INSTITUTE OF NATURAL AND APPLIED SCIENCES ...period of this study. First of all I would like to...

INSTITUTE OF NATURAL AND APPLIED SCIENCES UNIVERSITY OF CUKUROVA

PhD THESIS

Eda SONBAŞ

INTERPRETATION OF PHOTOMETRIC AND SPECTRAL OBSERVATIONS OF GAMMA RAY BURST AFTERGLOWS

DEPARTMENT OF PHYSICS

ADANA, 2009

ÇUKUROVA ÜNİVERSİTESİ

FEN BİLİMLERİ ENSTİTÜSÜ

GAMA IŞIN PATLAMALARI ARDIL IŞINIMLARININ FOTOMETRİK VE TAYF GÖZLEMLERİNİN YORUMU

EDA SONBAŞ

DOKTORA TEZİ

FİZİK ANABİLİM DALI

Bu tez ................................ Tarihinde Aşağıdaki Jüri Üyeleri Tarafından Oybirliği ile Kabul Edilmiştir. İmza:.............................. İmza:........................... İmza:............................ Prof.Dr. Aysun AKYÜZ Prof.Dr.Mehmet Emin ÖZEL Prof. Dr.Gülsen ÖNENGÜT DANIŞMAN ÜYE ÜYE İmza:.............................. İmza:.............................. Doç. Dr. H. Mustafa KANDIRMAZ Yard. Doç. Dr. Muhittin ŞAHAN ÜYE ÜYE Bu tez Enstitümüz Fizik Anabilim Dalında hazırlanmıştır. Kod No: Prof. Dr. Aziz ERTUNÇ Enstitü Müdürü İmza ve Mühür Bu Çalışma Ç.Ü. Araştırma Fonu Tarafından Desteklenmiştir. Proje No: FEF 2008 D2 Not: Bu tezde kullanılan özgün ve başka kaynaktan yapılan bildirişlerin, çizelge, şekil ve fotografların kaynak gösterilmeden kullanımı, 5846 sayılı Fikir ve Sanat Eserleri Kanunundaki hükümlere tabidir.

Sevgili aileme,

I

ABSTRACT

Ph.D. THESIS

INTERPRETATION OF PHOTOMETRIC AND SPECTRAL OBSERVATIONS OF GAMMA RAY BURST AFTERGLOWS

Eda SONBAŞ

DEPARTMENT OF PHYSICS

INSTITUTE OF NATURAL AND APPLIED SCIENCES

UNIVERSITY OF CUKUROVA

Supervisor: Prof. Dr Aysun AKYÜZ

Year:2009, Page: 119

Jury: Prof. Dr. Aysun AKYÜZ

Prof. Dr. M. Emin ÖZEL

Prof. Dr. Gülsen ÖNENGÜT

Assoc. Prof.. Dr. H. Mustafa KANDIRMAZ

Assist. Prof. Dr. Muhittin ŞAHAN

The short (~0.1-100 seconds) and intense bursts of high energy gamma-ray

radiation (>100 keV) are called Gamma-Ray Bursts (GRBs). They occur uniformly

in the universe approximately once per day at very high distances to Earth. Their

nature and origin have remained a mystery of modern astronomy since GRB were

discovered serendipitously in 1967. The results of afterglow observations with

ground based telescopes and scientific observations sent after 1990 indicated that

GRBs are cosmological objects. In this work, the association of GRBs – SNe

(Supernovae) are studied with the optical spectra of afterglow of some selected

GRBs by the 6m-BTA (Bolsoi Azimuthal Telescope) telescope at SAO-RAS

(Special Astrophysical Observatory of Russian Academy of Sciences). The host

galaxy properties and afterglows are also studied with the photometrical observations

of some selected GRBs.

Key words: GRB: Gamma Ray Burst, afterglow, SN (Supernova),

spectroscopy, photometry

I I

ÖZ

DOKTORA TEZİ

GAMA IŞIN PATLAMALARI ARDIL IŞINIMLARININ FOTOMETRİK

VE TAYF GÖZLEMLERİNİN YORUMU

Eda SONBAŞ

ÇUKUROVA ÜNİVERSİTESİ

FEN BİLİMLERİ ENSTİTÜSÜ

FİZİK ANABİLİMDALI

Danışman: Prof. Dr. Aysun AKYÜZ

Yıl:2009, Sayfa: 119

Jüri: Prof. Dr. Aysun AKYÜZ

Prof. Dr. M. Emin ÖZEL

Prof. Dr. Gülsen ÖNENGÜT

Doç. Dr. H. Mustafa KANDIRMAZ

Yard. Doç. Dr. Muhittin ŞAHAN

Gama Işın Patlamaları (GIP), >100 keV enerjilerde gama ışınları yayan kısa ( ~

0.1–100 sn) ve şiddetli patlamalardır. GIP dünyanın çok uzağında yaklaşık günde bir

kez olur ve düzgün dağılım gösterirler. Tesadüfen keşfedildiği tarihten bu yana

GIP’nın kaynağının ve doğasının tam olarak anlaşılamaması, onları modern

astronominin en gizemli konusu yapmaktadır. Özellikle 1990 sonrasında gönderilen

uydular tarafından yapılan gözlemler ve yerden yapılan ardıl ışınım gözlemleriyle

GIP’nın kozmolojik kaynaklar olduğu anlaşılmıştır. Bu çalışmada, öncelikle GIP’nın

genel özellikleri derlenmiştir. GIP – SN (Süpernova) ilişkisi ardıl ışnımlarının 6m' lik

BTA (Bolsoi Azimuthal Telescope) ile yapılan tayfsal gözlemleriyle çalışılmıştır.

Ayrıca bazı GIP ardıl ışınımları ve evsahibi galaksilerin özellikleri fotometrik

gözlemlerle çalışılmıştır.

Anahtar Kelimeler: GIP: Gama Işın Patlamaları, Ardıl Işınım, SN:

Süpernova, Tayf, Fotometri

III

ACKNOWLEDGEMENTS

I would like to thanks many people for their help and support during the

period of this study.

First of all I would like to thank my advisor, Prof. Dr. Aysun Akyüz, for her

vulnerable direction and motivation keeping me on right ways both my graduate

studies and life, for her patience, and supporting me when I fell least hopeful in this

long period. I would not have been imagined to finish my thesis without her.

I would also like to thank to Prof. Vladimir V. Sokolov for the supervising

and directing my research in Special Astrophysical Observatory of Russian Academy

of Science (SAO - RAS). His helps, insights, and directions are very important and

crucial for me to complete this thesis. Further I would also thank to him to say me

every time ‘So, cheer up! We will overcome!’.

I would like to thank the director of SAO-RAS, Yuri Yu. Balega and his wife

for their support and hot behavior.

I would like to thank the members of the SAO-RAS, including researchers

Timur Fatkhullin, Tatyana Sokolova, Alexander Moskvitin, Dr. Viktorya Kamorava,

and secretary Ekaterina Filippova

I would like to thank to A.J. Castro – Tirado and Antonio Ugarte Postigo

from Spain for their direction on observations of afterglow observations of Gamma

Ray Bursts.

I would like to thank Prof. Dr. M. Emin Özel for all his support with his

constructive suggestion during my graduate studies.

I would also thank to my colleges and friends İlham Nasiroğlu, Şükriye

Cihangir, and the staff of Space Science and Solar Energy Research and Application

Center (UZAYMER), Hakki Görgülü for their help during my works in UZAYMER.

Finally I would especially like to thank my sister Buket Sonbaş, my father

Hamza Sonbaş, and my mother Aysel Sonbaş for their encouragement, support and

love, for sharing my excitements, happiness and difficult times.

The works of this dissertation is based on the papers in part or in full ‘Sonbas,

E. et al. 2008. Stellar Wind Envelope Around The Massive Supernova Progenitor

IV

XRF/GRB 060218/SN 2006aj. Astrophysical Bulletin, Volume 63, Issue 3, pp.228-

243’, ‘Castro-Tirado, A. J. et al. 2008. Flares from a candidate Galactic magnetar

suggest a missing link to dim isolated neutron stars. Nature, Volume 455, Issue 7212,

pp. 506-509’, and ‘Moskvitin et al. 2008. Gamma-ray bursts and practical

cosmology. 2008pc2..conf..228M’

V

TABLE OF CONTENTS SAYFA

ABSTRACT............................................................. .......................................... I

ÖZ............................................................................ .......................................... II

ACKNOWLEDGEMENTS..................................... .......................................... III

TABLE OF CONTENTS......................................... ......................................... V

LIST OF TABLES.................................................... ......................................... VII

LIST OF FIGURES.................................................. .......................................... VIII

1.INTRODUCTION................................................. ........................................... 1

2. PREVIOUS WORK.................................. ....................................................... 4

3. BRIEF OVERVİEW AND METHOD… ....................................................... 15

3.1. Theory of Gamma Ray Bursts.................................................................. 15

3.2. Progenitors of GRBs……....................................................................... 17

3.2.1. Collapsar / Hypernova Models...................................................... 20

3.2.2. The Supranova Model……………............................................... 21

3.2.3. Supernova Models of GRBs.......................................................... 22

3.2.4. Observational Evidence................................................................ 24

3.2.5. Role of Binary Star Systems in GRB Events................................ 27

3.2.6. The Fireball Model........................................................................ 28

3.2.7. GRB Afterglows........................................................................... 31

3.3. Spectroscopic Observations and Analysis Technique for the GRB

Afterglows...............................................................................................

