∼ ∼ orientation of the nuclear jet is close to the dividing line between radio galaxies and quasars. We adopt an orientation angle of 45◦ to the line of sight. The hot-spot in ∼ the north-western lobe implies that it still receives jet material. For an inclination angle,

φ 45◦ and a jet velocity of c, the travel time from the core to the hot-spot is 0.35 Myr, ∼ ∼ However, this will be aﬀected by light-travel time eﬀects due to the orientation of the 4.5 Concluding remarks 73 source axis. The high brightness asymmetry and the observed levels of polarization (cf.

Bridle et al. 1981) suggest that the north-western lobe is approaching us, indicating that the observed time diﬀerence between the ejection of the last material and its arrival at the approaching hot-spot is 0.1 Myr. To be able to see the hot-spot as well as the inner ∼ structure, the interruption of jet activity must be less than 0.1 Myr, within which period ∼ of time the inner double must also form. The estimated spectral age of the inner double assuming a break frequency of 100 GHz is consistent with this scenario. ∼

4.5 Concluding remarks

Estimation of the time scales of recurrent activity is important to constrain models of such activity (e.g. Kaiser, Schoenmakers, & Röttgering 2000; Brocksopp et al. 2011), as well as to understand possible effects on the host galaxy and its evolution. GMRT and VLA observations of this highly misaligned DDRG have shown that the spectral age of the inner double is likely to be <0.1 Myr, which is similar to the time scale of interruption ∼ of jet activity. The overall linear size of 190 kpc, is much smaller than most DDRGs ∼ which are known to be over a Mpc in size. For example in the well-studied DDRG J1453+3308, the dynamical and spectral ages of the diffuse outer lobes are 215 and 50 ∼ Myr, while that of the inner double is only 2 Myr, suggesting a much longer time-scale ∼ of interruption (Kaiser et al. 2000; Konar et al. 2006). Reliable identification of signs of episodic activity in sources of different sizes is required to explore the entire range of time scales of episodic activity. For example in the compact steep-spectrum source CTA 21, a compact double-lobed structure with a size of 12 mas (Jones 1984; Kellermann et al. ∼ 1998), and diffuse emission on scales extending to 300 mas (Dallacasa et al. 1995), ∼ led Salter et al. (2010) to suggest that CTA 21 may be undergoing repeated cycles of activity. This is consistent with the suggestion of (Reynolds & Begelman 1997) that jet activity in compact sources may be intermittent on time scales of 104 to 105 yr. Objects 74 4.Aradiostudyofthedouble-doubleradiogalaxy3C293 such as 3C293 help us explore the implications of the intermediate range of time scales of episodic nuclear activity. For example the model by Kaiser et al. (2000) where long time-scales of 107 yr is required for the dispersion of the clouds of entrained material ∼ would not be applicable here. Here, given the small size of the inner double, the hotspots are likely to have formed by interaction of the jets with the interstellar medium of the host galaxy. Chapter 5

Double-double radio galaxies from the FIRST survey

The radio structures and optical identifications of a sample of 242 sources classified as double-double radio sources by Proctor (2011) from a morphological study of sources in the FIRST (Faint Images of the Radio Sky at Twenty centimeters) survey (2003 April release, 811,117 entries) have been examined. We have been able to confirm only

23 of these as likely to be double-double radio galaxies (DDRGs), whose structures could be attributed to episodic nuclear activity in their host galaxies. A further 63 require either higher-resolution radio observations or optical identiﬁcations to determine whether these are DDRGs. The remaining sources are unlikely to be DDRGs. We have examined the luminosities, sizes and symmetry parameters of the DDRGs and the constraints these place on our understanding of these sources.

Result presented in this Chapter is published in Nandi, S.; Saikia, D. J. 2012, BASI, 40, 121.

75 76 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

5.1 Introduction

An important and interesting question in our understanding of active galactic nuclei

(AGN) is whether their nuclear activity is usually episodic in nature. There have been suggestions that black holes grow during AGN phases with the total life time of the active phases ranging from 1.5 108 to 109 yr (e.g. Marconi et al. 2004). Recurrent activ- ∼ × ity could also have signiﬁcant implications in feedback processes in active galaxies and on the evolution of galaxies themselves. Although there has been increasing evidence of recurrent activity in AGN from both radio and X-ray observations (see Saikia & Jamrozy 2009), it is not clear how ubiquitous this phenomenon might be. For example from deep low-frequency observations Sirothia et al. (2009b) did not ﬁnd any unambiguous evidence of recurrent activity in a sample of 374 small- sized sources. Marecki (2012) has interpreted the structure of the highly asymmetric giant radio galaxy J1211+743 (Pirya et al. 2011) to be due to recurrent nuclear activity. To understand these varied aspects as well as the range of time scales of episodic activity and possible reasons for it (see Kaiser,

Schoenmakers, & Röttgering 2000; Czerny et al. 2009; Brocksopp et al. 2011), it is necessary to signiﬁcantly increase the number of sources with evidence of recurrent activity beyond the couple of dozen or so that are presently known (see Saikia & Jamrozy 2009). At present, there appears to be a reasonably wide range of time scales of episodic activity which has been estimated from spectral and dynamical ages of the outer and inner lobes of a few DDRGs. These range from 105 yr for 3C293 (Joshi et al. 2011) to 108 yr for ∼ ∼ the largest Mpc-scale DDRGs (e.g. Schoenmakers et al. 2000b; Konar et al. 2006).

Approximately 10 per cent of AGNs are luminous at radio wavelengths, referred to as radio-loud objects. Drawing an analogy with X-ray binary systems in our Galaxy, a number of authors have suggested that radio activityin AGN could itself be episodic in nature (Nipoti, Blundell, & Binney 2005; Körding, Jester, & Fender 2006). One of the clear- est signs of episodic activity is seen in radio-loud objects where there are more than one 5.1 Introduction 77 pair of distinct outer lobes, which can be unambiguously ascribed to diﬀerent cycles of activity. Although most of these sources exhibit two cycles of activity, and are referred to as DDRGs, there are two examples of sources which appear to possibly exhibit three cycles of activity, namely B0925+420 (Brocksopp et al. 2007) and J140948.85 030232.5 (Hota − et al. 2011), the latter being associated with a spiral host galaxy. Almost all the sources exhibiting evidence of a double-double structure are associated with galaxies, with the possible exception of 4C02.27 which is associated with a quasar (Jamrozy, Saikia, & Konar 2009). Evidence of episodic activity has also been reported from a combination of X-ray and radio observations, where inverse-Compton scattered X-ray emission from old electrons in the relic lobes are seen along with synchrotron emission from the recent cycle of activity. Sources where this has been reported include 3C191 and 3C294 (Erlund et al. 2006), as well as the well-studied object Cygnus A (Steenbrugge, Blundell, & Duﬀy 2008; Steenbrugge, Heywood, & Blundell 2010).

To understand the nature of these sources and possible reasons for their episodic activity, we need to enlarge the sample of objects. As a first step we have focused on the FIRST survey (Becker, White, & Helfand 1995), where Proctor (2011) has done a classification of the structures of sources into different categories. Of interest here are the 242 sources Proctor (2011) has classified as DDRGs based on the identification of at least four different components in the radio images. Since the mere existence of four components does not guarantee the source to be a DDRG, we have examined the radio structure as well as the optical identifications of all the 242 sources to identify those we believe to be good examples of DDRGs, possible examples which require further observations and also sources which do not appear to be DDRGs. We describe briefly the methodology we have adopted in Section 5.2, describe the DDRGs identified from this survey in Section 5.3, and discuss the nature of these sources in Section 5.4. 78 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

5.2 Methodology of classiﬁcation of the DDRGs

As mentioned earlier, we have examined the radio structures and optical identiﬁcations of each of the 242 sources listed by Proctor (2011) as DDRGs. To identify the optical objects we have used the Sloan Digital Sky Survey (SDSS) Data Release 8 (DR8)1. If the source is not covered in DR8 we have examined the Digital Sky Survey DSS R-band images to

find the optical identifications. We consider the optical object to be identified with the radio core component if the optical position is within an arcsec of the radio peak position of an unresolved or slightly resolved compact component. Normally the difference is within 0.5 arcsec. However, in some cases the presence of extended emission near the ∼ core at 1400 MHz may shift the centroid and higher frequency observations are required to locate the core more accurately. In sources without a radio core we consider the optical object to be identified with the radio source if it lies close to the axis defined by the inner lobes. The extended lobe-like emission in the source J1158+2625 is not visible in the

NRAO VLA Sky Survey (NVSS) image and is likely to be spurious. This source has not been considered further.

Sources where radio emission from two cycles of activity are clearly distinguish- able with the inner structures being more compact, and with an optical identification located between the inner doubles have been classified as DDRGs. Such sources without an optical identification have been classified as candidate DDRGs, as illustrated in the Appendix B. In these cases deep optical imaging is required to examine whether one of the inner compact components may be co-incident with an optical identification. Deep multi-frequency radio observations would also help determine whether the radio spectrum is flat, as is usually the case for radio cores. Alternatively it could help detect possible backflow in the form of tails if these compact inner components are hot-spots.

A source where the optical object is co-incident with one of the inner compact

1The SDSS Web Site is http://www.sdss.org/ 5.2 MethodologyofclassiﬁcationoftheDDRGs 79 components is clearly not a DDRG, and is listed as a non-DDRG in this study. For example the sources J0759+4051 and J1434+0441 discussed in the Appendix B belong to this category.

A DDRG is expected to have diffuse outer lobes due to the an earlier cycle of activity and more compact hot-spots in the inner lobes which are due to a more recent cycle of AGN activity. With inadequate resolution the components of radio emission on either side of the host galaxy may not be resolved clearly to reveal the detailed structure. These components may be due to either two cycles of activity or the secondary inner peaks may be caused by backflow from the outer hot-spot. We have fitted two dimensional gaussians to all such sources. Those where the inner components are clearly resolved and more extended than the outer ones are likely to be due to peaks of emission caused by back-flowing plasma and have been classified as non-DDRGs; those where the inner components are more compact but higher resolution observations are required to determine their structure more reliably are referred to as candidate DDRGs in our study. J0032 0019, shown in the Appendix B, represents the kind of sources where the inner − components are due to backflowing plasma from the hot-spots.

The wide-angle tailed sources (WATs) often have more compact emission closer to the host galaxy with more diﬀuse tails of emission extending from these features (e.g. Blanton et al. 2003; Mao et al. 2010). Detailed spectral studies are required to establish the episodic nature of these sources. These WATs are listed separately in Appendix B and will not be discussed here.

This leaves us with a sample of 23 DDRGs and 63 candidate DDRGs, where further observations are required to establish whether these might be DDRGs. The list of DDRGs are presented in Table 5.1, some of its properties in Table 5.2, while the list of candidate DDRGs are presented in Appendix B (Table B.1). The list of non-DDRGs and the WAT sources are presented in the Appendix B (Tables B.2 and B.3 respectively). 80 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

Table 5.1: DDRGs from the FIRST survey

Source Opt. z RAoptical Decoptical RAcore Deccore Notes name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J0746+4526 G (0.517) 07:46:17.92 +45:26:34.47 J0804+5809 08:04:42.79 +58:09:34.94 J0855+4204 G (0.279) 08:55:49.15 +42:04:20.12 J0910+0345 G (0.588) 09:10:59.10 +03:45:31.68 1 J1039+0536 G 0.0908 10:39:28.21 +05:36:13.61 J1103+0636 G (0.449) 11:03:13.29 +06:36:16.00 J1158+2621 G 0.1120 11:58:20.13 +26:21:12.08 11:58:20.12 +26:21:12.04 2 J1208+0821 G (0.600) 12:08:56.78 +08:21:38.57 12:08:56.78 +08:21:38.44 J1238+1602 S 12:38:21.20 +16:02:41.43 J1240+2122 G (0.357) 12:40:13.48 +21:22:33.04 J1326+1924 G 0.1762 13:26:13.67 +19:24:23.75 J1328+2752 G 0.0911 13:28:48.45 +27:52:27.81 13:28:48.43 +27:52:27.55 3 J1344 0030 G (0.579) 13:44:46.92 00:30:09.28 − − J1407+5132 G (0.324) 14:07:18.49 +51:32:04.88 J1500+1542 G (0.456) 15:00:55.18 +15:42:40.64 J1521+5214 G (0.537) 15:21:05.90 +52:14:40.15 J1538 0242 G (0.575) 15:38:41.31 02:42:05.52 − − J1545+5047 G 0.4309 15:45:17.21 +50:47:54.18 J1605+0711 G (0.268) 16:05:13.74 +07:11:52.56 J1627+2906 G (0.722) 16:27:54.63 +29:06:20.00 J1649+4133 S 16:49:28.32 +41:33:41.58 J1705+3940 G (0.701) 17:05:17.83 +39:40:29.25 J1706+4340 S 17:06:25.43 +43:40:40.41 1: separation of optical position from nearest radio peak, ∆radio opt is 2.5 arcsec; 2: known DDRG (Owen − ∼ & Ledlow 1997; Saikia & Jamrozy 2009); 3: misaligned DDRG. In all the Tables the photometric redshifts, taken from SDSS, are enclosed within parentheses.