35

3.4. The Code 'SYNOW' for Spectral Interpretation..................................... 38

3.5. PEGASE................................................................................................. 42

4. INTERPRETATION OF OBSERVATIONS AND RESULTS.................... 43

4.1. Observations of The Stellar – Wind Envelope Around The Massive

Supernova Progenitor XRF / GRB 060218 / SN 2006aj.......................

43

4.1.1. Review of Earlier Observations.................................................. 43

4.1.2. The Energies of the X – Ray and UV Flashes............................. 49

4.1.3. Evolution of Temperature........................................................... 49

4.1.4. The Size of The Stellar – Wind Envelope................................... 50

4.1.5. Shock Velocity............................................................................ 52

VI

4.1.6. Early Spectroscopic Observations of the Optical Afterglow of

XRF 060218 / SN 2006aj Made With the 6 – m Telescope.........

53

4.1.7. Comparison of the Observed Spectra of XRF 060218 / SN

2006aj With Synthetic Spectra.....................................................

54

4.1.8. Results and Discussion............................................................... 61

4.1.8.1. Evolution of the Spectra of XRF 060218 / SN 2006aj

and Other Massive SNe…...............................................

61

4.1.8.2. Hydrogen Signatures in The Spectra of Massive SNe.... 66

4.1.8.3. The Colgate “Shock – Breakout” Effect in XRF 060218

/ SN 2006aj and Other SNe – the Light Curves,

Spectra, Luminosities, and Sizes.....................................

67

4.1.8.4. Asymmetry of Type Ib and Ic Supernova Explosions… 70

4.2. SYNOW Modeling of XRF 080109 / SN 2008D................................ 71

4.2.1. Velocities of Photosphere and Envelopes for both SN 2006aj

and SN 2008D............................................................................

77

4.3. Modeling Wide-Band Spectra of GRB 021004 and GRB 060218

Host Galaxies.......................................................................................

79

4.3.1. The Host Galaxy of GRB 021004.............................................. 80

4.3.2. The Host Galaxy of GRB 060218.............................................. 82

4.3.3. Modeling of Spectra................................................................... 83

4.3.4. Results of Modeling................................................................... 86

4.4. Observations of SWIFT J195509+261406 in The Field of GRB

070610 with 6-m BTA Telescope........................................................

91

5. CONCLUSION..……………………….............. ............……………........ 98

REFERENCES………… …………………………………………………… 102

CIRRICULUM VITAE... …………………………………………………… 113

APPENDIX – 1.. …………………………………………………………… 114

VII

LIST OF TABLES

PAGE

Table 2.1 Spectroscopically confirmed GRBs and SNe connections..... 14

Table 4.1 Log of early spectra of the event XRF 060218 / SN 2006aj... 47

Table 4.2 The model parameters set of SYNOW computation of 2.55 d spectrum of XRF 060218 / SN 2006aj................................

59

Table 4.3 The model parameters set of SYNOW computation of 3.55 d spectrum of XRF 060218 / SN 2006aj................................

61

Table 4.4 Parameters for the spectrum of SN 2008D obtained 6.48 days after explosion…………………………………………

77

Table 4.5 Parameters for the spectrum of SN 2008D obtained 27.62 days after explosion…………………………………………

77

Table 4.6 Expansion velocities of photospheres (vphot), hydrogen (VH), and helium (VHe) envelopes of SN 2006aj and SN 2008D………………………………………………………..

78 Table 4.7 Photometry of the GRB 021004 host galaxy……………….. 82

Table 4.8 Photometry of the GRB 060218 host galaxy……………….. 83

Table 4.9 Model parameters of the GRB 021004 host galaxy corresponding to minimum of χ2 for the metallicity Z = Zּס...

90

Table 4.10 The model parameters of the GRB 021004 host galaxy corresponding to minimum of χ2 for the metallicity Z = 0.1 Zּס……………………………………………………………

90 Table 4.11 The model parameters of the GRB 060218 host galaxy

corresponding to minimum of χ2 for the metallicity Z = Zּס...

91 Table 4.12 The model parameters of the GRB 060218 host galaxy

corresponding to minimum of χ2 for the metallicity Z = 0.1 Zּס……………………………………………………………

91

Table 4.13 6m – BTA observations of GRB 070610 field……………... 96

VIII

LIST OF FIGURES

PAGE

Figure 3.1. Observations of GRB OTs and their host galaxies............................... 17

Figure 3.2. Schematic representation of the progenitor models of GRBs............... 19

Figure 3.3. Classification of supernovae................................................................. 24

Figure 3.4. Asymetry of supernovae....................................................................... 26

Figure 3.5. Schematic representation of fireball model.......................................... 29

Figure 3.6. Radio afterglow light curve of GRB 970508........................................ 34

Figure 3.7. The sample of a flat image...……………............................................. 36

Figure 3.8. Raw data of bias................................................................................ 36

Figure 3.9. Unprocessed arc image......................................................................... 37

Figure 3.10. Master flat and master bias subtracted image of the object.................. 38

Figure 3.11. The image of object with subtracted pair.............................................. 38

Figure 3.12. The resulting summed spectrum of XRF 080109................................. 39

Figure 3.13. The “detached” and “undetached” cases in the SYNOW code............. 40

Figure 3.14. The format of “in.dat” file.................................................................... 42

Figure 4.1. The early afterglow light curve of Swift............................................... 44

Figure 4.2. X – ray and UVOT (Ultra Violet/Optical Telescope) light curves of

GRB060218..........................................................................................

45

Figure 4.3. Observed spectra of the afterglow of XRF 060218 / SN 2006aj taken

with the 6-m telescope..........................................................................

46

Figure 4.4. Optical (U, B, V, R, I) and Infrared (J) light curves of XRF 060218 /

SN 2006aj............................................................................................

48

Figure 4.5. The temperature and radius evolution of thermal soft component of

GRB 060218.........................................................................................

52

Figure 4.6. Field of the optical transient................................................................. 55

Figure 4.7. SYNOW modeling of the 2.55 days spectrum of GRB 060218............ 60

Figure 4.8. SYNOW modeling of the 3.55 days spectrum of GRB 060218............ 61

Figure 4.9. The flux-calibrated and de-reddened spectra of SN 2006aj................. 63

Figure 4.10. VLT and LICK spectra ant its synthetic fit with the Monte-Carlo

IX

spectrum synthesis code of SN 2006aj................................................. 65

Figure 4.11. The earliest optical spectrum of SN 1993J........................................... 70

Figure 4.12. The lightcurve of SN 2008D................................................................. 73

Figure 4.13. Observations of SN 2008D with BTA / SCORPIO on Jan. 16.05,

2008......................................................................................................

74

Figure 4.14. Observations of SN 2008D with BTA / SCORPIO on Feb. 6.18,

2008......................................................................................................

74

Figure 4.15. BTA / SCORPIO spectra of SN 2008D................................................ 75

Figure 4.16. SYNOW modeling of SN 2008D obtained on Jan. 16......................... 76

Figure 4.17. SYNOW modeling of SN 2008D obtained on Feb. 6........................... 77

Figure 4.18. Photospheric velocities......................................................................... 79

Figure 4.19. Minimum velocities of ions.................................................................. 80

Figure 4.20. Curve of extinction laws used in the modeling..................................... 88

Figure 4.21. Result of modeling of GRB 021004...................................................... 89

Figure 4.22. Results of modelling GRB 021004 at Z = 0.1 Zּ89 .................................ס

Figure 4.23. Results of modeling GRB 060218 at Z = Zּ90 ........................................ס

Figure 4.24. Results of modelling GRB 060218 at Z = 0.1 Zּ90 .................................ס

Figure 4.25. 94

Figure 4.26. Light curve of SWIFT J195509+261406 95

Figure 4.27. The magnitude distribution of the optical flares detected from

SWIFTJ195509+261406 in the Ic band................................................

96

1.INTRODUCTION Eda SONBAŞ

1

1. INTRODUCTION

Gamma Ray Bursts (GRBs) are the most luminous events with 1051 – 1052 ergs

energy release in the Universe after the Big Bang. These bursts can be described

highly dense ‘gamma-ray flashes’ in the range of a few seconds to several minutes

duration.

GRBs have represented one of the outstanding scientific puzzles since their

discovery in 1967 (Klebesadel, Strong & Olson, 1973). The discovery of GRB

afterglows in X – rays (Costa et al. 1997), visible (van Paradijs et al. 1997), and radio

(Frail et al. 1997) has caused a considerable progress in their understanding. Today

GRBs studies are one of the most active and vibrant areas of astrophysics

(Djorgovski et al. 2003).

The most important development have been made with the observations of

BATSE (Burst And Transient Source Experiment) which was the detector on CGRO

(Compton Gamma Ray Observatory) launched in 1991. Almost 3000 bursts have

been detected by BATSE with scanned the whole sky during 9 years of operation.

According to BATSE results GRBs have been divided into two subgroups as “long

bursts” and “short bursts” according to their duration. Generally, long duration bursts

last more than two seconds and short duration bursts last less than two seconds. The

isotropic distribution of these bursts supported the idea of their cosmological

distances.

Since the Italian – Dutch satellite BeppoSAX has started its mission in the

beginning of 1997, a number of bursts have been localized less than an arc minute

certainity in radius. The burst coordinates are distributed to the community within 4-

8 hours. Thus observers can investigate GRBs in other wavelengths than the

observation of BeppoSAX. Therefore, the distance scale phenomenon has been

solved with the host galaxy observation of GRB 970228 which is the first GRB

detected in optical waveband. As the result of these observations the information

about the cosmologic origin has been improved.

The GRB satellite SWIFT is also capable to make X – ray, UV, and optical

observations of the sources. SWIFT has detected too many bursts since its launch on


2

November 20, 2004. In the case of possible GRB, SWIFT slew into the burst

direction while scanning the whole sky. Thus the position of events is quickly sent to

the ground to allow observation by other observatories. The first short duration GRB

(GRB 050509b) location, the furthest (z = 6.29) GRB (GRB 050904), and the

brightest (with 5th magnitude) GRB (GRB 080319B) ever seen have been identified

with SWIFT observations.

The multiwavelength emission satisfies the predictions of “standard”

relativistic fireball model (Meszaros & Rees 1993). And the central engines that

power this extraordinary events are thought to be the core collapse of massive stars

as recently proven by the spectroscopic signature of a peculiar, very energetic SN

(SN 2003dh) in the nearest “classical” GRB afterglow detected so far, at z = 0.1685

(Hjorth et al. 2003).

Scientists have been working on the different possible models about the

source of GRBs since 1970s. In recent years the most acceptable models contain

either the birth signal of a black hole with the death of massive stars or production of

the collision of two neutron stars or other exotic events as the causes these strong

radiation. Basically, these models are explored as the source of this tremendous

energy of explosion.

The physics of GRB has been developed with their afterglow observations.

The proposed progenitor and central engine models can be tested via afterglows of

these events. In addition to these information the distance scales and host galaxy

properties are also investigated. Up to now a number of GRB afterglows have been

studied with multiwavelength observations. With GRB 980425 observations it is

concluded that this event was a normal GRB produced with a supernova (SN, SN

1998bw). Therefore the progenitor of this event can be found. The distance

measurements have been established with the optical observation of GRB 970228

afterglow and host galaxy. Another important afterglow observation include the

observation of GRB 030329. The light curve of this burst showed a complex

evolution with bump and breaks. The appearance of the supernova, SN 2003 dh in

the spectrum of GRB 030329 was the first direct spectroscopic evidence that some


3

GRBs may originate from core-collapse supernova. This idea was also confirmed

with the observation of another GRB, GRB 060218.

2. PREVIOUS WORK Eda SONBAŞ

4

2. PREVIOUS WORK

Gamma Ray Bursts (GRBs) are short (~ 0.1 – 100 seconds) and intense bursts

of > 100 keV gamma – rays. GRBs were first detected serendipitously on July 2,

1967 by VELA which was bound to United States Ministry of Defense. Main

purposes of the Vela satellites were monitoring the nuclear test ban treaty of Soviet

Union. These mysterious flashes were thought to come from Earth's direction when

they were first detected. However Klebesadel et al. (1973) reported that these

mysterious flashes are new and highly puzzling cosmic phenomena. These results

were then confirmed with the observations of Soviet Konus experiment and IMP-6

(Interplenetary Monitoring Platform).

It is generally known that GRBs release ~1051 – 1053 ergs or more energy in a

few seconds. This huge energy which is 1019 times more than the luminosity of sun

comes from about the volume of sun. These intense and prompt explosions suggest

e± formation and γ fireball (Cavallo & Rees 1978, Meszaros 2002). According to

fireball shock model, external and internal shocks which go forward in the reverse

direction are created by the fireball. Fireball model also explains the afterglow of

GRBs (Meszaros 2003, Gehrels et al. 2002, Piran 1998, 2001, Goodman 1986,

Paczynsky 1986, Paczynisky and Rhoads 1993, Katz 1994, Meszaros & Rees 1997,

Vietri 1997, Zhang & Meszaros 2003).

Important developments have been achieved in GRB physics with the

discovery of their afterglows. These afterglow observations satisfy the predictions of

the “standard” relativistic fireball model (Meszaros & Rees 1997). Afterglow is the

radiation after the explosion in gamma-rays. Afterglow emission can be observable

from minutes to weeks after the GRB event in X – rays, optical, UV, and radio

wavelengths. The important features of GRBs can be derived by afterglow emissions.

They also give information about the redshift of GRBs. The calculated values of

redshift provide distance and locations of GRBs. The measurements of some of the

GRB redshifts indicate that these events are in cosmological distances (Meszaros

1999, 2003, Waxman 2000, Piran 2001, Ghisellini 2001, Rykoff et al. 2004, Yost et

al. 2003).


5

Their nature, location and origin could not be explained exactly for almost four

decades. In this period these bursts were observed at gamma rays (Hurley 1992), and

occasionally X – ray wavelengths (Murakami et al. 1988, Yoshida et al. 1989,

Connors & Hueter 1998). Most important developments in GRB research were

recorded with the launch of Compton Gamma – Ray Observatory (CGRO) in 1991.

Most of the bursts were recorded by BATSE (Burst and Transient Source

Experiment) detector on CGRO. BATSE had been designed to study the mystery of

these events and its observations have opened new horizons for understanding the

distribution and origin of GRBs.

In 1970s, contributions to the GRB events have been made by IMP – 6, OSO –

7 (Orbiting Solar Observatory) and SAS – 2 (Small Astronomy Satellite) satellites.

SAS – 2 satellite was launched on 1972 November 15 and has been designed to

detect > 35 MeV gamma – rays. It started to detect GRB events four days after its

launch. SAS – 2 satellite detected 27 GRBs during the 6 months of its orbiting

period. Eventually, 55% part of sky that covers the bulk of the galactic plane was

observed (Fichtel et al. 1975, M.E. Özel 1982).

The statistics and recording efficiency of GRB events reached significant

numbers after 1977. Pioneer (Evans et al. 1979), Sun – Earth Explorer – 3

(Kouveliotou et al. 1981) and Venera 11 and 12 (Mazets et al. 1981) satellites have

started their missions besides Helios – 2, and HEAO – 1 (High Energy Astronomy

Observatories) satellites. Besides these satellites, gamma – ray observations have

been made successfully with balloon experiments (Sommer et al. 1978). Thus, the

recorded number of explosions increased to 130 in 1980. The number and sensitivity

of satellites, that scan the entire sky continuously, increased and the important

developments in GRB astronomy has been achieved.

CGRO, was the second great observatory of NASA. It was launched on April

5, 1991. CGRO, has made a very big contribution to GRB astronomy until it re-

entered the Earth' s atmosphere on June 4, 2000. CGRO had four detectors in the 30

keV – 30 GeV energy range of electromagnetic spectrum. These detectors were as

follows;


6

BATSE, Burst and Transient Source Experiment searched the sky within

20 – 103 keV range. BATSE was a sky monitor on CGRO. Eight BATSE detectors

were placed on the corners of CGRO. The main aim of these detectors was to study

the phenomenon of GRBs. Moreover, BATSE also recorded data from pulsars,

terrestrial gamma – ray flashes, soft gamma repeaters, black holes and other exotic

astrophysical objects. The origin of GRBs has been discussed extensively after

BATSE moved into the orbit. 2700 GRBs were recorded during its mission and the

sky distribution map of GRBs seen by BATSE was created.

OSSE, Oriented Scintillation Spectrometer Experiment has been designed

to obtain overall observations of astrophysical sources in the energy range of 0.05 –

10 MeV. It was also designed to make gamma – ray and neutron observations for

above 10 MeV solar flare studies (http://cossc.gsfc.nasa.gov/osse). OSSE was

capable to make hard X – ray and soft gamma – ray observations of GRB sources.