⊕ In the second Column G and S represent galaxy and star respectively, as classiﬁed in SDSS. Spectroscopic observations are required to determine whether the stellar objects are quasars. 5.2 MethodologyofclassiﬁcationoftheDDRGs 81

J0746+4526 FIRST 45 27 45

30 J0804+5809 FIRST

58 10 15 15

00 00

26 45 09 45

30 30 DECLINATION (J2000) DECLINATION (J2000)

15 15

00 00

08 04 52 50 48 46 44 42 40 38 36 34 25 45 RIGHT ASCENSION (J2000) Cont peak flux = 2.1649E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

30 07 46 22 20 18 16 14 12 RIGHT ASCENSION (J2000) Cont peak flux = 1.1176E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J0855+4204 FIRST

42 05 30 J0910+0345 FIRST

15 03 46 00

04 45 45 45

15 30

DECLINATION (J2000) 00 DECLINATION (J2000)

03 45 15

15 00 09 11 01.0 00.5 00.0 10 59.5 59.0 58.5 58.0 57.5 57.0 00 RIGHT ASCENSION (J2000) Cont peak flux = 2.4355E-02 JY/BEAM 08 55 53 52 51 50 49 48 47 46 45 44 Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64) RIGHT ASCENSION (J2000) Cont peak flux = 7.3320E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J1158+2621 FIRST

26 23 00

J1103+0636 FIRST

22 30

06 36 45

21 30

00 DECLINATION (J2000)

DECLINATION (J2000) 00

20 30 35 45

00 11 03 18 16 14 12 10 08 RIGHT ASCENSION (J2000) Cont peak flux = 5.5655E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64) 19 30

11 58 28 26 24 22 20 18 16 14 RIGHT ASCENSION (J2000) Cont peak flux = 2.5573E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

Figure 5.1: FIRST images of the DDRGs listed in Table 6.1. In all the Figures the + sign when shown denotes the position of the optical object. 82 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

J1238+1602 FIRST

16 03 45

J1208+0821 FIRST 30

15 08 22 00

21 45 02 45

30 DECLINATION (J2000) DECLINATION (J2000) 30

15 12 09 01 00 08 59 58 57 56 55 54 53 00 RIGHT ASCENSION (J2000) Cont peak flux = 3.4427E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64) 01 45

12 38 25 24 23 22 21 20 19 18 17 RIGHT ASCENSION (J2000) Cont peak flux = 1.0612E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J1240+2122 FIRST

21 23 30

J1326+1924 FIRST 15 19 24 45

40 00 35

22 45 30

25 30 20

15 15 DECLINATION (J2000) DECLINATION (J2000)

10 00 05

00 21 45 13 26 16 15 14 13 12 11 RIGHT ASCENSION (J2000) Cont peak flux = 4.3811E-03 JY/BEAM 30 Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

12 40 17 16 15 14 13 12 11 10 RIGHT ASCENSION (J2000) Cont peak flux = 3.5430E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J1328+2752 FIRST J1344-0030 FIRST

-00 29 00

27 54 30

53 30

30 00

52 30 DECLINATION (J2000) DECLINATION (J2000) 15

00 30

51 30 45

00 31 00

13 28 54 52 50 48 46 44 42 13 44 50 49 48 47 46 45 44 RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) Cont peak flux = 1.0769E-02 JY/BEAM Cont peak flux = 8.6657E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64) Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

Figure 5.1: (continued) 5.2 MethodologyofclassiﬁcationoftheDDRGs 83

J1500+1542 FIRST

15 43 45

J1407+5132 FIRST 30 51 33 00

32 45 15

30 00

42 45 00 DECLINATION (J2000) 31 45 DECLINATION (J2000) 30

30 15

00 00 14 07 28 26 24 22 20 18 16 14 12 10 RIGHT ASCENSION (J2000) Cont peak flux = 1.4652E-02 JY/BEAM 41 45 Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

15 00 58 57 56 55 54 53 52 RIGHT ASCENSION (J2000) Cont peak flux = 5.7980E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J1521+5214 FIRST J1538-0242 FIRST

-02 41 00

52 15 15 15

00 45

42 00 14 45

15 DECLINATION (J2000) DECLINATION (J2000)

30 30

43 00

00 15 15 21 09 08 07 06 05 04 03 02 RIGHT ASCENSION (J2000) 15 38 44 43 42 41 40 39 38 RIGHT ASCENSION (J2000) Cont peak flux = 3.7233E-03 JY/BEAM Cont peak flux = 8.0249E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64) Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J1605+0711 FIRST

07 13 30

00 J1545+5047 FIRST

12 30 50 48 15

00 00

47 45 11 30 DECLINATION (J2000) DECLINATION (J2000)

30 00

15 45 22 20 18 16 14 12 RIGHT ASCENSION (J2000) Cont peak flux = 7.8260E-03 JY/BEAM 10 30 Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

16 05 17 16 15 14 13 12 11 10 RIGHT ASCENSION (J2000) Cont peak flux = 1.5743E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

Figure 5.1: (continued) 84 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

J1627+2906 FIRST

29 07 15 J1649+4133 FIRST

41 34 05 00 00

06 45 33 55

50 30 45

15 40

35 DECLINATION (J2000) DECLINATION (J2000) 00 30

25 05 45

30 15 16 49 31 30 29 28 27 26 RIGHT ASCENSION (J2000) Cont peak flux = 5.4144E-03 JY/BEAM 15 Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64) 16 27 58 57 56 55 54 53 52 51 RIGHT ASCENSION (J2000) Cont peak flux = 1.4034E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J1705+3940 FIRST

39 41 30 J1649+4133 FIRST

41 34 05 15 00

33 55 00

50 40 45

30 40

35 DECLINATION (J2000) DECLINATION (J2000) 15

00 25

20 39 45

15 16 49 31 30 29 28 27 26 RIGHT ASCENSION (J2000) 30 Cont peak flux = 5.4144E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64) 17 05 22 21 20 19 18 17 16 15 14 RIGHT ASCENSION (J2000) Cont peak flux = 1.7788E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

J1706+4340 FIRST

43 41 30

40 45

30 DECLINATION (J2000)

39 45 17 06 32 30 28 26 24 22 20 RIGHT ASCENSION (J2000) Cont peak flux = 5.7409E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 32, 64)

Figure 5.1: (continued) 5.2 MethodologyofclassiﬁcationoftheDDRGs 85

2.5

1.5

Log luminosity ratio: outer/inner 0.5

-0.5 0 100 200 300 400 500 600 700 800 Linear size: inner (kpc) Figure 5.2: The ratio of the luminosities of the outer to the inner doubles vs the projected linear size of the inner double. The + signs represent the DDRGs identiﬁed from the FIRST survey, while the open circles represent those from Saikia et al. (2006).

These tables are arranged as follows. Column 1: source name; Column 2: optical iden- tiﬁcation; Column 3: redshift; Columns 4 and 5: The right ascension (hh:mm:ss.ss) and declination (dd:mm:ss.ss) of the optical objects in J2000 co-ordinates; Columns 6 and 7: The right ascension (hh:mm:ss.ss) and declination (dd:mm:ss.ss) of the radio core positions; Column 8: Notes on individual sources. Since we are not discussing the WAT sources in detail, notes are not provided for Table B.3.

Some of the observed properties of the FIRST DDRGs are listed in Table 5.2 which is arranged as follows. Column 1: source name; Column 2: optical identiﬁcation;

Column 3: redshift; Columns 4 and 5: projected linear size of the inner and outer double- lobed sources in kpc; Column 6: the locations of the components farther/closer from the core for the inner double. The symmetry parameters mentioned in Columns 7 10 are − all in the same sense as given in Column 6. Columns 7 and 8 represent respectively the armlength or separation ratio for the inner and outer doubles; Columns 9 and 10: ﬂux density ratios for the inner and outer doubles; and Columns 11 and 12: log of radio luminosity at an emitted frequency of 1400 MHz for the inner and outer doubles. The FIRST images of all the DDRGs are shown in Fig. 5.1. 86 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

Table 5.2: Some of the observed properties of the sample of DDRGs Source Opt. Red- lin lo Cmp. Rθ(in) Rθ(o) Rs(in) Rs(o) Pin Po Id. shift kpc kpc W/Hz W/Hz (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) J0746+4526 G (0.517) 95 630 N/S 1.86 1.35 0.79 0.19 25.41 26.17 J0804+5809 W/E 2.00 0.99 1.43 0.81 J0855+4204 G (0.279) 40 516 N/S 1.17 0.86 1.34 0.87 24.64 25.30 J0910+0345 G (0.588) 34 225 E/W 2.17 0.73 0.84 0.99 25.85 25.66 J1039+0536 G 0.0908 29 163 W/E 1.25 0.74 1.03 1.29 24.05 24.97 J1103+0636 G (0.449) 71 652 E/W 1.28 0.93 1.51 0.72 24.99 25.56 J1158+2621 G 0.1120 143 503 N/S 1.70 1.01 1.81 1.31 24.33 24.80 J1208+0821 G (0.600) 121 687 W/E 1.07 1.04 0.93 0.96 24.49 25.76 J1238+1602 S S/N 1.23 0.91 1.12 0.79 J1240+2122 G (0.357) 68 584 S/N 2.00 1.07 4.40 0.72 25.10 25.42 J1326+1924 G 0.1762 27 149 E/W 1.20 1.25 1.06 0.65 23.68 24.56 J1328+2752 G 0.0911 95 355 N/S 1.53 1.61 1.85 0.63 23.76 24.39 J1344 0030 G (0.579) 86 643 N/S 1.25 1.06 1.30 0.35 25.46 25.60 − J1407+5132 G (0.324) 87 613 W/E 1.40 0.84 1.51 0.62 24.43 26.15 J1500+1542 G (0.456) 128 485 S/N 1.54 0.76 1.68 1.27 24.93 24.92 J1521+5214 G (0.537) 64 391 S/N 1.50 0.74 2.09 0.81 24.99 25.28 J1538 0242 G (0.575) 59 534 N/S 1.16 0.93 0.73 1.24 25.05 26.30 − J1545+5047 G 0.4309 66 483 E/W 1.12 0.84 0.67 1.23 24.95 25.69 J1605+0711 G (0.268) 311 576 S/N 1.30 1.06 0.54 0.69 25.05 25.38 J1627+2906 G (0.722) 74 702 N/S 2.20 0.98 0.80 1.46 25.40 26.30 J1649+4133 S E/W 1.25 0.94 1.38 1.66 J1705+3940 G (0.701) 130 528 S/N 1.70 0.96 2.52 1.27 25.84 26.07 J1706+4340 S S/N 1.24 0.98 8.67 0.59 5.2 MethodologyofclassiﬁcationoftheDDRGs 87

2.5

1.5 Armlength ratio: outer

0.5 0.5 1 1.5 2 2.5 3 3.5 4 Armlength ratio: inner 3

2.5

1.5

Flux density ratio: outer 1

0.5

0 0 2 4 6 8 10 Flux density ratio: inner Figure 5.3: Upper panel represents arm-length ratio of the outer double vs the arm-length ratio of the inner double where as Lower panel represents flux density ratio of outer double vs flux density ratio of inner double. The + signs represent the DDRGs identified from the FIRST survey, while the open circles represent those from Saikia et al. (2006). 88 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

5.3 Results and discussions

5.3.1 Linear sizes

A comparison of the sizes of the DDRGs selected from the FIRST survey (Table 5.1) with the sample compiled by Saikia, Konar, & Kulkarni 2006 shows that the median sizes of both the inner and outer doubles are smaller for the FIRST sources. Most of the sources in the Saikia et al. (2006) compilation were from lower-resolution surveys such as the NVSS, and there was a natural bias against selecting DDRGs with small inner and outer doubles. The median values of the inner and outer doubles in the FIRST DDRGs are

75 and 530 kpc respectively, compared with 165 and 1100 kpc from the Saikia et al. ∼ ∼ compilation. Studies of speciﬁc smaller sources such as 3C293 (Joshi et al. 2011) and 4C02.27 (Jamrozy, Saikia, & Konar 2009) suggest that time scales of episodic activity could be as small as 105 yr, compared with time scales of 108 yr for Mpc-scale objects ∼ ∼ (e.g. Schoenmakers et al. 2000b; Konar et al. 2006), making it diﬃcult to understand episodic activity in terms of periodic supply of fuel by an interacting companion. The smaller sizes of FIRST DDRGs demonstrate clearly that a wide range of time scales of episodic activity are possible, which may extend to even CSS and GPS scale objects as has been speculated for the archetypal GPS object CTA21 (Salter et al. 2010).

In Fig. 5.2 we plot the ratio of the luminosities of the outer double to that of the inner one versus the projected linear size of the inner double. Schoenmakers et al. (2000b) suggested an inverse correlation between these two parameters with the outer lobes being more luminous than the inner ones, while Saikia et al. (2006) found the inner ones to be more luminous in the smallest inner doubles and the overall correlation to have a reduced level of significance. Considering the FIRST DDRGs along with the Saikia et al. sample, we find that while there may be an upper envelope to this diagram suggesting an inverse relation, the correlation is not statistically significant. The range 5.3 Results and discussions 89 of the ratio of luminosities appears large for sources <200 kpc, varying by a factor of ∼ over 100. Note that the only source in the new sample where the inner double is much more luminous than the outer one has a size of only 34 kpc, the second smallest in the sample. The evolution of the smallest sources could be affected by the dense interstellar medium of the host galaxy leading to more luminous componentsdue to a moreefficient dissipation of energy as well as better confinement by the medium (e.g. Jeyakumar et al. 2005, and references therein).

5.3.2 Arm-length and brightness ratios

We have re-examined the trend for the inner doubles to be more asymmetric in both brightness and location compared with the outer ones for the sample compiled by Saikia et al. (2006) Such studies might provide clues towards understanding the environments in which these sources are evolving as well as any possible asymmetries in the oppositely- directed radio jets. For the FIRST sample the estimates of arm-length ratios have been made from the positions of the optical objects and it would be useful to make these estimates from higher frequency images with higher resolution where the cores are likely to be identified. Considering both the samples together there is a marginal trend for the inner double to be more asymmetric in the location of the outer lobes, but the two distributions are not significantly different (Fig. 5.3). However, the asymmetries in the outer and inner lobes are often not in the same sense, suggesting that these are not likely to be due to the effects of orientation and relativistic motion.