Rapid response and high energy sensitivity of OSSE could contribute restrictions to

illuminate burst/afterglow connection and afterglow mechanism. OSSE has worked

burst and afterglow data of five brightest events that were well observed with OSSE

in pointed observations. Four of these bursts have > 50 keV radiation after the main

event. The spectral and temporal specifications of afterglows detected by OSSE were

defined and these results were compared with other low energetic afterglow

observations (Matz et al. 1999).

COMPTEL, The Imaging Compton Telescope viewed the sky in 8 – 30

MeV range. The spectra and location catalog of GRBs localized by COMPTEL was

created during the mission of CGRO (http://wwwgro.unh.edu/bursts/cgrbpage.html).

Spatial analysis of bursts localized by COMPTEL was investigated whether these

GRB locations are consistent with an isotropic angular distribution of sources. The

long lasting (162s) gamma-ray burst, GRB 940217, was detected by COMPTEL

(Winkler et al. 1995). The spectral observation of COMPTEL is consistent with

EGRET and BATSE observations. Moreover, COMPTEL recorded 29 GRBs during

the period from April, 1991 to May, 1995 (Kippen et al. 1998).

http://cossc.gsfc.nasa.gov/osse

http://wwwgro.unh.edu/bursts/cgrbpage.html


7

EGRET, Energetic Gamma Ray Experiment has gamma-ray energy range

from about 30 MeV to 30 GeV. EGRET provided a detailed highest–energy gamma–

ray window for CGRO. It was 10 to 20 times larger and more sensitive than other

high energetic detectors sent before. Diffuse gamma-ray emission, gamma-ray

bursts, cosmic rays, pulsars, and active galaxies (blazars) have been studied with the

detailed observations by EGRET. One of the long – lasting GRB detected by EGRET

is GRB 940217. This high energy explosion emission has continued in the hard X –

ray during ~ 5000 s after the explosion (Hurley at al. 1994). EGRET also discovered

several long duration GRBs (Dingus et al. 1997).

The results from EGRET showed that the energies of GRBs can be higher than

1 GeV and brought new approaches to these explosions. Therefore GRBs show

bimodal distribution with two time scales of short bursts of t < 2s and long bursts of

t > 2s GRB durations (Kouveliotou et al. 1993).

The Italian – Dutch Satellite BeppoSAX has obtained high a resolution X – ray

image of GRB 970228 afterglow in 1997 (Costa et al. 1997). BeppoSAX was

launched on April 30, 1996 and made successful observations from 1996 to 2003. it

provided some opportunities for optical observations of GRBs like the determination

of the position of GRBs with identification of GRB afterglow in X – rays.

BeppoSAX led to the understanding of those sources being extragalactic. GRB

observations were made in high energy X – ray band (2 – 25 keV) with BeppoSAX.

GRB positions were discovered with WFC (Wide Field Camera) and the error box

was established in the direction of known positions with Narrow Field Instruments

(NFI) in the range of 0.1 – 10 keV and then GCN (GRB coordinates Network)

distributes the locations of detected GRBs and ground based observers attempt to

detect optical, radio, and infrared afterglow observations. The burst coordinates can

be localized to about an arc minute position accurate within 4 – 6 hours by these X –

ray data that allowed the identification of afterglows at optical and longer

wavelengths with follow – up observations (van Paradijs et al. 1997, Frail et al.

1997).

Distance measurements have been made with the results of optical observations

and the host galaxy of some of these bursts were identified. Thus the cosmological


8

distance for these bursts was confirmed (Metzger et al. 1997, Kulkarni et al. 1999b).

56 GRB afterglows have been localized with X – ray and optical observations and

more than a dozen of them detected extending to radio waveband (Frail et al. 1999,

Weiler 2002). Some of the GRBs whose afterglows were observed by BeppoSAX are

as follows; GRB 960720, GRB970111, GRB970228, GRB 970402 and GRB

970508, GRB 971214, GRB 980329, GRB 980425 (Frontera et al. 1999).

One of the most important GRB observations of BeppoSAX was the detection

of GRB 970228. The long-lived multi-wavelength emission of GRB 970228 was

discovered in X-rays (Costa et al. 1997) and optical (van Paradijs et al. 1997) but no

emission in radio band (Bloom et al. 2001). GRB 970228 was discovered with the

NFI which was the first X – ray afterglow source in the BeppoSAX capabilities. X –

ray emission was reported from ~8 hr to ~13 days after the GRB event (Costa et al.

1997b, Yoshida et al. 1997) and optical emission was from ~17 hr to ~37 days of the

event (Groot et al. 1997a, 1997b, Metzger et al. 1997a, van Paradijs et al. 1997).

GRB 970228 was also detected with RXTE (Rossi X – ray Timing Explorer)

(Remillard et al. 1997), BATSE and Ulysses. The earliest optical data of the burst

was obtained 15.4 hrs after the event (Pedichini et al. 1997, Galama et al. 1997).

Then the brightness of optical source decreased monotonically and was fainter in R –

band by a factor 40, 6 days after the explosion (Pedersen et al. 1998). The redshift

determination failed despite the early afterglow spectroscopy of the event. Then the

redshift was determined as z = 0.695 investigating a faint galaxy spectroscopy at the

same location of the GRB 970228 afterglow (Bloom et al. 1998).

GRB970508 was discovered by BeppoSAX on May 8, 1997, 15 s after the

event. It was also detected by BATSE on CGRO (Kouveliotou et al. 1997). The

source location was established with 10´ precision within 3 hr. In this error circle an

X – ray source (Band 1997) and a radio source (Frail et al. 1997, Taylor et al. 1997)

were also detected. As the result of spectroscopic observations the redshift z ≥ 0.835

was found. Therefore the first distance determination for a GRB was made. The first

measurements of the redshift of GRBs showed that these sources were located in

cosmological distances. The total energy output was also estimated as 1052 ergs with

an assumed redshift of z = 1 (Waxman 1997). GRB990123 observed by BeppoSAX


9

is important to prove that GRBs have large amounts of energy. If the explosion

energy is released equally from all directions it has 1045 W radiation energy that is

1019 times larger than the energy of the Sun (1026 W).

The ESA mission INTEGRAL was launched on the Russian Proton rocket on

October 2002. Currently, the high spatial and spectral resolution gamma – ray

observations are delivered by INTEGRAL. These observations are being made in the

15 keV – 100 MeV energy range using its two main instruments SPI spectrometer

and IBIS imager. Moreover, INTEGRAL also records sources in the 4 – 35 keV X –

rays using JEM – X (Joint European X – Ray monitor) and in the 500 – 600 nm

optical band using OMC (Optical Monitoring Camera). INTEGRAL is one of the

GCN observatories therefore it is an important monitor to provide accurate location

of GRBs for other space and ground based observatories. BAS (Burst Alert System)

trigger system of INTEGRAL determines the position of GRBs. The first GRB

observed by Integral is GRB 030227. 56 bursts have been detected and localized

from November 2002 to September 2008 by INTEGRAL. Some of GRBs were

observed by INTEGRAL are as follows; GRB 030406, GRB 030501, GRB 031203,

GRB 040106, GRB 040323, GRB 040730, GRB 050223, GRB 060901, GRB

070311, GRB 070707, GRB 080723B (Moran et al. 2004, Vianello et al. 2009). The

source catalog of INTEGRAL consists of more than 400 high – energy sources

detected in the range 17 – 100 keV. The scientific goals of INTEGRAL are to study

compact objects, extragalactic astronomy, stellar nucleosynthesis, galactic structure,

particle processes and acceleration, and identification of high energy sources.

Soft Gamma Repeaters (SGRs) are a small subgroup of the classical gamma –

ray bursts. None of the SGR > 20 keV were detected before INTEGRAL. SGR 1806

– 20 and SGR 1900+14 have been observed for the first time by deep observations of

IBIS with a ~1036 erg s−1 luminosity in the 20 – 100 keV range.

The small scientific space satellite HETE 2 (High Energy Transient Sources

Experiment ) is designed to detect the localizations of GRBs. It was launched on

October 9, 2000. HETE 2 is a satellite with international participation designed by

MIT (Massachusetts Instute of Technology) USA. It is sensitive to 1 keV – 500 keV

photons. The location of the GRBs are computed by HETE 2 then the precise


10

coordinates of GRBs are communicated to the GCN for immediate follow – up

observations and to allow detailed studies of the initial phases of GRBs (Vanderspek

et al. 1999). The main aims of HETE 2 mission is to observe the GRBs in the

multiwavelength regime and HETE instruments are also able to make a survey of X

– ray sky and sensitive to prompt flashes from X – ray sources in the sky.

HETE 2 observed more than several hundred GRBs and localized 84 of them

during the year from 2000 to 2006 (Lamb et al. 2004,

http://grbhuntsville2008.cspar.uah.edu/content/Posters/P073_Munz.pdf). HETE 2

has identified the location of ~25 – 30 GRBs per year. The afterglows of 21 of these

bursts were observed in X – rays, optical and radio band. One of the important

observations of HETE 2 is to discover the optical afterglow of GRB 030329 (=SN

2003dh). GRB 030329 is one of the brightest GRBs ever detected and it is linked

with a supernova (SN 2003dh). GRB 011130, GRB 020531, GRB 020813, GRB

021004, GRB 030226, GRB 030328, GRB030329, GRB 030528, and GRB 030723

are the important bursts detected by HETE 2 (Richer et al. 2003). HETE 2 has

detected a short duration GRB which is GRB 050709 . The afterglow emission of the

burst has been observed from gamma-ray band to optical. The mystery of short

GRBs has been solved with these multiwavelength observations (Ricker et al. 2005).

In addition to these satellites RXTE (Rossi X – ray Timing Explorer) has been

making GRB observations since December 1995. The PCA (Proportional Counter

Array) has localized X – ray afterglow in the BATSE LOCBURST error regions.

RXTE is sensitive to 2 – 250 keV energy range. X – ray light curves of 15 GRBs

detected by the All – Sky Monitor (ASM) on RXTE in 1,5 – 12 keV energy range

were presented. These soft X – ray light curves were compared with high energy

experiments (BATSE, Konus – Wind, the BeppoSAX Gamma Ray Burst monitor,

and the burst monitor of Ulysses), and these light curves simple relativistic fireball

could be linked with synchrotron shock mechanisms (Smith et al. 2002).

Chandra is an X – ray satellite which is sensitive to 0.1 – 10 keV energy range.

It was launched July 23, 1999. Chandra provides contribution to solve the mystery of

GRBs. As a result of photometric and spectral analyses of Chandra, the important

information about nature, progenitor, and host galaxies of GRBs were indicated.

http://grbhuntsville2008.cspar.uah.edu/content/Posters/P073_Munz.pdf


11

Chandra data also showed the surrounding medium of the bursts. Moreover, Chandra

observation of GRB 010222 afterglow points out the evidence for these massive

explosions taking places at the places where stars are born. There are also several

theories about the nature of gamma – ray bursts. The important theories of GRBs

include the combinations of merging neutron stars and black holes or explosion of

massive stars. These are all called Hypernovae. Recent Chandra X – ray observations

supporting hypernova model were found with the observation of Chandra X – ray

observatory (http://www.chandra.harvard.edu).

GRB 980425 / SN 1998bw have been observed 1281 days after the gamma-ray

burst using Chandra X – ray observatory. Eight X – ray point sources were localized

in the GRB error box. Well-pointed radio location of SN 1998bw and one of these

point sources coincided. So this point source is probably the remnant of the

explosion. As a result, the contribution of GRB afterglows to the ultraluminous X –

ray source population were pointed out (Kouveliotou et al. 2004).

ROTSE (Robotic Optic Transient Source Experiment) telescopes are very

important for working on GRB afterglows with an international group. It is dedicated

to quick detection and observation of optical afterglows of GRBs. Four ROTSE

telescopes were installed at several remote locations. Two of them were installed at

the northern hemisphere (Turkey and USA) and others were placed in southern

hemisphere (Namibia and Australia). Therefore, continuous observations of the night

sky (24 hours per day) can be possible.

A dozen explosions were observed after approximately 2 years in ROTSE

operation. One of the important observations of ROTSE is GRB 990123. It was

observed optically from the ground based observatories while main part of the

explosion is in progress (Akerlof et al. 1999).

Rapid response ability of ROTSE III led to the study of some of the bursts

properties in terms of energetics, temporal structure, and spectral index which are

required to understand the GRB behavior during the transition from the prompt to

afterglow phase. The reviewed prompt optical observations of ROTSEIII include

GRB 060111B, GRB 060729, GRB 060904B, GRB 061007, GRB061121, GRB

060927 given by Yuan et al. (2009).

http://www.chandra.harvard.edu)


12

Another important gamma – ray burst observatory Swift was launched on

Nowember 20, 2004 as a part of NASA’ s medium explorer (MIDEX) program.

Swift is dedicated to solve the gamma – ray burst mystery. Swift delivers a burst’s

location to give the opportunity to observe the burst’s afterglow both ground based

and space based telescopes. Swift has observed 237 bursts during 2.5 years of

operation. It is also capable of X – ray (with X-Ray Telescope (XRT)), UV, and

optical (Ultraviolet and Optical Telescope (UVOT)) observations. Swift opens a new

phase to solve mysteries of short duration GRBs (Gehrels et al. 2005, Barthelmy et

al. 2005b). GRB 060218 was detected by Swift that is associated with a supernova

(Campana et al. 2006). Swift X – ray observations also give a new insight about the

fireball model and nature of central engine of GRBs (Zhang et al. 2006, Sakamoto et

al. 2007).

Some of the important bursts observed by Swift are as follows; GRB 050128

(X – ray afterglow was observed by Swift, Campana et al. 2005), GRB 050801

(optical, ultraviolet and X- ray observations were obtained by Swift, Massimiliano et

al. 2007), GRB 080319B (could be observed with naked human eye that has 5.3

visual magnitude, Racusin, et al. 2008).

The last large high-energy gamma-ray mission GLAST (The Gamma-Ray

Large Area Space Telescope) was launched on June 11, 2008. The main instrument

LAT (Large Area Telescope) studies the sky in 20 MeV - 300 GeV energy range.

GLAST is 30 times more sensitive than the highly successful EGRET detector. LAT

and GBM (Gamma-Ray Burst Monitor) are the main detectors of GLAST. High

energy phenomena of active galaxies, the optical – UV extragalactic background

light, pulsars, GRBs are thought will gain a new and important sight with LAT.

Spectral and timing information of GRBs will be provided in the 10 keV – 30 MeV

range with GBM (http://www-glast.sonoma.edu/).

Also 10 afterglows have been discovered from 1997 to 2007 at the Special

Astrophysical Observatory, Russian Academy of Sciences by 6-m BTA and 1-m

Zeiss telescopes by the group under the leadership of Dr. V. V. Sokolov and A.J.

Castro-Tirado. Some of these bursts are as follows; GRB 030329, GRB 050408,

http://www-glast.sonoma.edu/)


13

GRB 060218, GRB 070610. The observations of these bursts have led to solve the

mystery of GRBs.

It is known that GRBs are the high energy explosions in the Universe. The

progenitors of these explosions are trying to be understood since their discovery.

Different progenitor models are assigned by theorists to explain progenitors of bursts

with afterglow observations of GRBs in different wavelengths. These models are

divided in two classes. One of them involves massive stars that collapse directly into

a black hole at the end of its life named as Collapsar or Hypernovae (Woosley 1993,

Mac Fadyen & Woosley 1999, Woosley 2002, Zhang & Meszaros 2003). It may

produce an energetic long-duration (t > 2s) GRB. Another class of the progenitor

model involves the merger model of two compact objects (Neutron Star (NS) – NS,

Black Hole (BH) – NS, White dwarf (WD) – BH mergers) as a result of orbital

angular momentum lost by gravitational wave radiation (Eichler et al. 1989, Fryer et

al. 1999, Zhang & Fryer 2001, Belczynski et al. 2001, Zhang & Meszaros 2003). The

merger of two compact objects model explains the progenitor of short (t < 2s) GRBs.

Collapsar is a fast rotating massive star which is stripped of its outer hydrogen

envelope at relativistic speeds with a Lorentz factor of ~150. It collapses into a black

hole that is surrounded with a flow disk.

Another alternative progenitor model is a neutron star with an extremely high

magnetic field that is named as magnetar (Eichler et al. 1989, Belcynski et al. 2001).

Long GRBs are generally believed to be produced by the core collapse of

massive stars (Woosley & Heger 2006). On April 25, 1998 BeppoSAX and

BATSE/CGRO recorded a GRB (GRB980405). Approximately two and half days

later a powerful component of the radiation of a supernova (SN1998bw) in the

spectra and in the light curve was detected in the BeppoSAX error box on the GRB

(Galama et al. 1998). Finally, the coincidence of these two events indicated the

association of the two events (Li – Xin Li, 2008). So far four pairs of GRBs-SNe

connection has been established since GRB980425/SN1998bw. They are

summarized in Table 2.1 (from the paper of Li – Xin Li, 2008), from which it is seen

that such spectroscopically confirmed connections occur very rarely and for rather


14

close GRBs. It can be assumed that such observations can be carried out once every

two-three years (Chapman et al. 2007).

Table 2.1 Spectroscopically confirmed GRBs and SNe connections ( Li – Xin Li, 2008).