A comparison of the flux density ratios of the inner and outer doubles also shows this to be not significantly different. There is also no significant trend for the farther component to be brighter for either the inner or outer doubles. A closer brighter component is expected when the jet on this side is interacting with denser material as it propagates outwards (Fig. 5.3). 90 5.Double-doubleradiogalaxiesfromtheFIRSTsurvey

5.4 Concluding remarks

Early studies of DDRGs suggested that these are likely to be associated with giant radio sources, yielding time scales of episodic activity of 108 yr or so (e.g. Schoenmakers et al. ∼ 2001; Saikia, Konar, & Kulkarni 2006; Saikia & Jamrozy 2009 and references therein). Identiﬁcation of 23 DDRGs from the FIRST survey has shown that these often occur in sources with overall projected sizes of hundreds of kpc as is seen in the misaligned DDRG 3C293 (Joshi et al. 2011). Salter et al. (2010) also speculated that radio emission from the archetypal GPS source CTA21 may be episodic. Examples of DDRGs appear to occur over a wide range of size scales, and surveys of diﬀerent resolutions are required to be able to identify these objects. Detailed spectral and dynamical age studies will help us explore the range of time scales of episodic nuclear activity. In this chapter we also list the candidate DDRGs which require further observations to determine whether it is a DDRG, and also the ones we classify as non-DDRGs and WATs in the Appendix B. Chapter 6

Conclusion and discussion

The multifrequency investigations of radio galaxies and DDRGs, as carried out in this thesis have clariﬁed several issues like their morphology, radio spectra, the injection spectral indices, time scale of episodic activity and ages. As described in Chapter 1, the main objectives of this thesis is spectral ageing analysis of large radio sources, to examine any sign of restarting of nuclear activity in these sources, estimate time scales of episodic activity and to explore new candidates which show signs of episodic activity. The summary and conclusions of the work on the basis of our objectives presented in this thesis and strategies for the future work are described in the ensuing sections.

6.1 Summary of conclusions

3C46: We have selected one large FRII type radio galaxy 3C46 to estimate its spectral indices and spectral ages as well as to examine any possible evidence of episodic activity. We studied 3C46 over a large frequency range. We examined the variation of spectral age through out the source. Like all other radio galaxies, 3C46 also shows an increase of spectral age with distance from hots spots. The maximum spectral age determined for 3C46 is nearly 15 Myr which is quite similar for other 10 large radio galaxies ∼ examined by Jamrozy et al. (2008). Signatures of episodic activity are most often seen

91 92 6. Conclusion and discussion in the form of morphological structures such as a new double in between the old lobes or steep-spectrum extended core component which can be a composite of several radio components. Our study do not show any diﬀuse features due to earlier cycle of activity or any steep-spectrum radio core for 3C46.

3C293: The morphologies of the DDRGs and the variations in the spectral index over them have been used to probe the history of the central engine. Such sources are excellent opportunities for studying the time scales of episodic activity. The range of time scales varies widely, starting from 105yr it goes up to 108yr. We have studied a misaligned ∼ DDRG, 3C293 over a large frequency range. 3C293 is relatively small in linear size with bright steep-spectrum extended core component. High resolution observation of the central core region resolves the complex asymmetric structure of the inner double which is extended only up to 4.2kpc. The newly formed jets move through the ISM of the ∼ host galaxy just like CSS or GPS sources. The outer and inner doubles are asymmetric in both brightness and location relative to the radio core. Though our observations cannot resolve the inner components of the central region, from a determination of the spectra of the inner and outer doubles over a large frequency range, we have estimated time scale of episodic activity to be 105yr which is signiﬁcantly smaller compared with most other ∼ known DDRGs.

New DDRGs and symmetry parameters: It is important to identify more DDRGs to clarify several unanswered questions about their restarting of nuclear activity, time scales between two episodes of AGN activity, propagation of jets in different media and their interaction with different external environments. Study of symmetry parameters of the outer and inner doubles can probe the environments through which the jets propagate. Saikia et al. (2006) reported that the inner doubles appear to be more asymmetric in both its armlength and its flux density ratios than the outer doubles. On the other hand some DDRGs (e.g. 4C29.30, J1453+3308) show a significant core variability where as some of 6.1 Summary of conclusions 93 the DDRGs (e.g. J0840+2949, J1453+3308, J1548 3216) show a trend of very similar − injection spectral indices for inner and outer doubles (Saikia & Jamrozy 2009). It has been also reported that the accumulation of CO (e.g. 3C293) and HI (e.g. 3C236) gas towards the central region of DDRG may be responsible for their restarting of nuclear activity. Moreover some DDRGs are also identified with merger galaxies. In order to understand whether all the above mentioned features are an intrinsic property of the DDRGs or not, a larger sample is required.

Proctor (2011) reported 242 sources as DDRGs based only on their radio structures as seen in the FIRST 20 cm survey. To ﬁnd out new samples of DDRGs, we made a detailed study of their radio morphology and the optical host positions. We identiﬁed only

23 sources which are good examples of DDRGs, along with 63 sources, which require further observations to confirm their episodic nature. The symmetry parameters of the inner and outer doubles of these 23 DDRGs have been examined. We find the armlength and the flux density ratios of the inner doubles to the outer doubles are not significantly different. The overall projected linear sizes of these samples are quite small. In table 6.1 we have shown some example of DDRGs with their time scale of recurrent activity as well as the size of their inner and outer lobes. The table 6.1 is arranged as follows. Column 1: source name; Columns 2 and 3: projected linear size of the inner and outer double- lobed sources in kpc; Column 4: time scale in yr and Colums 5: References. Here we can see the linear size most of the DDRGs are quite large except 3C293 (J1352+3126) and Cygnus A (J1959+4044). The time scales of these two sources are relatively small than other listed DDRGs. So, these new 23 DDRGs will help us explore the range of time scales of episodic nuclear activity from multi-frequency radio observations. 94 6. Conclusion and discussion

Table 6.1: DDRGs from the FIRST survey

Source lin lo timescale Notes name kpc kpc yr (1) (2) (3) (4) (8) J0016 4722 460 1447 7x107 1 − ∼ J0041+3224 171 969 20x106 2 ∼ J0840+2949 36 639 108 3 ∼ J0929+4146 652 1875 108 4 ∼ ⊕ J0935+0204 70 470 <106 5 J1158+2621 138 483 ∼108 4 ∼ ⊕ J1242+3838 251 602 108 4 ∼ ⊕ J1352+3126 4.2 190 105 6 ∼ J1453+3308 159 1297 108 4 ∼ ⊕ J1548 3216 313 961 (5 10)108 7 − ∼ − J1835+6204 369 1379 <106 8 J1959+4044 136 ∼106 9 1: Saripalli, Subrahmanyan, & Udaya Shankar 2002; 2: Jamroz∼ y, Konar, Saikia, Stawarz, Mack, & Siemiginowska 2007; 3:Saikia, Konar, & Kulkarni 2006; 4: Kaiser, Schoenmakers, & Röttgering 2000; 5: Jamrozy, Saikia, & Konar 2009; 6: Joshi, Nandi, Saikia, Ishwara-Chandra, & Konar 2011; 7: Safouris, Subrahmanyan, Bicknell, & Saripalli 2008; 8: Schoenmakers, de Bruyn, Röttgering, & van der Laan 2000a; 9:Steenbrugge, Blundell, & Duﬀy 2008

⊕ Estimated from dynamical age.

6.2 Future work:

Each wavelength regime provides a lot of new informations about the astrophysical systems. A multi-wavelength approach of studying radio galaxies will be useful to probe several unanswered question about their central AGN activity, time scale of episodic activity, host galaxy properties etc.

Low-frequency studies of DDRGs: In order to enlarge the sample of DDRGs with estimates of time scales of episodic activity, we would like to make a low-frequency study of 23 sources which we believe to be good examples of DDRGs. Low-frequency imaging is important to investigate the older relic lobes formed due to earlier cycle of activity. An observing proposal1 has been accepted further low frequency studies using the GMRT.

1proposal code 23_056; by S. Nandi, D. J. Saikia, M. Singh, H. C. Chandola 6.2 Future work: 95

This will provide information on the diﬀerences in spectra between the outer and inner lobes as expected for DDRGs, estimate the dynamical ages of the inner and outer lobes, and time scales of the episodic nuclear activity.

Optical follow up of AGN with recurrent activity: Most of the radio galaxies are associated with evolved elliptical host galaxies. However, there are some exceptional cases of spiral-host radio galaxies (e.g., Hota et al., 2011). It will be very interesting to character- ize the spectral natures of the optical hosts of the double-double radio systems, listed by Nandi & Saikia (2012) as well as to identify the possible associations of spiral hosts. A further development in this regard will be a comparative study between the time scales of star formation in these host galaxies and the synchrotron ages of the radio lobes. O’Dea et al. (2001) shows multiple episodes of star formation in the host of DDRG 3C236. They estimated the age of two star forming regions as 107 yr and the age for other two regions ∼ as 108-109 yr which is comparable to the synchrotron age of the giant radio source. ∼ One can also study the possible dependency of radio activity of the central AGNs on the morphology of the host galaxies and their metallicity using optical spectra of the host galaxies. Certainly, this work has direct implication towards the understanding the activities of central AGNs in these highly chaotic systems.

The inﬂuence of host galaxy properties on its radio morphology and the occur- rence of episodic AGN activities has not been largely explored. The gas properties of the central part of the host may provide important information regarding the origin of radio emission and their episodic nature. To sustain a steady state process, a regular supply of fresh gas towards the central engine is needed. It is still not very clear whether this gas is injected during gravitational or tidal interaction of the host with its surrounding or is an eﬀect of internal dynamics of gas and stars present in the parent galaxy (Ledlow & 96 6. Conclusion and discussion

Owen 1996). Saripalli & Mack (2007) suggested that accumulation of large gas clouds in the center of the galaxy may interrupt the jet activity and restart new episodes. Taking a sample of 43 powerful radio galaxies, mostly FRII type, Heckman et al. (1986) found that 35% of the hosts in the samples show strongly peculiar optical morphologies such as dust features, tidal bridges, shells etc. Colina & Perez-Fournon (1990) and de Juan et al.

(1994) reported that the 50% FRI sources of their samples have nearby companion or they are peculiar in optical morphology. Ramos Almeida et al. (2011) compared hosts of 46 powerful radio galaxies with various samples of quiescent elliptical which have similar surface brightness and concluded that morphological disturbance is quite high for their samples. So it would be very interesting to ﬁnd out the characteristics of the host galaxy of DDRGs and to compare them with the morphologies of classical FRI and FRIIs and understand the possible causes of episodic nuclear activities.

In addition to deep imaging, surface photometry of such new samples will be useful to conﬁrm the morphologies as well as total mass of these systems. It will be an extension of the work by Ramos Almeida et al. (2011) and O’Dea et al. (2001). The possible association of star formation scenario with the jet activities of radio counterpart can be addressed through Hα photometry of the hosts.

Millimeter and submillimeter follow up: The radio galaxies are often found to be strong sources at millimeter and submillimeter wave lengths. This could be due to synchrotron emission from compact core or thermal emission from the heated dust grains. Both the possibilities are quite important in case of radio galaxies. If the emission is nonthermal it gives information about the energetics of the synchrotron source, while the thermal detection implies the presence of large amount of interstellar dust and gas of the host galaxy (Knapp & Patten, 1991). Study of these radio galaxies both in radio and Infrared 6.2 Future work: 97 regime helps to understand the evolution process and recurrent activity of the central

AGN. Sometimes the radio observations do not show any spectral break over a large radio frequency range. In such cases we have to assume a break frequency or some upper limit to estimate the spectral age. Taking a sample of 26 radio galaxies Knapp & Patten (1991) found presence of spectral break between the centimeter and infrared wavelength. So sub-millimeter and far infrared wavelength observations are required to estimate a reliable spectral break for these sources. Using radio, sub-millimeter and far infrared wavelengths a reliable age estimation can be done. More over to understand the recurrence of activities in the AGNs, there have been attempts to detect CO gas at millimeter wavelengths in the central regions of these sources. The study of these sources in radio, sub-millimeter and far infrared wavelengths allow us to understand dynamics and time scales of recurrent activities, the nuclear fueling characteristics of DDRGs. Neutral hydrogen 21-cm observations: Although there have been suggestions that there is a high detection rate of H gas in DDRGs, and the spectra often have complex line pro- ﬁles, the number of detections are still small (cf. Saikia & Jamrozy, 2009, and references therein). This needs to be clearly extended to a large sample of sources with evidence of recurrent activity, perhaps with the new generation of radio telescopes since the cores are often quite weak to be able to detect H in absorption. 98 6. Conclusion and discussion Appendix A

Appendix A: Catalogue of 320 sources detected at 154MHz

The images of three of the extended sources detected at 154 MHz are shown here, along with an extract from the catalogue of 320 sources detected at this frequency in the image

2 with an angular resolution of 15 12.5 arcsec along PA 73◦. The catalogue consists of × sources with a peak flux density 7 times the local rms noise value. Amongst the extended ≥ sources detected at this frequency, Fig. A.1 shows a wide-angle tailed (WAT) source associated with a galaxy at a redshift of 0.061. This was detected 2.5◦ north of the phase ∼ centre. A lower resolution image of the source shows diffuse emission towards the south- east which has been estimated to have a spectral index of 1.02 between 154 and 1400 ∼ MHz using the NVSS image, while the WAT has a spectral index of 0.91. The diffuse ∼ blob has a tail extending approximately in the direction of the WAT host galaxy, and is likely to be related to the WAT source. There is no known optical counterpart for this source.