GRB/SN redshift, z Mpeak, bolometric

980425/1998bw 0.0085 −18.65 ± 0.20 030329/2003dh 0.1687 −18.79 ± 0.23 031203/2003lw 0.1055 −18.92 ± 0.20 060218/2006aj 0.0335 −18.16 ± 0.20

15

3.BRIEF OVERVIEW AND METHOD Eda SONBAŞ

3. BRIEF OVERVIEW AND METHOD

3.1 Theory of Gamma-Ray Bursts

Cosmic GRBs are bright, short (from parts of second to hundreds seconds)

bursts of mostly soft gamma-radiation with total fluencies (time integrated total

energy per unit area) from ~ 10−7erg cm−2 to 10−3 erg cm−2. Events of such levels are

registered with an average rate of about 0.8 bursts per day (Fishman et al. 1994).

Their light curves often show a complicated and multipeak structure with

variabilities at time scales up to tens of milliseconds, and energy spectra with a

maximum up to ~1 MeV (Schaefer et al. 1994). However, the burst radiation is also

observed down to the X – ray range, 2 – 30 keV (Amati et al. 2002). Before 1997 all

information about these events was obtained only from observations with all-

directional gamma-ray detectors placed on space platforms.

The problem of identification of such burst sources on the other wavelengths

and setting of their distance scale has become important since right after their

identification.

After optical identification and recognizing the cosmological nature of

several long-duration GRBs, they turned out to be the most powerful explosions in

the Universe observable. The luminosity during the burst (of duration from several to

100 seconds) can achieve L ~1052 erg s−1 (isotropic equivalent), whereas, for

example, Active Galactic Nuclei (AGNs) can have L ~1048 erg s−1 (when averaged

over a long time) and supernovae (SNe) can have L ~1045 erg s−1 . The short

timescales (i.e., ≤ miliseconds) of the gamma-ray emission in a GRB suggest already

very small dimensions for the source of the order of tens of kilometers.

Besides, the Universe is transparent in gamma-rays up to zo ~10 (Moskvitin et

al. 2008). That is why the observation of GRBs is a powerful tool for studying

physical conditions of environment and the processes of birth and death of stars at

redshifts up to zo and possibly, even more: at present, the redshift of the most distant

GRB 050904 is z = 6.29 (Kwai et al. 2006) and GRB 090323 is z = 8.2

(http://science.nasa.gov/headlines/y2009/28apr_grbmask.html?list884969).

http://science.nasa.gov/headlines/y2009/28apr_grbmask.html?list884969)

16


Furthermore, GRBs and their afterglows are of great interest for studies

related to stellar astrophysics, the interstellar and intergalactic medium, and, the most

important, they also reveal themselves as unique probes of the high redshift Universe

closer to the Big Bang. After the first X – ray identifications carried out by

BeppoSAX in 1997 (Costa et al. 1997), the fast localization of a GRB and follow-up

observation of their afterglow became possible practically in all wavelengths, from

X-rays to radio. The optical range is the most important for studying the relation of

GRBs with known objects. For studying GRBs at large redshifts, the near infrared

(NIR) range has become the most informative because optical line emissions of

several GRB afterglows (the GRB optical transients) were found to shift there.

The observation strategy of GRB optical transients (OTs) and host galaxies

can be described as shown in Figure 3.1.

17


Figure 3.1: Summary of astrophysical processes and questions related to GRBs are included firstly, the observational data acquision of GRB OTs and their host galaxies then the interpretation of these acquised data (http://w0.sao.ru/hq/grb/astronomy-6m.html).

3.2. Progenitors of GRBs

Although, the studies on GRBs show great progress in recent years, the origin

of these explosions is still not known thoroughly. While pulsars, quasars and X – ray

http://w0.sao.ru/hq/grb/astronomy-6m.html)

18


sources (binary or otherwise) could be satisfactorily explained shortly after their

discovery, the true origin of GRBs are still not fully known almost 40 years. In this

sense, these explosions are quite unique events in the astronomy. GRBs were divided

into two subgroups according to their durations as long (t > 2s) and short (t < 2s)

bursts (Kouveliotou et al. 1993, Katz & Canel 1996). Therefore, it is thought that,

these two subgroups can be explained by different progenitor models. Core collapse

of massive stars at the end of their evolution (also referred as Collapsar/ Hypernova

or Supranova events) could explain the progenitor of long duration GRBs (Woosley

1993, Meszaros 2002, Pacyniski 1998, Fryer 1999).

If the long-duration GRBs can be associated with the collapse of short-lived

massive stars, with peak emission at sub-MeV energies (where dust extinction is not

an issue) the fluxes of these events is expected to be detected eventually from almost

any redshift. Since gamma-rays do not suffer any significant absorption, GRBs are

expected to trace the massive SFR even in very dense molecular regions of any

galaxy in the Universe, most probably, of the dust-enshrouded actively star-forming

galaxies at high redshifts.

However, progenitors of short duration GRBs were thought to be the mergers

of any combinations of two compact objects (i.e., Neutron Star – Neutron Star,

Neutron Star – Black Hole, Neutron Star, White Dwarf) in a binary star system

(Eichler et al. 1989, Meszaros & Rees 1997). The schematic model for these theories

are summarized in Figure 3.2.

19


Figure 3.2: Schematic representation of the progenitor models of GRBs.(www.daviddarling.info/images/gamma-ray_bursts.jpg). In the long GRB section as a final product a black hole in formed with an accretion disc and possibly powerful jet of gamma-rays.

http://www.daviddarling.info/images/gamma-ray_bursts.jpg

20


3.2.1. Collapsar / Hypernova Models

Rotating massive stars (M > 25 Mּס) without a hydrogen envelope are thought to be

the starting points of a “Collapsar” model for GRBs. At the ends of their evolution,

the central regions of these stars collapse to a black hole that is surrounded by an

accretion disk (Woosley 1993, Mac Fadyen & Woosley 1999, Woosley et al. 2002).

The collapse along the star’s equator is slowed by the conservation of angular

momentum while the envelope of star falls into the newly formed black hole. The

forming accretion disc powers the GRB (Woosley 1993, Mac Fadyen & Woosley

1999). A long duration GRB ( ≥ 10s ) can be formed by such collapsars with the

large enough masses. A Wolf – Rayet star that is stripped of its hydrogen produce the

collapsar. A collapsar event can be explained with several models:

According to the Type I (standard) collapsar model, a black hole forms

promptly with collapsing of the iron core of a ~ 25 Mּס massive star. According to the

calculations, a collapsar’s angular momentum value should lie in the range of J ~ 3

× 1016 – 2 × 1017 cm s-2 (Mac Fadyen & Woosley 1999). In this case the massive star

must have lost most of its hydrogen envelope and sufficiently large ( > 10 Mּס)

helium core was retained. As a result, the evolved He core collapses to a black hole

and creates a (long duration) Gamma Ray Burst. If the angular momentum is lower

than the value of ~ 3 × 1016 – 2 × 1017 cm s−2, an accretion disc will not form and

core matter, more or less directly, collapse into a black hole. In the case of a too large

angular momentum, the disk does not loose energy efficiently (via neutrino

emission) and creates a weak outflow. This kind of massive stars may have low

metallicities (Fryer et al. 1999)

The most important problem of the collapsar models is the uncertainty in the

angular momentum. There are different opinions to know how much angular

momentum is held in the stellar core of collapsing massive stars. Heger (1998) has

found that, with the results of massive stars evolution models that the angular

momentum of the cores of most stars are similar to the range proposed by Mac

Fadyen & Woosley (1999) which is given above. Spruit & Phinney (1998) have

disputed this result since the rigid rotation of stars, magnetic fields will couple the

cores and envelope of such stars (Fryer et al. 1999).

21


According to a Type II scenario, on the other hand, the companions in a

binary star system remove their hydrogen envelopes during the so called ‘common

envelope’ phase. If the primary star is massive enough after common envelope

evolution, it will collapse into a black hole. This way, when this system still remains

as a binary, in later phases it can evolve into a BH – NS pair from a WD – BH binary

(Fryer et al. 1999).

These two scenarios may both fail due to magnetic breaking of the core’s

rotation. If this is the case, it is possible to form a collapsar event if the two stars

have comparable masses, and evolve together in the main sequence (Spruit &

Phinney, 1998). Than two helium cores formed will merge and form a more massive

helium core that has a large enough angular momentum, during their common

envelope evolution. This helium core will then collapse and a GRB event will occur


3.2.2. The Supranova Model

Supranova model is similar to collapsar model in many details. The

production of GRB is foreseen during the collapse of a massive star. In the model a

large amount of material enriched in heavy elements surrounds the explosion that

will form the GRB. The shell of matter is at a sufficiently far distance to GRB to

cause the observed delay in some GRB’s. However this matter does not prevent us to

see the burst. As was reported in the several bursts; X – ray emission lines can be

produced with the irradiation of the surrounding material during the initial burst or

afterglow emission phase (Woosley et al. 2002).