Approximately 2.5◦ to the south-east of the phase centre we detect the large radio galaxy J1400+3019 (Fig. A.2, left panel), which is associated with a galaxy at a redshift of 0.206, and has an overall angular size of 649 arcsec, indicating a projected linear size of 2170 kpc (Parma et al. 1996; Ishwara-Chandra & Saikia 1999). The spectral index of

99 100 A.AppendixA:Catalogueof320sourcesdetectedat154MHz

NVSS-1400 MHz

33 06 33 06 33 06

05 05 05

04 04 04

03 03 03 DECLINATION (J2000)

DECLINATION (J2000) 02 DECLINATION (J2000) 02 02

01 01 01

00 00 00

13 53 20 15 10 05 00 52 55 50 13 53 20 15 10 05 00 52 55 50 13 53 20 15 10 05 00 52 55 50 RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) Cont peak flux = 1.4686E+01 JY/BEAM Cont peak flux = 1.5763E+01 JY/BEAM Cont peak flux = 5.0243E-02 JY/BEAM Levs = 4.000E-03 * (-2, 2, 3, 5, 7, 10, 14, 20, 28, 40, 56, 80, 112, 160) Levs = 7.000E-03 * (-2, 2, 3, 5, 7, 10, 14, 20, 28, 40, 56, 80, 112, 160) Levs = 1.200E-03 * (-1, 1, 2, 4, 8, 16, 32)

Figure A.1: A WAT source associated with a galaxy at a redshift of 0.061 along with evidence of diﬀuse emission. The left panel shows the higher resolution GMRT image 2 at 154 MHz with a restoring beam of 15 12.5 arcsec along PA 73◦, while in the middle × 2 panel the restoring beam is 41 39 arcsec along PA 159◦. The right panel shows the NVSS image at 1400 MHz. Crosses× mark the position of the optical galaxy. this source between 154 and 1400 MHz is 1.01. ∼ Approximately 30 arcmin towards the west of 3C293, the 154-MHz image shows an FRI galaxy, J1356+3126 (Fig. A.2, right panel), which resembles the morphology of

3C31. There are three galaxies in the vicinity of the centroid, of which the one closest to the peak, SDSS J135620.77+312627.3 with a redshift of 0.151 is likely to be the identiﬁ- cation. The nearby companion SDSS J135621.36+312631.7 is only 8.7 arcsec towards ∼ the north-east. Although its redshift has not been measured, its separation would be 23 ∼ kpc at a redshift of 0.151 and could be a companion galaxy. It may be relevant to note that NGC383, the galaxy associated with 3C31 has a nearby companion and is a member of a chain of galaxies Arp 331 (e.g. Strom et al., 1983). The overall angular size of the radio source J1356+3126 is 120 arcsec which corresponds to a projected linear size of ∼ 310 kpc. Its spectral index is 0.6 between 154 and 1400 MHz. ∼ 101

30 22

18 31 27 30

DECLINATION (J2000) 16

26 45

DECLINATION (J2000) 15

12 25 45

14 00 55 50 45 40 35 RIGHT ASCENSION (J2000) 13 56 24 23 22 21 20 19 18 17 16 15 Cont peak flux = 1.5763E+01 JY/BEAM RIGHT ASCENSION (J2000) Cont peak flux = 1.4686E+01 JY/BEAM Levs = 9.000E-03 * (-2, 2, 3, 5, 7, 10, 14, 20, Levs = 2.000E-03 * (-2, 2, 3, 5, 7, 10, 14, 20, 28, 40, 56, 80, 112, 160) 28, 40, 56, 80, 112, 160, 224, 320, 448, 640)

Figure A.2: A known giant radio galaxy with an FRII-type structure (left panel) and an FRI-type galaxy resembling the morphology of 3C 31 (right panel). Crosses in the images mark the positions of the optical galaxies. The restoring beam for the giant radio galaxy, 2 J1400+3019, is 41 39 arcsec along a PA=159◦, while for the FRI galaxy, J1356+3126, ×2 it is 15 12.5 arcsec along PA=73◦. × 102 A.AppendixA:Catalogueof320sourcesdetectedat154MHz

Table A.1: Sources within 2.2◦ of the phase centre of the GMRT observations at 154 MHz (HPBW 3.1 ). All sources with a peak flux density 7-σ, where σ is the primary beam cor- ∼ ◦ ≥ 1 rected local rms noise in units of mJy beam− are listed. The values of σ range from 8.6 mJy beam 1 near 3C293 to typical values of 3 4 mJy beam 1 in regions without strong sources. There − − − are a total of 320 sources. All but three of the weaker sources (J1352+3059, J1352+3039 and J1357+3111) are seen in the NVSS images at 1400 MHz. Using 5 times the rms noise in the NVSS images yields spectral indices steeper than 1.2 for these three sources. Columns 1 and 2: ∼ the right ascension (h:m:s) and declination (d:m:s) in J2000 co-ordinates; columns 3 and 4: the 1 peak and integrated flux densities in units of mJy beam− and mJy respectively; column 5: the angular size of the source in arcsec where U denotes an unresolved source. The flux densities and source sizes have been estimated as described in Sirothia et al. (2009a).

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:42:08.04 31:19:32.6 172.1 330.4 30.7 13:42:31.58 32:03:29.1 36.6 36.6 U 13:42:51.04 31:37:56.4 67.2 67.2 U 13:42:51.39 31:11:04.6 204.3 518.0 95.4 13:43:03.02 31:43:50.0 2341.9 3961.1 59.5 13:43:07.66 31:27:50.3 319.2 497.4 44.2 13:43:08.30 30:31:10.7 34.0 34.0 U 13:43:13.10 32:08:01.0 33.0 33.0 U 13:43:28.67 31:16:30.0 61.2 61.2 U 13:43:53.47 32:43:28.5 36.6 36.6 U 13:44:00.76 32:11:27.3 148.1 148.1 U 13:44:26.84 31:34:51.7 424.3 934.2 39.5 13:44:36.79 31:04:25.1 179.3 221.8 12.6 13:44:39.72 32:01:23.1 1519.7 2003.7 61.6 13:44:54.55 31:51:43.5 34.0 34.0 U 13:44:58.49 30:15:53.1 334.6 699.0 85.5 13:45:10.70 32:10:53.7 404.8 674.3 23.6 13:45:29.69 32:27:06.8 44.6 44.6 U 13:45:38.98 30:59:15.5 41.2 41.2 U 13:45:41.18 30:08:35.9 36.0 36.0 U 13:45:41.63 31:24:05.6 181.1 336.2 31.1 13:45:43.02 31:33:41.6 102.2 125.4 14.6 103

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:45:48.10 32:32:28.0 35.6 35.6 U 13:45:49.14 31:41:19.9 32.2 32.2 U 13:45:49.56 32:07:57.9 34.4 34.4 U 13:45:50.50 31:03:25.4 49.8 49.8 U 13:45:54.80 33:07:55.7 56.3 56.3 U 13:45:57.01 30:18:19.9 41.3 41.3 U 13:45:57.88 32:47:26.6 33.1 33.1 U 13:45:59.25 32:29:25.1 47.4 47.4 U 13:46:12.01 30:17:01.9 66.2 66.2 U 13:46:12.15 32:34:18.2 230.8 326.3 33.6 13:46:17.27 30:06:47.6 123.9 271.8 41.5 13:46:21.31 32:49:00.9 303.2 495.1 30.6 13:46:27.07 30:20:58.4 54.6 54.6 U 13:46:29.65 32:51:19.1 86.3 165.7 28.0 13:46:34.13 31:29:40.5 81.6 159.7 18.3 13:46:40.04 32:34:17.8 27.7 27.7 U 13:46:48.00 30:05:01.9 153.8 153.8 U 13:46:51.53 29:45:34.2 53.5 53.5 U 13:46:54.61 32:19:03.4 88.1 88.1 U 13:46:57.66 31:16:53.8 165.1 165.1 U 13:46:58.72 30:03:28.5 64.1 64.1 U 13:47:01.80 31:09:12.5 46.4 354.3 98.5 13:47:15.55 30:33:17.0 582.1 783.1 78.4 13:47:16.33 32:17:16.7 45.0 45.0 U 13:47:17.65 31:24:42.4 59.5 163.8 64.7 13:47:19.13 32:21:46.6 103.7 103.7 U 13:47:19.87 31:21:19.8 69.4 69.4 U 13:47:26.99 31:17:23.9 84.5 84.5 U 13:47:31.19 30:45:52.4 170.8 221.4 38.4 13:47:31.71 32:03:45.3 50.4 50.4 U 13:47:34.69 31:28:15.0 21.5 21.5 U 13:47:37.60 31:59:15.5 57.4 57.4 U 13:47:40.17 32:12:01.5 29.0 29.0 U 104 A.AppendixA:Catalogueof320sourcesdetectedat154MHz

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:47:47.99 32:58:18.9 405.0 1112.1 74.1 13:47:52.54 33:22:20.3 70.9 70.9 U 13:47:54.69 32:06:13.7 128.9 156.0 13.4 13:47:55.14 32:00:42.7 110.6 163.0 72.0 13:47:58.59 33:00:33.4 52.8 52.8 U 13:48:00.52 33:03:57.0 72.2 72.2 U 13:48:01.86 31:05:48.3 38.4 38.4 U 13:48:03.04 32:26:07.2 31.7 31.7 U 13:48:03.34 30:10:18.0 231.9 543.9 56.6 13:48:04.30 30:52:30.1 38.3 63.4 27.3 13:48:06.18 31:15:51.9 56.6 77.4 22.9 13:48:06.82 33:23:28.9 65.8 65.8 U 13:48:08.72 30:49:06.2 47.1 47.1 U 13:48:11.57 30:37:56.8 42.6 42.6 U 13:48:13.71 32:51:21.9 135.8 242.5 31.3 13:48:17.11 31:03:10.1 34.3 70.4 48.5 13:48:18.18 33:05:53.5 33.8 33.8 U 13:48:20.32 30:20:02.9 117.3 320.7 134.8 13:48:20.71 32:22:14.0 24.1 24.1 U 13:48:30.41 32:54:58.8 46.9 46.9 U 13:48:31.84 30:39:25.4 40.1 40.1 U 13:48:42.13 31:33:04.7 163.5 191.0 26.7 13:48:42.32 33:10:18.4 60.1 60.1 U 13:48:48.02 29:39:18.9 2048.4 4951.0 44.3 13:48:49.83 31:10:18.5 110.0 155.3 28.8 13:48:51.08 32:02:28.4 63.9 121.7 48.6 13:48:52.13 32:09:25.3 25.8 25.8 U 13:48:55.10 31:57:32.4 394.0 502.4 41.4 13:48:56.82 33:17:10.4 140.2 193.0 16.5 13:48:57.33 29:28:53.7 45.7 45.7 U 13:48:59.60 32:59:47.6 44.3 44.3 U 13:49:03.00 31:40:09.4 34.4 34.4 U 13:49:04.19 32:18:32.2 42.0 75.4 29.0 105

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:49:08.61 30:51:50.2 43.5 43.5 U 13:49:11.10 31:32:07.2 59.7 59.7 U 13:49:11.48 31:40:31.6 32.8 32.8 U 13:49:11.68 31:33:49.2 30.5 30.5 U 13:49:14.00 33:03:16.9 71.7 71.7 U 13:49:29.28 31:22:54.0 53.6 53.6 U 13:49:34.12 30:04:34.0 39.9 39.9 U 13:49:37.94 31:44:47.2 65.5 65.5 U 13:49:39.76 30:08:48.8 76.0 76.0 U 13:49:48.26 31:01:54.9 86.9 164.8 29.3 13:49:50.51 29:35:37.5 74.0 74.0 U 13:49:52.63 30:15:22.9 55.7 189.5 118.2 13:49:58.19 31:04:13.1 23.8 23.8 U 13:49:58.52 32:49:56.1 72.2 133.4 54.4 13:50:00.33 30:04:15.2 124.7 197.0 16.6 13:50:06.10 31:03:54.3 23.8 23.8 U 13:50:11.22 29:42:54.0 49.2 49.2 U 13:50:11.84 31:32:34.9 24.0 24.0 U 13:50:11.97 30:22:44.8 705.9 1274.6 89.3 13:50:16.38 32:36:59.4 36.2 36.2 U 13:50:18.36 30:37:39.1 72.5 171.0 73.6 13:50:19.39 30:22:39.8 72.7 72.7 U 13:50:20.80 30:32:05.0 191.3 249.3 27.7 13:50:31.47 30:57:24.9 191.2 282.2 72.7 13:50:31.90 33:01:34.2 49.0 215.7 118.6 13:50:32.50 30:10:47.9 39.7 39.7 U 13:50:33.48 30:07:51.7 28.8 60.2 32.2 13:50:35.43 30:40:17.3 53.8 53.8 U 13:50:36.64 30:37:30.0 79.0 79.0 U 13:50:39.43 29:31:05.9 39.9 39.9 U 13:50:45.65 31:28:33.1 136.6 237.7 32.0 13:50:46.59 30:13:25.7 101.8 101.8 U 13:50:47.83 33:07:30.7 139.8 159.8 13.0 106 A.AppendixA:Catalogueof320sourcesdetectedat154MHz