The supranova (suppramassive) model of a GRB event was suggested by

Vietri & Stella (1998). When suppramassive neutron star assumed during its

evolution start losing too much angular momentum, centrifugal forces cannot balance

gravitational force and star collapses directly into a black hole. Supranova model of a

GRB is explained in two steps. In the first step; an extremely massive star makes a

core collapse during which the outer layers of the star are ejected. A rapidly rotating

black hole is formed by the collapsed core that is surrounded by the rotating disk of

matter.

22


In the second step; a jet of high energy particles is produced by this black

hole system. The X – ray and/or gamma ray bursts (lasting only several minutes) are

produced by the shockwaves within the jet. X – ray emission (afterglow) is also

produced through the interaction of the jet and ejected supernova shell. This

interaction can last for days or even months. The delay between the formation of the

black hole and the jet formation mechanism has not clearly been understood yet.

Another alternative to the supranova model a massive magnetized NS formed

as the result of supernova explosion that is provided rapid rotation to the NS. Until

the gravitational collapse becomes unstable, this suppramassive NS lose its rotation

that is trigger a GRB event by goes to the BH formation. Suppramassive neutron

star sweeps the surrounding medium during the supernova explosion and it is

proposed so far that, this may be in the most baryon-clean environment. This

environment is thought to be the in the prompt high energy region with the high (>

100) Lorentz factor. The quickly spinning remnant loses energy through magnetic

dipole radiation. Because of the short mean free paths of neutrinos and their prompt

collapse of the polar caps the explosion is adiabatic in addition to this, it can easily

reach centrifugal equilibrium in the equatorial belt with ~ 0.1 Mּס mass. Energy

extraction mechanism is due to the pointing flux conversion into a magnetized

relativistic wind. Quickly decaying or nondetectable afterglows will be produced by

this model (Vietri and Stella 1998, Königl 2003).

3.2.3. Supernova Models for GRBs The study of core-collapse supernovae (SNe) is particularly interesting in

general because it is believed that long-duration Gamma Ray Bursts (GRBs) are

somehow connected with core collapse SNe. Actually long-duration GRBs are to be

produced by the core collapse of massive stars. The spectroscopic and photometric

observations of (core collapse) SNe in general are of great importance for

understanding the mechanism of the explosion of the massive SN itself, because the

explosion mechanism of most massive progenitors is still a puzzle. In principle core-

collapse SN can be naturally explained by the axially-symmetrical explosion of

23


massive progenitors. Such axially-symmetrical explosions can be understood by the

nebular phase observations of core-collapse SNe.

SNe are the violent explosions at the end of star’s life. They are classified

according to their spectra and light curves. If hydrogen lines are present in their

spectra, these are classified as Type II. Type II SNe are further divided into three

sub-classes by their light curve shape and spectral details: Type IIL, Type IIP, Type

IIn. First two types SNe show a linear decline and “plateu” in their light curves. Type

II SNe with narrow lines is classified as Type IIn. Type I supernova on the other

hand does not show hydrogen lines in their spectra. This class of supernova events

are also divided into three sub-classes. If SNe that have silicon lines are classified as

Type Ia. Otherwise SNe that show helium lines but do not show silicon lines in the

spectra are classified as Type Ib. Type Ic supernovae are very similar to Type Ib SNe

with lack of conspicuous He I lines. Basic properties of supernovae are shown in

Figure 3.3. Type Ib and Ic supernovae show many similarities therefore, they are

sometimes collectively called Type Ibc supernovae. Generally Type Ibc and Type II

supernovae are classified as core-collapse supernovae because these SNe are

believed to be produced by core collapse of massive stars (Li – Xin Li 2008). In the

core collapse supernovae the final stage of nuclear fusion in hydrostatic burning

contains the heavy element iron. At this stage the iron core becomes larger and

because of the silicon shell burning, its mass becomes larger than the Chandrasekhar

mass limit (~ 1.44 solar masses); thus the core cannot be stabilized by electron

degeneracy pressure hence it collapses (Janka et al. 2007)

24


Figure 3.3 Classification of supernovae

3.2.4 Observational Evidence All GRBs, that are associated with SNe, were identified with a quite definite class of

the stripped envelope SNe Ibc . SN Ib and Ic have been identified and observed for a

long time ago and their most likely progenitors are believed to be Wolf – Rayet stars

surrounded by a dense wind envelope, resulting from the massive star's evolution.

During the explosion of the evolved core a shock develops and traverses the

envelope and produces bright and short X – ray and UV flashes. These flashes may

last several hours and the duration of the flash depends on how massive and extended

was the wind envelope that surrounded the progenitor star before the SN explosion.

The interaction between the shock produced by the SNe explosion and the envelope,

also known as the “shock break-out” effect had been predicted some time ago before

the identification of GRBs (Colgate 1968), and was revised by several authors

(Bisnovaty-Kogan et al. 1975, Blinnikov et al. 2002). Massive explosions occur

25


when the nuclear reactions inside the core is exhausted. Then the core collapses

under its own mass and a shock wave are generated outward through star by the

stellar breakdown. The material that is in front of the shock wave is heated by the

moving shock wave to several 105 K degrees. For a short time, the star shines mostly

in ultraviolet and X-rays. This is known as the shock breakout

(http://news.nationalgeographic.com/news/2008/06/080612-supernova.html). In

particular, Calzavara and Matzner (2004) dedicated their paper to future systematic

observations of this effect. However, before the XRF 060218 / SN 2006aj event the

effect in question could have been fully observed only for a small number of massive

SNe. Another example as a very short duration flash– due to the compact size of a 20

– 30 Rּס blue supergiant effect of the interaction of the shock was also observed in the

famous SN 1987A of Type II (Ensman and Burrovs 1992).

Another well studied event is SN 1993J. This SN was first classified by its

spectrum as Type II because of existence of hydrogen lines in its early spectra.

However, after a certain time this SN changed its type to Ib (Filippenko et al. 1993);

that is hydrogen confidently ceased to be in their spectra. Its early spectra show a

strong UV excess, which is considered to be the characteristic of the shock

interaction.

The general belief is that core collapse supernovae connected with X-ray

Flash (XRF)GRB events can be naturally explained by an aspherical (axially-

symmetrical) explosions of massive SNe. The common assumption is that in the case

of an XRF, the observer is located outside the cone of radiation in which for some

reasons, the bulk of gamma-ray radiation is concentrated. The asphericity is

generally observed in the nebular phase observations (Pian & Mazzali 2006,Leonard

et al. 2006, Maeda et al. 2008, Modjaz et al. 2008, Valenti et al. 2008).

In among popular Type Ibc SNe is SN 2003dj. It is also highly aspherical.

Howewer this bright and very energetic SNe was not associated with a GRB event

(Pian & Mazzali 2006, Hurley et al. 2003). This SN has shown a doubled peaked

[OI] emission line profile. The assumption for this feature was the aspherical

explosion similar to the earlier example of SN 1998bw. Explanation is that due to the

location of line of sight of the observer (which is viewing the event along the

equator), we cannot see any GRB event connected with SN 2003jd (Figure

http://news.nationalgeographic.com/news/2008/06/080612-supernova.html

26


3.4)(Flippenko et al. 1993). Thus, the double-peaked [OI] emission must be observed

for SNe which were not accompanied with GRBs, like SN 2008D. On the other hand

single peak of [OI] emission is observed in the nebular phase of SNe which are

accompanied with GRBs, as in the case of GRB 060218/SN 2006aj.

Figure 3.4 Nebular line profiles of two supernovae according to line of sight of observer (Pian & Mazzali 2006). Nebular line profiles observed from an aspherical explosion model depend on the orientation. The figure shows the properties of the explosion model computed in 2D: Fe (colored in blue) is ejected near the jet direction and oxygen (brown) in a disc-like structure on and near the equatorial plane. Density contours (covering 2 orders of magnitude and divided into 10 equal intervals in log scale) reflect the dense disc-like structure. Synthetic [O I] 6300, 6363 Ǻ lines (red lines) computed in 2D are compared with the spectra of SN 1998bw and SN 2003jd (black lines). (from P.Mazzali et al. 2005), “An Asymmetric, Energetic Type Ic Supernova Viewed Off-Axis, and a Link to Gamma-Ray Bursts”)

SNe explosion observations have much longer history than GRB

observations. Therefore understanding of SNe events should be much clearer than

GRBs. However some details of explosion mechanisms of SNe are not solved yet (Li

– Xin Li 2008, Janka, et al. 2007). Therefore the early spectroscopical observations

of SNe that are connected with GRBs is the key moment for the further

27


understanding of explosion mechanism of core-collapse SNe and other questions

which are not properly solved yet.