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:50:48.29 33:12:12.3 626.9 1269.2 64.2 13:50:49.26 33:05:31.0 70.4 70.4 U 13:50:49.94 29:42:51.3 45.0 45.0 U 13:50:50.05 30:09:12.3 244.4 319.4 26.6 13:50:52.54 30:34:52.8 289.5 370.7 48.0 13:50:56.90 33:31:46.3 179.0 443.3 32.7 13:51:01.11 30:42:44.1 33.4 33.4 U 13:51:01.97 32:34:35.9 51.7 51.7 U 13:51:02.66 31:14:28.3 108.5 108.5 U 13:51:03.14 30:53:56.1 338.7 678.4 57.1 13:51:11.21 33:17:38.2 44.5 44.5 U 13:51:12.55 33:21:51.7 127.0 249.6 37.0 13:51:16.62 30:55:53.0 60.2 248.7 115.5 13:51:20.97 33:07:09.9 197.3 197.3 U 13:51:21.99 31:33:41.1 200.5 200.5 U 13:51:23.30 30:07:37.8 61.1 61.1 U 13:51:26.10 30:24:41.8 167.9 191.7 31.9 13:51:27.82 31:04:01.9 121.6 147.8 35.2 13:51:31.09 30:25:01.7 28.4 28.4 U 13:51:35.18 29:33:53.7 108.4 201.0 41.3 13:51:39.06 33:21:57.7 127.6 127.6 U 13:51:41.54 31:04:43.4 40.4 40.4 U 13:51:42.82 30:02:55.7 33.2 33.2 U 13:51:47.14 32:06:14.7 36.8 36.8 U 13:51:53.81 33:05:52.3 56.4 56.4 U 13:52:05.31 32:52:27.8 32.8 32.8 U 13:52:12.58 32:08:15.0 48.7 48.7 U 13:52:15.83 32:29:09.6 37.6 37.6 U 13:52:15.99 32:58:11.6 56.6 239.8 89.6 13:52:16.76 31:26:55.2 15216.0 21060.6 131.5 13:52:19.73 30:44:33.1 183.8 213.9 10.9 13:52:21.29 32:49:05.0 31.8 31.8 U 13:52:22.92 29:49:43.7 49.6 76.7 29.6 107

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:52:23.49 31:25:45.8 213.5 895.2 74.5 13:52:26.47 30:20:07.5 719.3 1096.8 60.9 13:52:33.06 32:39:00.1 23.8 23.8 U 13:52:33.60 30:59:38.2 39.6 39.6 U 13:52:38.04 32:31:49.5 38.7 38.7 U 13:52:49.41 32:13:00.8 43.9 61.9 25.9 13:52:53.40 30:39:08.5 25.8 25.8 U 13:52:54.50 29:32:36.8 41.3 41.3 U 13:52:55.32 31:38:05.1 133.7 172.8 19.0 13:52:56.75 29:25:58.4 133.3 133.3 U 13:52:59.74 33:04:40.3 94.9 987.4 167.5 13:53:07.07 33:36:22.7 51.0 187.8 62.0 13:53:11.60 32:05:40.9 986.2 1046.2 15.5 13:53:12.52 32:54:03.9 106.6 134.1 16.2 13:53:13.35 31:12:45.0 42.7 42.7 U 13:53:17.18 30:40:32.6 73.9 73.9 U 13:53:19.15 31:09:06.2 73.0 73.0 U 13:53:25.00 29:34:13.5 67.1 67.1 U 13:53:27.54 32:18:21.2 355.3 413.6 8.3 13:53:31.12 32:45:57.9 153.5 306.7 59.5 13:53:37.94 31:33:21.9 39.7 39.7 U 13:53:38.99 30:40:24.7 53.0 189.3 109.0 13:53:40.02 29:26:35.7 82.3 82.3 U 13:53:41.66 33:13:36.6 49.5 175.2 87.5 13:53:43.30 30:51:56.6 138.6 242.4 51.8 13:53:43.74 31:47:06.7 37.7 37.7 U 13:53:44.66 33:04:39.8 49.3 104.9 46.8 13:53:45.38 31:51:51.6 3207.9 5079.4 62.6 13:53:47.44 30:05:50.7 47.0 47.0 U 13:53:49.17 32:09:10.1 53.2 53.2 U 13:53:50.97 29:56:19.1 70.4 70.4 U 13:53:53.86 30:31:09.8 43.6 43.6 U 13:53:56.31 29:33:07.3 289.9 378.9 29.5 108 A.AppendixA:Catalogueof320sourcesdetectedat154MHz

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:54:00.30 33:36:38.7 81.8 81.8 U 13:54:00.73 30:56:27.2 77.2 138.1 57.4 13:54:00.92 32:57:27.7 77.6 168.3 53.4 13:54:05.23 31:39:01.8 1319.6 1623.7 46.2 13:54:07.44 33:16:54.1 113.0 113.0 U 13:54:08.85 30:48:34.7 72.0 72.0 U 13:54:11.72 30:42:20.1 25.1 25.1 U 13:54:12.81 30:18:45.5 115.4 115.4 U 13:54:14.12 30:38:25.1 64.0 64.0 U 13:54:15.39 30:34:06.6 53.4 53.4 U 13:54:16.46 30:42:05.8 35.5 35.5 U 13:54:25.97 31:52:17.8 37.4 37.4 U 13:54:26.35 30:05:35.3 79.0 100.7 19.1 13:54:30.62 33:13:45.1 39.3 39.3 U 13:54:32.01 32:03:30.6 847.1 1703.2 106.2 13:54:33.34 30:56:04.0 225.9 267.9 47.2 13:54:33.74 29:57:32.4 416.0 512.7 52.1 13:54:37.29 31:44:21.1 65.2 120.9 44.1 13:54:39.32 31:51:51.9 36.4 36.4 U 13:54:48.26 29:21:05.0 104.4 157.3 20.8 13:54:50.43 33:24:15.3 51.1 243.0 59.3 13:54:51.32 32:47:24.5 175.8 229.7 56.9 13:54:51.43 32:31:58.1 89.2 89.2 U 13:54:59.24 30:36:07.1 714.5 764.2 13.1 13:55:01.25 29:43:17.1 592.9 714.9 10.9 13:55:01.36 33:06:27.1 54.9 54.9 U 13:55:04.58 31:32:12.6 31.3 31.3 U 13:55:06.00 30:34:59.2 165.5 243.1 47.0 13:55:06.07 31:51:31.1 90.8 90.8 U 13:55:11.79 32:57:42.8 134.3 134.3 U 13:55:11.91 32:56:12.7 109.1 109.1 U 13:55:12.94 32:17:35.9 379.1 750.8 54.1 13:55:13.15 32:51:59.0 33.1 33.1 U 109

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:55:14.91 31:56:26.7 204.1 204.1 U 13:55:16.45 29:21:54.3 102.7 366.0 91.2 13:55:19.29 29:25:34.7 90.2 187.7 36.9 13:55:21.78 31:44:39.2 240.3 869.9 64.8 13:55:21.79 31:28:15.9 88.8 88.8 U 13:55:22.69 30:34:32.2 89.8 89.8 U 13:55:25.45 29:33:11.6 700.6 817.4 77.4 13:55:26.94 32:34:58.6 92.0 92.0 U 13:55:28.65 32:19:52.2 47.9 47.9 U 13:55:30.13 30:31:30.2 175.9 195.2 13.9 13:55:30.30 31:12:36.5 108.4 108.4 U 13:55:31.93 32:48:28.1 44.8 44.8 U 13:55:35.14 30:33:21.7 53.8 79.4 25.1 13:55:35.89 29:54:00.8 131.7 131.7 U 13:55:37.65 32:29:15.3 60.9 60.9 U 13:55:39.20 29:32:19.6 77.7 77.7 U 13:55:41.11 30:24:15.1 174.0 174.0 U 13:55:46.12 32:38:04.6 566.6 649.8 20.2 13:55:51.35 31:37:34.9 33.3 33.3 U 13:55:56.72 33:29:06.5 46.8 46.8 U 13:55:58.28 31:08:32.9 52.2 52.2 U 13:55:58.79 31:19:39.7 172.3 192.9 15.1 13:56:03.56 30:59:00.6 34.1 34.1 U 13:56:04.82 31:45:07.6 107.7 107.7 U 13:56:07.51 30:04:55.0 105.1 172.2 21.2 13:56:08.03 31:53:05.2 792.1 846.2 9.9 13:56:15.53 30:52:20.2 1037.0 1459.4 60.0 13:56:15.67 30:06:40.4 114.0 114.0 U 13:56:17.08 32:21:14.6 211.0 211.0 U 13:56:19.79 32:03:53.5 58.8 58.8 U 13:56:19.96 31:26:28.0 97.9 436.7 85.6 13:56:24.73 31:17:06.4 173.5 282.7 39.5 13:56:29.14 30:11:24.3 42.7 42.7 U 110 A.AppendixA:Catalogueof320sourcesdetectedat154MHz

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:56:30.33 29:41:19.8 811.5 1007.1 71.3 13:56:36.64 30:36:17.5 29.5 29.5 U 13:56:38.08 31:02:30.2 212.3 212.3 U 13:56:39.46 30:16:51.3 57.2 57.2 U 13:56:44.86 32:19:46.2 2869.4 8854.4 161.2 13:56:46.56 32:53:48.0 51.0 51.0 U 13:56:48.06 30:22:40.1 338.8 619.2 95.6 13:56:50.60 32:04:47.9 176.5 291.9 48.7 13:56:54.91 30:56:41.1 54.3 211.7 86.3 13:57:00.91 30:24:46.7 132.6 201.7 31.0 13:57:00.99 33:19:12.8 50.5 50.5 U 13:57:02.35 31:54:57.9 48.8 48.8 U 13:57:02.81 31:48:13.3 271.4 311.2 32.6 13:57:02.95 31:11:51.9 27.0 27.0 U 13:57:04.05 29:32:29.9 93.8 93.8 U 13:57:08.64 32:58:21.7 120.5 283.1 47.9 13:57:09.71 32:50:44.3 673.5 815.9 10.7 13:57:13.88 31:03:46.1 232.7 362.1 20.6 13:57:14.42 29:57:56.2 48.9 48.9 U 13:57:20.42 33:11:31.2 52.6 52.6 U 13:57:20.54 32:03:15.7 86.0 86.0 U 13:57:22.13 30:41:55.5 1214.5 1470.0 32.6 13:57:23.95 33:01:11.7 94.8 94.8 U 13:57:26.29 29:58:27.9 103.7 162.0 26.7 13:57:37.43 30:46:07.6 37.8 37.8 U 13:57:41.43 30:09:06.4 51.8 51.8 U 13:57:55.22 31:39:06.7 628.1 877.0 52.5 13:58:01.75 32:37:31.3 55.2 55.2 U 13:58:16.15 29:55:33.7 152.2 152.2 U 13:58:19.54 31:18:35.4 1596.0 2819.0 141.1 13:58:26.09 30:28:12.0 136.6 220.7 20.2 13:58:26.41 29:55:44.2 58.0 58.0 U 13:58:27.93 29:57:16.4 48.3 48.3 U 111

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 13:58:27.94 30:31:48.4 75.1 75.1 U 13:58:31.39 29:54:59.9 37.7 37.7 U 13:58:31.88 32:22:35.4 41.7 41.7 U 13:58:33.04 30:32:32.7 32.3 32.3 U 13:58:42.27 31:00:47.6 47.6 47.6 U 13:58:48.19 31:25:55.4 95.7 186.2 34.2 13:58:49.23 32:21:51.9 46.2 46.2 U 13:58:50.25 30:36:19.9 144.4 211.9 23.1 13:58:53.28 30:26:22.7 246.4 433.4 34.9 13:58:53.86 32:21:07.3 30.7 30.7 U 13:59:00.94 31:30:59.7 78.0 115.6 23.2 13:59:01.21 31:33:59.9 112.1 112.1 U 13:59:06.16 29:54:29.8 85.9 125.2 23.9 13:59:14.46 31:41:50.2 35.1 35.1 U 13:59:18.09 30:25:02.5 114.5 270.4 21.1 13:59:24.76 32:30:42.9 78.6 263.7 49.7 13:59:26.41 31:51:20.4 40.1 40.1 U 13:59:27.82 30:14:31.9 46.6 46.6 U 13:59:32.18 30:22:01.0 41.1 41.1 U 13:59:36.41 31:14:09.4 36.5 36.5 U 13:59:37.44 30:06:44.5 38.0 115.9 61.0 13:59:47.69 30:32:26.9 39.7 39.7 U 13:59:53.27 32:25:31.4 110.0 110.0 U 13:59:54.23 31:45:22.8 460.1 559.8 14.7 14:00:06.35 30:36:17.2 47.3 47.3 U 14:00:06.91 32:29:10.8 50.8 50.8 U 14:00:10.86 31:23:29.0 68.5 112.7 29.4 14:00:13.88 32:39:46.4 58.8 58.8 U 14:00:18.08 32:42:32.7 43.0 71.3 35.5 14:00:34.47 30:12:09.6 64.3 64.3 U 14:01:05.43 30:20:18.2 73.2 73.2 U 14:01:21.44 31:37:48.8 30.7 30.7 U 14:01:22.31 31:05:08.1 132.7 243.1 26.5 112 A.AppendixA:Catalogueof320sourcesdetectedat154MHz

Table A.1 –(continued)

RA (h:m:s) Dec (d:m:s) Speak Sint Size 1 mJy beam− mJy ′′ (1) (2) (3) (4) (5) 14:01:32.99 31:08:55.6 37.7 132.1 44.1 Appendix B

Appendix B: List of the candidate DDRGs and non-DDRGs

In this Appendix, we list the candidate DDRGs from the FIRST survey which require further observations to determine whether these might be DDRGs, the ones we classify as non-DDRGs from the list given by Proctor (2011), and the wide-angle tailed sources.

Examples of sources which we classify as candidate DDRGs or non-DDRGs are shown in Figs. B.1 and B.2 respectively.

J1428+2410 FIRST J1448+0357 FIRST J1540+6344 FIRST

63 45 00

24 11 30 03 58 00

44 45

15 57 45

00 30

10 45 DECLINATION (J2000) DECLINATION (J2000) 15 DECLINATION (J2000)

00 30 00

43 45 15 56 45 14 28 46 45 44 43 42 41 40 14 48 52 51 50 49 48 47 46 15 41 04 02 00 40 58 56 54 52 50 RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000) Cont peak flux = 3.6382E-02 JY/BEAM Cont peak flux = 1.3270E-01 JY/BEAM Cont peak flux = 3.6945E-03 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 64) Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 64) Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 64)

Figure B.1: Examples of candidate DDRGs which require an optical identiﬁcation to determine whether the inner structure is due to either an inner double or a radio core and part/knot of a jet.

113 114 B.AppendixB:ListofthecandidateDDRGsandnon-DDRGs

J0759+4051 FIRST J0032-0019 FIRST -00 19 00

40 52 45

15 30

30 15

00 45 DECLINATION (J2000) DECLINATION (J2000) 51 45

20 00

15 15

00 32 50 49 48 47 46 45 RIGHT ASCENSION (J2000) 07 59 14.5 14.0 13.5 13.0 12.5 12.0 11.5 11.0 10.5 10.0 Cont peak flux = 8.6736E-02 JY/BEAM RIGHT ASCENSION (J2000) Cont peak flux = 2.0320E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 64) Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 64)

J1434+0441 FIRST

04 42 15

41 45 DECLINATION (J2000) 30

00 14 34 22.5 22.0 21.5 21.0 20.5 RIGHT ASCENSION (J2000) Cont peak flux = 1.1740E-02 JY/BEAM Levs = 4.000E-04 * (-1, 1, 2, 4, 8, 16, 64)

Figure B.2: Examples of sources which we have classified as non-DDRGs. The inner emission in J0032 0019 is likely to be due to backflow from the hot-spots, while in the case of J0759+4051− and J1434+0441, the optical identification is co-incident with one of the components of the inner double which is likely to be the radio core. The other feature is possibly a knot in the jet. 115

Table B.1: Candidate DDRGs

Source Opt. z RAoptical Decoptical RAcore Deccore Notes

name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J0012 1050 G (0.322) 00:12:08.59 10:50:43.68 00:12:08.64 10:50:43.11 1 − − − J0040 1001 G (0.646) 00:40:31.07 10:01:29.19 1 − − J0142 0000 G (0.478) 01:42:48.84 00:00:34.84 1 − − J0202+0015 1,2 J0251+0039 1,2 J0848+2925 1,2 J0901+1212 G (0.473) 09:01:31.55 +12:12:12.91 1 J0927+2932 G (0.728) 09:27:44.00 +29:32:34.43 09:27:43.88 +29:32:32.37 1,3 J1009+3608 2 J1044+1442 G 0.1547 10:44:34.63 +14:42:04.06 2 J1057+5326 2 J1107+1216 G 3 J1114+3658 2 J1120+1611 2 J1123+4757 G (0.816) 11:23:18.74 +47:57:59.26 1 J1154+3021 S 2 J1204+0756 2 J1233 0412 2 − J1246+3848 G (0.814) 12:46:08.23 +38:48:37.42 1 J1248+2022 1,2 J1306+1922 G (0.573) 13:06:00.63 +19:22:48.74 1 J1308+3955 G (0.104) 13:08:10.19 +39:55:05.42 1 J1327+4946 1,2 J1332+1821 G 0.5049 13:32:24.76 +18:21:00.49 1 J1335+2042 2 J1339+3756 G 0.3750 13:39:47.28 +37:56:52.82 13:39:47.26 +37:56:52.10 1 J1342+0225 1,2 J1342+0642 G (0.052) 13:42:51.27 +06:42:31.05 1 J1343+6306 S 2 J1351+2519 1,2 J1352 0325 2 − J1355+3525 G 0.1078 13:55:26.19 +35:25:44.12 1 116 B.AppendixB:ListofthecandidateDDRGsandnon-DDRGs

Table B.1-cont.

Source Opt. z RAoptical Decoptical RAcore Deccore Notes

name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J1357+3407 2 J1404+5622 2 J1417+2508 2 J1428+2410 2 J1432 0507 14:32:22.27 05:07:15.49 1 − − J1432+2634 G (0.648) 14:32:40.76 +26:34:53.37 1 J1434+2046 S 14:34:13.70 +20:46:13.74 14:34:13.57 +20:46:11.44 1 J1446+0047 S 14:46:36.92 +00:47:04.67 1 J1448+0357 2 J1450 0607 2 − J1459+3836 1,2 J1503+1154 S 15:03:11.75 +11:54:40.83 3 J1511+5650 G 0.6319 15:11:09.21 +56:50:52.08 15:11:09.19 +56:50:51.71 1 J1511+3633 G 0.1638 15:11:25.91 +36:33:36.03 1 J1540+3121 S 15:40:13.86 +31:21:25.08 1 J1540+6344 2 J1546+0844 G 0.1849 15:46:35.78 +08:44:24.63 1 J1551+2654 2 J1554+3233 2 J1604+2703 1,2 J1607+0625 G (0.612) 16:07:19.46 +06:25:46.90 1 J1611+1017 G (0.737) 16:11:14.84 +10:17:12.81 1 J1613+2929 1,2 J1613+5203 S 16:13:21.11 +52:03:01.75 1 J1619+2841 1,2 J1642+4501 G 0.1862 16:42:22.78 +45:01:04.21 1 J1706+3816 2 J1716+5023 1,2 J1720+3015 G (0.634) 17:20:50.99 +30:15:39.29 1 J1727+3727 1,2 J1733+4342 1,2

Notes 1: higher-resolution images required to clarify the structure; 2: optical identiﬁcation is either 117

unavailable or uncertain; 3: separation of optical position from possible core 2 arcsec. ∼ ⊕ In the second column G and S represent a galaxy and star respectively, as listed in SDSS. 118 B.AppendixB:ListofthecandidateDDRGsandnon-DDRGs

Table B.2: Non-DDRGs.

Source Opt. z RAoptical Decoptical RAcore Deccore Notes

name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J0013 0919 G (0.477) 00:13:57.23 09:19:49.29 − − J0032 0019 G (0.625) 00:32:47.93 00:19:36.96 − − J0056 1051 G 0.1959 00:56:41.46 10:52:03.11 00:56:41.46 10:52:03.44 1 − − − J0118+0114 J0729+3226 G (0.423) 07:29:06.37 +32:26:40.39 J0758+1617 G (0.626) 07:58:30.23 +16:17:36.26 3 J0759+4051 G (0.602) 07:59:12.22 +40:51:54.79 07:59:12.22 +40:51:54.42 1 J0838+2404 S 08:38:18.39 +24:04:49.65 08:38:18.40 +24:04:49.26 1 J0902+5707 S 1.5963 09:02:07.20 +57:07:38.09 09:02:07.22 +57:07:37.95 1 J0911+1255 G 0.0495 09:11:34.75 +12:55:38.12 J0912+0810 G (0.736) 09:12:20.19 +08:10:43.24 J0914+1006 4 J0928 0319 09:28:41.93 03:19:50.27 09:28:41.91 03:19:51.04 1 − − − J0942+2710 S 09:42:20.04 +27:10:31.72 09:42:20.04 +27:10:31.58 1 J0948+5758 G (0.337) 09:48:07.56 +57:58:53.24 09:48:07.69 +57:58:53.86 1 J0949+2044 G (0.345) 09:49:12.21 +20:44:14.50 J0953+1403 G 0.2376 09:53:42.24 +14:03:58.01 1 J0959+1558 G (0.767) 09:59:08.01 +15:58:30.32 J1026+3639 G (0.816) 10:26:07.98 +36:40:01.21 10:26:07.99 +36:40:01.47 1 J1028+4306 S 10:28:33.11 +43:06:26.93 10:28:33.14 + 43:06:26.94 1 J1033+0755 S 10:33:40.07 +07:55:57.7 10:33:40.06 +07:55:58.10 1 J1041+5233 5 J1049+0059 G 0.1064 10:49:14.08 +00:59:45.26 10:49:14.07 +00:59:45.18 1 J1053+3125 G (0.647) 10:53:04.71 +31:26:01.34 10:53:04.70 +31:26:01.13 1 J1054+0740 G 0.0968 10:54:52.10 +07:40:06.37 10:54:52.08 +07:40:06.74 1 J1103+0249 G (0.316) 11:03:38.13 +02:49:28.81 J1105+2317 J1139+6030 G (0.707) 11:39:31.52 +60:33:15.18 J1141+0802 G 0.2282 11:41:25.98 +08:02:16.51 11:41:26.01 +08:02:16.90 1 J1150+4046 G (0.380) 11:50:55.10 +40:46:32.09 11:50:55.11 +40:46:31.31 1 J1201+2256 G 0.2594 12:01:41.70 +22:56:46.41 12:01:41.68 +22:56:46.05 1 J1206+1626 G 0.3036 12:06:03.54 +16:26:34.92 2 119

Table B.2-cont.

Source Opt. z RAoptical Decoptical RAcore Deccore Notes

name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J1208+2220 G (0.636) 12:08:21.81 +22:20:03.89 12:08:21.99 +22:19:58.37 J1213+1343 G 0.1743 12:13:06.68 +13:43:17.79 J1214+5107 G (0.718) 12:14:46.64 +51:07:04.30 J1224+0203 S 0.4492 12:24:25.61 +02:03:09.68 12:24:25.54 +02:03:10.82 1 J1224+2358 G (0.612) 12:24:39.31 +23:58:56.69 J1230+1046 J1234+5753 G 0.1529 12:34:24.35 +57:53:27.33 12:34:24.26 +57:53:26.08 J1240+5334 G (0.267) 12:40:12.46 +53:34:37.37 12:40:12.49 +53:34:37.55 J1242+4244 G (0.388) 12:42:37.95 +42:44:03.09 J1243+1508 S 12:43:44.83 +15:08:20.66 12:43:44.83 +15:08:20.61 1 J1248+1725 G (0.581) 12:48:04.08 +17:25:50.56 6 J1248 0301 Q 1.0337 12:48:04.09 03:01:14.54 12:48:04.11 03:01:13.27 1 − − − J1248+5942 G (0.557) 12:48:35.68 +59:42:22.37 J1255+4405 G (0.211) 12:55:54.59 +44:05:21.82 6 J1319+0502 S 13:19:43.58 +05:02:43.02 13:19:43.54 +05:02:43.02 1 J1325+5736 G (0.139) 13:25:11.17 +57:36:01.58 2 J1339+2812 S 13:39:04.29 +28:12:41.19 13:39:04.29 +28:12:41.18 1 J1339 0637 13:39:07.10 06:37:04.96 13:39:07.22 06:37:05.09 1 − − − J1339+1024 G (0.464) 13:39:13.64 +10:24:47.88 13:39:13.42 +10:24:46.44 1 J1343+4627 G 0.2248 13:43:00.36 +46:27:19.99 13:43:00.36 +46:27:19.87 1 J1344+3317 Q 0.6862 13:44:15.75 +33:17:19.13 13:44:15.72 +33:17:18.74 1 J1349 0149 G 0.2098 13:49:18.71 01:49:21.40 13:49:18.76 01:49:20.22 2 − − − J1351+0728 G 0.1500 13:51:10.81 +07:28:46.07 13:51:10.81 +07:28:46.78 7 J1400+0736 G (0.718) 14:00:06.93 +07:37:00.48 J1402+6105 G (0.735) 14:02:14.73 +61:05:31.35 J1410+3749 8 J1412+2301 G (0.486) 14:12:50.38 +23:01:16.61 J1413+0741 G (0.539) 14:13:15.25 +07:41:43.63 J1420+3552 G (0.549) 14:20:37.55 +35:52:51.18 J1423+2448 G (0.585) 14:23:21.33 +24:48:15.05 14:23:21.24 +24:48:15.07 1 J1425 0456 14:25:12.21 04:56:34.71 14:25:12.25 04:56:35.70 2 − − − J1430+5217 G 0.3675 14:30:17.34 +52:17:35.32 14:30:17.14 +52:17:35.74 6 J1431+0538 G (0.245) 14:31:03.50 +05:38:12.45 120 B.AppendixB:ListofthecandidateDDRGsandnon-DDRGs

Table B.2-cont.

Source Opt. z RAoptical Decoptical RAcore Deccore Notes

name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J1431+1922 G 0.2136 14:31:49.14 +19:23:00.12 14:31:49.15 +19:23:59.83 1 J1434+0441 S 14:34:21.56 +04:41:37.21 14:34:21.57 +04 41 38.23 1 J1437+2445 G 0.0862 14:37:15.00 +24:45:32.21 14:37:15.02 +24:45:33.21 1 J1439+5314 G (0.266) 14:39:34.48 +53:14:37.44 14:39:34.47 +53:14:39.65 1 J1439+2824 G (0.367) 14:39:58.41 +28:24:22.66 14:39:58.44 +28:24:22.89 1 J1444+3817 J1445 0626 − J1448+2954 9 J1448+4923 G (0.494) 14:48:59.59 +49:23:40.84 J1450+0001 Q 1.9679 14:50:49.93 +00:01:44.34 14:50:49.94 +00:01:44.25 J1450 0315 14:50:54.99 03:15:52.32 − − J1459+1655 G (0.608) 14:59:36.39 +16:55:25.59 14:59:36.43 +16:55:25.62 1 J1500+1327 G 0.1116 15:00:03.96 +13:27:45.57 15:00:03.99 +13:27:45.80 1,2 J1501+5455 G 0.3388 15:01:17.98 +54:55:18.37 15:01:17.95 +54:55:18.02 1 J1501+4012 G (0.466) 15:01:21.53 +40:12:19.88 15:01:21.57 +40:12:20.04 1 J1525+1253 G 0.2571 15:25:11.68 +12:52:58.03 15:25:11.68 +12:52:57.83 1 J1527+1822 G (0.472) 15:27:37.11 +18:22:51.31 15:27:37.10 +18:22:50.80 1 J1530+2316 G 0.0899 15:30:07.96 +23:16:16.02 15:30:07.94 +23:16:15.74 1 J1530 0703 15:30:58.81 07:03:31.74 15:30:58.90 07:03:32.40 1 − − − J1540+4925 G (0.472) 15:40:28.65 +49:25:14.72 15:40:28.67 +49:25:14.55 1 J1540+0132 Q 0.7743 15:40:47.88 +01:32:07.17 15:40:47.88 +01 32 06.84 1 J1545 0330 15:45:37.96 03:30:46.63 − − J1548+3633 G 0.1963 15:48:05.70 +36:33:40.57 15:48:05.74 +36:33:40.72 1 J1550+2246 G (0.435) 15:50:05.00 +22:46:03.46 15:50:04.97 +22 46 05.82 1 J1557+1618 G 0.0370 15:57:49.61 +16:18:36.59 J1558+0759 G (0.227) 15:58:34.46 +07:59:46.13 2 J1600+1306 S 16:00:58.77 +13:06:59.59 16:00:58.79 +13:06:59.64 1 J1601+3423 G (0.669) 16:01:30.46 + 34:23:12.42 J1608+2828 G 0.0502 16:08:21.14 +28:28:43.29 16:08:21.15 +28:28:43.54 1,2 J1614+3210 G (0.663) 16:14:49.71 +32:10:54.46 J1614+0240 2 J1615+1245 J1616+2809 121

Table B.2-cont.

Source Opt. z RAoptical Decoptical RAcore Deccore Notes

name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J1617+5451 S 16:17:56.89 +54:51:13.80 16:17:56.94 +54:51:13.90 1,2 J1618+1211 G 0.3934 16:18:03.56 +12:11:32.62 16:18:03.60 +12:11:33.70 1 J1620+2017 G (0.056) 16:20:21.74 +20:17:07.32 1 J1623+4318 G (0.687) 16:24:00.24 +43:18:41.14 16:24:00.27 +43:18:41.65 1 J1627+5019 G (0.450) 16:27:24.54 +50:19:40.92 16:27:24.54 +50:19:40.89 1 J1628+3906 G (0.499) 16:28:11.44 +39:06:36.70 16:28:11.42 +39:06:36.65 1 J1628+5750 G (0.262) 16:28:57.46 +57:50:44.77 J1631+1855 G (0.107) 16:31:04.91 +18:55:22.77 16:31:04.95 +18:55:22.26 1 J1633+0847 G (0.191) 16:33:00.85 +08:47:36.44 16:33:00.83 +08:47:36.61 1,7 J1637+4130 S 1.1781 16:37:02.20 +41:30:22.22 J1643+2642 6 J1646+5549 16:46:19.05 +55:49:58.13 16:46:19.19 +55:49:58.54 1 J1647+5154 G (0.303) 16:47:05.21 +51:54:11.47 2 J1649+1651 G (0.563) 16:49:48.74 +16:51:47.14 J1658+3408 J1659+2602 G (0.494) 16:59:52.81 +26:02:38.69 16:59:52.80 +26:02:39.42 J1700+4851 G (0.204) 17:00:35.32 +48:51:03.95 J1702+5312 17:02:01.36 +53:12:19.02 17:02:01.38 +53:12:34.37 1 J1703+5533 J1705+2839 G (0.300) 17:05:02.05 +28:39:06.94 J1710+4611 G (0.759) 17:10:36.66 +46:11:11.75 J1710+4239 G (0.176) 17:10:40.73 +42:39:45.08 17:10:40.75 +42:39:44.80 1 J1713+5812 J1723+3017 G (0.645) 17:23:08.07 +30:17:16.32 17:23:08.04 +30:17:15.79 6 J1725+3452 S 17:25:07.68 +34:52:15.97 J1728+4855 9 J1728+4421 G (0.569) 17:28:56.98 +44:21:37.30 J1732+5634 G 0.3330 17:32:50.22 +56:34:26.61 17:32:50.18 +56:34:27.07 1,2 J2143 0858 − J2149 0004 G (0.650) 21:49:01.07 00:04:22.19 − − J2259 0925 − J2315 0026 G 0.0909 23:15:42.11 00:26:07.05 23:15:42.09 00:26:07.34 1 − − − J2322 0941 G (0.273) 23:22:08.21 09:41:58.00 23:22:08.25 09 41 58.67 1 − − − 122 B.AppendixB:ListofthecandidateDDRGsandnon-DDRGs

Table B.2-cont.

Source Opt. z RAoptical Decoptical RAcore Deccore Notes

name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) (8) J2353+0012 G 0.1663 23:53:55.59 +00:12:56.40 23:53:55.65 +00:12:56.12

Notes 1: optical identification is associated with one of the inner objects; 2: FRI type structure; 3: although inner lobe appears larger, high-resolution observations would be helpful to confirm its classification; 4: optical identification unclear (there are two possible candidates); 5: northern compact component possibly unrelated; 6: X-shaped source; 7: two independent sources nearby; 8: optical object associated with westernmost component; 9: two optical objects along the radio axis with the western one being the likely identification; 10: optical object associated with the northernmost component.

⊕ In the second column G, S and Q represent galaxy, star and quasi stellar object (QSO), as listed in SDSS. 123

Table B.3: Candidate DDRGs from the FIRST survey classiﬁed as WATs.

Source Opt. z RAoptical Decoptical RAcore Deccore name Id.⊕ hh:mm:ss.ss dd:mm:ss.ss hh:mm:ss.ss dd:mm:ss.ss (1) (2) (3) (4) (5) (6) (7) J0041 0925 G 0.05779 00:41:50.17 09:25:47.43 00:41:50.12 09:25:47.99 − − − J0123 1025 G 0.1537 01:23:12.52 10:25:10.38 − − J0930+5930 G (0.618) 09:30:06.88 +59:30:32.82 09:30:06.86 +59:30:32.44 J0948+4218 G (0.372) 09:48:42.95 +42:18:31.35 09:48:42.98 +42:18:31.33 J1028+0345 G (0.108) 10:28:23.47 +03:45:31.51 J1035+4255 G 0.13602 10:35:02.61 +42:55:48.34 10:35:02.59 +42 55 47.68 J1046+1210 J1119+5808 J1151+0422 G (0.150) 11:51:46.90 +04:22:22.73 J1153+2341 S 11:53:18.07 +23:41:13.52 11:53:18.15 +23:41:13.22 J1258+1110 G (0.699) 12:58:22.78 +11:10:39.56 J1335+5352 S 13:35:40.85 +53:52:24.40 J1342+0643 G (0.619) 13:42:38.95 +06:43:25.33 J1406+5504 G 0.2505 14:06:55.11 +55:04:02.87 14:06:54.72 +55 04 01.72 J1421 0743 − J1432+4436 G 0.1815 14:32:43.85 +44:36:14.21 14:32:43.87 +44:36:14.75 J1453+2108 G (0.634) 14:53:05.17 +21:08:31.64 14:53:05.18 +21:08:31.50 J1458+2754 G 0.2288 14:59:00.23 +27:54:00.78 14:59:00.20 +27:54:00.10 J1510+0544 G (0.146) 15:10:56.10 +05:44:41.19 J1534+0556 G (0.257) 15:34:39.92 +05:56:17.45 J1547 0046 − J1706+2006 G (0.305) 17:06:16.33 +20:06:48.98 J2151 0005 G (0.810) 21:51:27.90 00:05:07.87 − − ⊕ In the second column G and S represent galaxy and star respectively, as listed in SDSS. 124 B.AppendixB:ListofthecandidateDDRGsandnon-DDRGs Bibliography

Akujor, C. E., Leahy, J. P., Garrington, S. T., Sanghera, H., Spencer, R. E., & Schilizzi, R. T. 1996, MNRAS, 278, 1

Alexander, P. 1987, MNRAS, 225, 27

Alexander, P. & Leahy, J. P. 1987, MNRAS, 225, 1

Antonucci, R. 1993, ARA&A, 31, 473

Axford, W. I., Leer, E., & McKenzie, J. F. 1982, A&A, 111, 317

Baan, W. A. & Haschick, A. D. 1981, ApJ Lett, 243, L143

Baars, J. W. M., Genzel, R., Pauliny-Toth, I. I. K., & Witzel, A. 1977, A&A, 61, 99

Becker, R. H., White, R. L., & Helfand, D. J. 1995, ApJ, 450, 559

Begelman, M. C., Blandford, R. D., & Rees, M. J. 1984, Reviews of Modern Physics, 56, 255

Bell, A. R. 1978a, MNRAS, 182, 147

—. 1978b, MNRAS, 182, 443

Beswick, R. J., Peck, A. B., Taylor, G. B., & Giovannini, G. 2004, MNRAS, 352, 49

Beswick, R. J., Pedlar, A., & Holloway, A. J. 2002, MNRAS, 329, 620

Bianchi, S., Maiolino, R., & Risaliti, G. 2012, Advances in Astronomy, 2012

Blandford, R. D. & Ostriker, J. P. 1978, ApJ Lett, 221, L29

Blandford, R. D. & Rees, M. J. 1974, MNRAS, 169, 395

125 126 BIBLIOGRAPHY

Blanton, E. L., Gregg, M. D., Helfand, D. J., Becker, R. H., & White, R. L. 2003, AJ, 125, 1635

Blundell, K. M. & Rawlings, S. 2000, AJ, 119, 1111

Blundell, K. M., Rawlings, S., & Willott, C. J. 1999, AJ, 117, 677

Bridle, A. H., Fomalont, E. B., & Cornwell, T. J. 1981, AJ, 86, 1294

Brocksopp, C., Kaiser, C. R., Schoenmakers, A. P., & de Bruyn, A. G. 2007, MNRAS, 382, 1019

—. 2011, MNRAS, 410, 484

Capetti, A., de Ruiter, H. R., Fanti, R., Morganti, R., Parma, P., & Ulrich, M.-H. 2000, A&A, 362, 871

Carilli, C. L., Perley, R. A., Dreher, J. W., & Leahy, J. P. 1991, ApJ, 383, 554

Chandola, Y., Saikia, D. J., & Gupta, N. 2010, MNRAS, 403, 269

Colina, L. & Perez-Fournon, I. 1990, ApJ, 349, 45

Condon, J. J., Cotton, W. D., Greisen, E. W., Yin, Q. F., Perley, R. A., Taylor, G. B., & Broderick, J. J. 1998, AJ, 115, 1693

Croston, J. H., Hardcastle, M. J., Harris, D. E., Belsole, E., Birkinshaw, M., & Worrall, D. M. 2005, ApJ, 626, 733

Czerny, B., Siemiginowska, A., Janiuk, A., Nikiel-Wroczynski,´ B., & Stawarz, Ł. 2009, ApJ, 698, 840

Dallacasa, D., Fanti, C., Fanti, R., Schilizzi, R. T., & Spencer, R. E. 1995, A&A, 295, 27 de Juan, L., Colina, L., & Perez-Fournon, I. 1994, ApJS, 91, 507 de Koﬀ, S., Best, P., Baum, S. A., Sparks, W., Röttgering, H., Miley, G., Golombek, D., Macchetto, F., & Martel, A. 2000, ApJS, 129, 33 de Vries, W. H., O’Dea, C. P., Baum, S. A., Perlman, E., Lehnert, M. D., & Barthel, P. D. 1998, ApJ, 503, 156

Drury, L. O. & Voelk, J. H. 1981, ApJ, 248, 344

Eilek, J. A. & Arendt, P. N. 1996, ApJ, 457, 150 BIBLIOGRAPHY 127

Emonts, B. H. C., Morganti, R., Tadhunter, C. N., Oosterloo, T. A., Holt, J., & van der Hulst, J. M. 2005, MNRAS, 362, 931

Erlund, M. C., Fabian, A. C., Blundell, K. M., Celotti, A., & Crawford, C. S. 2006, MNRAS, 371, 29

Evans, A. S., Mazzarella, J. M., Surace, J. A., Frayer, D. T., Iwasawa, K., & Sanders, D. B. 2005, ApJS, 159, 197

Evans, A. S., Sanders, D. B., Surace, J. A., & Mazzarella, J. M. 1999, ApJ, 511, 730

Fanaroﬀ, B. L. & Riley, J. M. 1974, MNRAS, 167, 31P

Fassnacht, C. D. & Taylor, G. B. 2001, AJ, 122, 1661

Floyd, D. J. E., Perlman, E., Leahy, J. P., Beswick, R. J., Jackson, N. J., Sparks, W. B., Axon, D. J., & O’Dea, C. P. 2006, ApJ, 639, 23

Fouque, P., Durand, N., Bottinelli, L., Gouguenheim, L., & Paturel, G. 1992, Catalogue of optical radial velocities, Observatoires de Lyon et Paris-Meudon

Giovannini, G., Taylor, G. B., Feretti, L., Cotton, W. D., Lara, L., & Venturi, T. 2005, ApJ, 618, 635

Gopal-Krishna, Biermann, P. L., Gergely, L. Á., & Wiita, P. J. 2012, Research in Astronomy and Astrophysics, 12, 127

Gopal-Krishna & Wiita, P. J. 2000, A&A, 363, 507

Gregorini, L., Padrielli, L., Parma, P., & Gilmore, G. 1988, A&AS, 74, 107

Hardcastle, M. J. 2009, in Astronomical Society of the Paciﬁc Conference Series, Vol. 407, The Low-Frequency Radio Universe, ed. D. J. Saikia, D. A. Green, Y. Gupta, & T. Venturi, 121

Hardcastle, M. J., Harris, D. E., Worrall, D. M., & Birkinshaw, M. 2004, ApJ, 612, 729

Haschick, A. D. & Baan, W. A. 1985, ApJ, 289, 574

Heavens, A. 1989, in Lecture Notes in Physics, Berlin Springer Verlag, Vol. 327, Hot Spots in Extragalactic Radio Sources, ed. K. Meisenheimer & H.-J. Roeser, 247–251

Heckman, T. M., Illingworth, G. D., Miley, G. K., & van Breugel, W. J. M. 1985, ApJ, 299, 41

Heckman, T. M., Smith, E. P., Baum, S. A., van Breugel, W. J. M., Miley, G. K., Illingworth, G. D., Bothun, G. D., & Balick, B. 1986, ApJ, 311, 526 128 BIBLIOGRAPHY

Hota, A., Sirothia, S. K., Ohyama, Y., Konar, C., Kim, S., Rey, S.-C., Saikia, D. J., Croston, J. H., & Matsushita, S. 2011, MNRAS, 417, L36

Ishwara-Chandra, C. H. & Saikia, D. J. 1999, MNRAS, 309, 100

Jaﬀe, W. J. & Perola, G. C. 1973, A&A, 26, 423

Jamrozy, M., Konar, C., Machalski, J., & Saikia, D. J. 2008, MNRAS, 385, 1286

Jamrozy, M., Konar, C., Saikia, D. J., Stawarz, Ł., Mack, K.-H., & Siemiginowska, A. 2007, MNRAS, 378, 581

Jamrozy, M., Saikia, D. J., & Konar, C. 2009, MNRAS, 399, L141

Jayaram N., C., Yashawant, G., & K.S., D., eds. 2003, Low Frequency Radio Astronomy (NCRA- TIFR)

Jennison, R. C. & Das Gupta, M. K. 1953, Nature, 172, 996

Jeyakumar, S. & Saikia, D. J. 2000, MNRAS, 311, 397

Jeyakumar, S., Wiita, P. J., Saikia, D. J., & Hooda, J. S. 2005, A&A, 432, 823

Jones, D. L. 1984, ApJ Lett, 276, L5

Jones, D. L. & Preston, R. A. 2001, AJ, 122, 2940

Jones, T. W., Ryu, D., & Engel, A. 1999, ApJ, 512, 105

Joshi, S. A., Nandi, S., Saikia, D. J., Ishwara-Chandra, C. H., & Konar, C. 2011, MNRAS, 414, 1397

Kaiser, C. R. & Alexander, P. 1999, MNRAS, 302, 515

Kaiser, C. R., Schoenmakers, A. P., & Röttgering, H. J. A. 2000, MNRAS, 315, 381

Kardashev, N. S. 1962, SvA, 6, 317

Kellermann, K. I., Sramek, R., Schmidt, M., Shaﬀer, D. B., & Green, R. 1989, AJ, 98, 1195

Kellermann, K. I., Vermeulen, R. C., Zensus, J. A., & Cohen, M. H. 1998, AJ, 115, 1295

Kirk, J. G. & Schneider, P. 1987, ApJ, 315, 425

Klein, U., Mack, K.-H., Gregorini, L., & Parma, P. 1995, A&A, 303, 427 BIBLIOGRAPHY 129

Knapp, G. R. & Patten, B. M. 1991, AJ, 101, 1609

Konar, C., Hardcastle, M. J., Croston, J. H., & Saikia, D. J. 2009, MNRAS, 400, 480

Konar, C., Jamrozy, M., Saikia, D. J., & Machalski, J. 2008, MNRAS, 383, 525

Konar, C., Saikia, D. J., Ishwara-Chandra, C. H., & Kulkarni, V. K. 2004, MNRAS, 355, 845

Konar, C., Saikia, D. J., Jamrozy, M., & Machalski, J. 2006, MNRAS, 372, 693

Körding, E. G., Jester, S., & Fender, R. 2006, MNRAS, 372, 1366

Kozieł-Wierzbowska, D., Jamrozy, M., Zola, S., Stachowski, G., & Ku´zmicz, A. 2012, MNRAS, 422, 1546

Kuehr, H., Witzel, A., Pauliny-Toth, I. I. K., & Nauber, U. 1981, A&AS, 45, 367

Laing, R. A. & Peacock, J. A. 1980, MNRAS, 190, 903

Lara, L., Márquez, I., Cotton, W. D., Feretti, L., Giovannini, G., Marcaide, J. M., & Venturi, T. 1999, A&A, 348, 699

Leahy, J. P., Muxlow, T. W. B., & Stephens, P. W. 1989, MNRAS, 239, 401

Leahy, J. P., Sparks, W. B., & Jackson, N. J. F. 1999, in Bulletin of the American Astronomical Society, Vol. 31, American Astronomical Society Meeting Abstracts #194, 950

Ledlow, M. J. & Owen, F. N. 1996, AJ, 112, 9

Liu, F. K., Wu, X.-B., & Cao, S. L. 2003, MNRAS, 340, 411

Liu, R., Pooley, G., & Riley, J. M. 1992, MNRAS, 257, 545

Machalski, J., Chy˙zy, K. T., Stawarz, Ł., & Kozieł, D. 2007, A&A, 462, 43

Machalski, J., Jamrozy, M., & Konar, C. 2010, A&A, 510, A84

Machalski, J., Jamrozy, M., & Saikia, D. J. 2009, MNRAS, 395, 812

Machalski, J., Kozieł-Wierzbowska, D., Jamrozy, M., & Saikia, D. J. 2008, ApJ, 679, 149

Mack, K.-H., Klein, U., O’Dea, C. P., Willis, A. G., & Saripalli, L. 1998, A&A, 329, 431

Mao, M. Y., Sharp, R., Saikia, D. J., Norris, R. P., Johnston-Hollitt, M., Middelberg, E., & Lovell, J. E. J. 2010, MNRAS, 406, 2578 130 BIBLIOGRAPHY

Marconi, A., Risaliti, G., Gilli, R., Hunt, L. K., Maiolino, R., & Salvati, M. 2004, MNRAS, 351, 169

Marecki, A. 2012, A&A, 544, L2

Martel, A. R., Baum, S. A., Sparks, W. B., Wyckoﬀ, E., Biretta, J. A., Golombek, D., Macchetto, F. D., de Koﬀ, S., McCarthy, P. J., & Miley, G. K. 1999, ApJS, 122, 81

Morganti, R., Oosterloo, T. A., Emonts, B. H. C., van der Hulst, J. M., & Tadhunter, C. N. 2003, ApJ Lett, 593, L69

Murgia, M., Fanti, C., Fanti, R., Gregorini, L., Klein, U., Mack, K.-H., & Vigotti, M. 1999, A&A, 345, 769

Myers, S. T. & Spangler, S. R. 1985, ApJ, 291, 52

Nandi, S., Pirya, A., Pal, S., Konar, C., Saikia, D. J., & Singh, M. 2010, MNRAS, 404, 433

Nandi, S. & Saikia, D. J. 2012, Bulletin of the Astronomical Society of India, 40, 121

Nipoti, C., Blundell, K. M., & Binney, J. 2005, MNRAS, 361, 633

O’Dea, C. P., Koekemoer, A. M., Baum, S. A., Sparks, W. B., Martel, A. R., Allen, M. G., Macchetto, F. D., & Miley, G. K. 2001, AJ, 121, 1915

Owen, F. N. & Ledlow, M. J. 1997, ApJS, 108, 41

Pacholczyk, A. G. 1970, Radio astrophysics. Nonthermal processes in galactic and extragalactic sources

Parma, P., de Ruiter, H. R., Mack, K.-H., van Breugel, W., Dey, A., Fanti, R., & Klein, U. 1996, A&A, 311, 49

Parma, P., Murgia, M., Morganti, R., Capetti, A., de Ruiter, H. R., & Fanti, R. 1999, A&A, 344, 7

Perley, R. A., Schwab, F. R., & Bridle, A. H., eds. 1989, Astronomical Society of the Paciﬁc Conference Series, Vol. 6, Synthesis imaging in radio astronomy

Pirya, A., Nandi, S., Saikia, D. J., & Singh, M. 2011, Bulletin of the Astronomical Society of India, 39, 547

Proctor, D. D. 2011, ApJS, 194, 31

Ramos Almeida, C., Tadhunter, C. N., Inskip, K. J., Morganti, R., Holt, J., & Dicken, D. 2011, MNRAS, 410, 1550 BIBLIOGRAPHY 131

Reynolds, C. S. & Begelman, M. C. 1997, ApJ Lett, 487, L135

Rudnick, L., Katz-Stone, D. M., & Anderson, M. C. 1994, ApJS, 90, 955

Safouris, V., Subrahmanyan, R., Bicknell, G. V., & Saripalli, L. 2008, MNRAS, 385, 2117

Saikia, D. J., Gupta, N., & Konar, C. 2007, MNRAS, 375, L31

Saikia, D. J. & Jamrozy, M. 2009, Bulletin of the Astronomical Society of India, 37, 63

Saikia, D. J., Jeyakumar, S., Wiita, P. J., Sanghera, H. S., & Spencer, R. E. 1995, MNRAS, 276, 1215

Saikia, D. J., Konar, C., & Kulkarni, V. K. 2006, MNRAS, 366, 1391

Saikia, D. J. & Kulkarni, V. K. 1994, MNRAS, 270, 897

Salter, C. J., Saikia, D. J., Minchin, R., Ghosh, T., & Chandola, Y. 2010, ApJ Lett, 715, L117

Saripalli, L. & Mack, K.-H. 2007, MNRAS, 376, 1385

Saripalli, L., Subrahmanyan, R., & Udaya Shankar, N. 2002, ApJ, 565, 256

Scheuer, P. A. G. 1974, MNRAS, 166, 513

Schoenmakers, A. P. 1999, PhD thesis, Utreht University

Schoenmakers, A. P., de Bruyn, A. G., Röttgering, H. J. A., & van der Laan, H. 2000a, MNRAS, 315, 395

—. 2001, A&A, 374, 861

Schoenmakers, A. P., de Bruyn, A. G., Röttgering, H. J. A., van der Laan, H., & Kaiser, C. R. 2000b, MNRAS, 315, 371

Schoenmakers, A. P., Mack, K.-H., de Bruyn, A. G., Röttgering, H. J. A., Klein, U., & van der Laan, H. 2000c, A&AS, 146, 293

Shostak, G. S., van Gorkom, J. H., Ekers, R. D., Sanders, R. H., Goss, W. M., & Cornwell, T. J. 1983, A&A, 119, L3

Singal, A. K., Konar, C., & Saikia, D. J. 2004, MNRAS, 347, L79

Sirothia, S. K., Dennefeld, M., Saikia, D. J., Dole, H., Ricquebourg, F., & Roland, J. 2009a, MNRAS, 395, 269 132 BIBLIOGRAPHY

Sirothia, S. K., Saikia, D. J., Ishwara-Chandra, C. H., & Kantharia, N. G. 2009b, MNRAS, 392, 1403

Smith, H. E. & Spinrad, H. 1980, PASP, 92, 553

Spergel, D. N., Verde, L., Peiris, H. V., Komatsu, E., Nolta, M. R., Bennett, C. L., Halpern, M., Hinshaw, G., Jarosik, N., Kogut, A., Limon, M., Meyer, S. S., Page, L., Tucker, G. S., Weiland, J. L., Wollack, E., & Wright, E. L. 2003, ApJS, 148, 175

Steenbrugge, K. C., Blundell, K. M., & Duﬀy, P. 2008, MNRAS, 388, 1465

Steenbrugge, K. C., Heywood, I., & Blundell, K. M. 2010, MNRAS, 401, 67

Strom, R. G., Fanti, R., Parma, P., & Ekers, R. D. 1983, A&A, 122, 305

Subrahmanyan, R. & Saripalli, L. 1993, MNRAS, 260, 908

Subrahmanyan, R., Saripalli, L., & Hunstead, R. W. 1996, MNRAS, 279, 257

Swarup, G., Ananthakrishnan, S., Kapahi, V. K., Rao, A. P., Subrahmanya, C. R., & Kulkarni, V. K. 1991, Current Science, Vol. 60, NO.2/JAN25, P. 95, 1991, 60, 95

Thompson, A. R., Clark, B. G., Wade, C. M., & Napier, P. J. 1980, ApJS, 44, 151

Thompson, A. R., Moran, J. M., & Swenson, G. W. 1986, Interferometry and synthesis in radio astronomy, ed. Thompson, A. R., Moran, J. M., & Swenson, G. W.

Urry, C. M. & Padovani, P. 1995, PASP, 107, 803 van Breugel, W., Heckman, T., Butcher, H., & Miley, G. 1984, ApJ, 277, 82

Vigotti, M., Grueﬀ, G., Perley, R., Clark, B. G., & Bridle, A. H. 1989, AJ, 98, 419

Wiita, P. J. & Gopal-Krishna. 1990, ApJ, 353, 476