3.2.5. Role of Binary Star Systems in GRB Events The exact mechanisms that produce GRBs are not yet fully known: however,

it is clear that models based on black hole accretion disks (BHAD) have increased

their prominence. Because this type of models can explain the large amounts of

energy and potentially short time scales induced. Moreover, BHAD models

contribute to magnetic fields with favorable conditions to extract energy. These

result either from the gravitational potential energy of the accretion disk itself or

from the rotational energy of spinning black holes (Fryer et al. 1999).

BHAD models for GRB production mechanisms principally, are based on,

either the evolution of a single massive star or the evolution of massive stars in a

binary system. During the evolution of a single massive star a helium core is

produced by mass loss and a collapsar event will occur as described above. The

evolution of massive stars in a binary system, however can be addressed with

different merger scenarios. The merger model of a double neutron star system is the

best studied one as a BHAD based progenitors. Moreover, these models lead to

similar conditions to the mergers of neutron star - black hole (NS - BH), a white

dwarf and a black hole (WD - BH) or a helium star with a BH or an NS (Fryer et al.

1999, Woosley et al. 2002)

Two massive stars are required to begin the standard scenario for forming

close-orbit double neutron star (NS-NS) systems. The more massive main star that

has > 25 Mּס mass, overfills its Roche Lobe and starts to transfer its mass to the lower

mass (25 Mּס > Mc > 10 Mּס) companion star. The main star then evolves and an NS is

formed with a supernova explosion. During the neutron star formation, binary system

remains bound if the system does not gain a kick. Then, this system passes to the X –

ray binary phase and companion star (Mc) which is in the common envelope,

expands. During the common envelope phase (in the standard model), NS spirals

around the more massive companion star and released orbital energy would lead to

eject companion’s hydrogen mantle. Then a neutron star – helium star binary will be

28


formed. The companion star (a Helium star) then reaches at its SN phase and an NS

is created with this explosion and finally a close NS – NS binary remains in there.

Double neutron stars merge via gravitational wave emission over time and produce a

GRB at their final collision (Fryer et al. 1999).

The formation of a BH – NS binary system and producing a GRB with

merging a binary system are identical to the scenario for the above mentioned double

neutron star scenario. Only the primary star mass should have a mass (Mps) above

the critical mass MPS ≥ 25 Mּס at the beginning so that a MBH ≈ 25 Mּס may result


Another binary merger scenario that may create a GRB is the White Dwarf –

Neutron Star (WD – NS) scenario. In this model the mass of WD should satisfy

Mmin,WD > 0.9 Mּס. The mass transfer from WD to NS makes the system unstable.

When the total mass of binary star system ensures the MWD + MNS > Mmax, NS

+0.3Mּס condition, NS will get enough mass to collapse to a BH. To generate a GRB

with the merge of WD – NS, at least a 0.3 Mּס disk must be produced (Fryer et al.

1999).

As a last alternative for, a WD – BH merger scenario we begin with a primary

star of mass < 10 Mּס and a companion star with 2 – 3 Mּס. In this binary system the

massive star completes its evolution first, and a He star is produced. They loose

(eject) their Hydrogen in the common envelope phase and then primary star

collapses. Later a red giant will be produced by the evolution of the companion star,

and a WD – BH binary will be produced. WD can loose too much material onto the

accretion disk of BH. A GRB can be produced using this accretion disc. WD must

have > 0.9 Mּס mass for this to occur (Fryer et al. 1999).

3.2.6. The Fireball Model Another popular model that attempts to explain GRBs is the so called, fireball

model (see Figure 3.5). It can explain how these large amounts of energy concentrate

in a small region of space. This model was first proposed starting from

29


Figure 3.5 Schematic representation of fireball model (www.rotse.net/summary/defs/fireball.html).

the evolution of a relativistic expanding shell by Piran (1999). A solution was

provided to the relativistic expanding of shell problem by Blandford and McKee

(1976). In the expansion phase, a radiation pressure from the actively radiating

central engine was predicted. This determined an acceleration period and the

transpiration of the radiation to a certain distance. Then a fireball formed and was

accelerated and relativistically expanding outwards with constant Lorentz factor (γ α

R). This relativistic expanding shell solves the compactness problem, The

compactness problem can be resolved when the source is moving relativistically with

a high Lorentz factor, γ. The estimation of total energy from the burst at a distance D

from us, fluence F as follows;

E = 4πD2F = 1050 ergs (D / 3000 Mpc)2 (F / 10−7 ergs cm-2)

(3.1)

http://www.rotse.net/summary/defs/fireball.html

30


The typical source size, Ri < cδt ≈ 3000 km, can be calculated using the rapid

temporal burst variability (δT ≈ 10−2 sec). γ rays in the fireball with energy E1, will

produce electron positron pairs through (γγ → e+e-) rays and lower energy photons

that have an energy E2

2

21 cm>EE e (3.2)

This condition the average optical depth with an fp fraction of photon pairs (E1, E2)

can be expressed as follows;

22

713

22 sec103000/1010

−

−

mδT

MpcD

cmergsFf=

cmRFDσf

=τ 2pei

2Tp

γγ (3.3)

where σT is the Thompson cross section and fp is the fraction of photon pairs. This

expression can also be described in the term of bulk Lorentz factor. This typical

optical depth is huge for pair production and not for high energy photons. But its

value should reduce to less than unity because of the optically thin observed spectra.

The emitted photon energy from the source is blue-shifted by the bulk Lorentz factor

11

12

⟩⟩−

=Γβ (3.4)

When the source energy is lower, fewer photons can produce pairs with a factor

smaller by Γ−2α (http://glast.pi.infn.it/Nicola/GRB_Model/node2.html). Here α is a

high energy photon spectral index. However, in the relativistic scheme, the radius of

fireball should be greater with Γ2 factor (i.e., Re < Γ2 c δT). Therefore, the optical

depth can then be expressed as;

22

27)24(

13

sec103000/1010 −

−+

Γ

≈mT

MpcD

cmergsFf p

δτ

αγγ (3.5)

http://glast.pi.infn.it/Nicola/GRB_Model/node2.html)

31


According to fireball model, a plasma “shell” is emitted by the hidden central

engine, with a bulk Lorentz factor, Γ, to the interstellar medium. These shells can

move with different velocities to produce shocks so that faster shells can reach

slower ones. This energy can be used for particle acceleration and magnetic field

generation. Energetic electrons loose energy by synchrotron emission. Highly

energetic photons are also produced by the inverse Compton scattering of

synchrotron photons a mechanism known as synchrotron self Compton mechanism

(Piran 2004, http://glast.pi.infn.it/Nicola/GRB_Model/node2.html).

Fireball model can also explain several observational data (Piran 2004);

• The total energies of GRBs are ~ 1051 erg that is about the binding energy of

a stellar compact object. This is enough to accelerate a mass of particles

~10−5 Mּס to relativistic velocities and this energy can be generated by the

“inner engine”.

• The relativistic flow must be collimated by the “inner engine” with a typical

opening angle (1o < θ < 20o) for most GRBs.

• GRBs are divided into two subgroups according to their duration as long (t >

2s) and short (t < 2s) bursts. This implies that there are two different inner

engines because of the duration of GRBs are determined by their inner

engine.

• The rate of GRBs (once per 3 × 105 yr per galaxy) also implies an inner

engine of stellar masses.

• The “inner engine” activity must also determine the time scales of GRBs

(duration of long GRBs are in the order of 50 sec) through the internal shock

models. The time scale of variability δt ~ 1 ms, indicates a compact (~ 100

km size) object. Since the duration of GRBs has to be longer than dynamic

time scale, a different time scale should be operating within the “inner

engine”. This shows a prolonged activity (δt/T << 1 for short burst) and

denies the “explosive” model in which energy comes from a single explosion

(Piran 2004).

3.2.7. GRB Afterglows

http://glast.pi.infn.it/Nicola/GRB_Model/node2.html)

32


After (sometimes during) the main gamma ray burst phase lower energy

photons from X – rays to radio are sometimes produced. This latest emission is

known as the afterglow. In must models, electrons are accelerated in relativistic

shocks that are produced by the expansion of a progenitor. The radiation emitted by

these electrons is the afterglow. This emission appears from radio to X – ray

wavelengths and is observable from minutes to weeks after the event (Paczyniski &

Rhoads 1993, Katz 1994, Yost et al. 2003). It was not known that GRBs have any

counterparts in other wavelengths until 1997. The first GRB afterglow was detected

in X – rays from GRB 970228 by BeppoSAX on Feb. 28, 1997 (Costa et al. 1997).

The determination of the exact position by BeppoSAX led to identify the optical

emission (Van Paradijs et al. 1997) from the same effect. The first radio afterglow

was soon detected in another burst event, in GRB 970508 (Frail et al. 1997).

Afterglow observations increased by time and they also led to the host galaxy

identification and redshift measurements for many events.

(a) The X – Ray Afterglows: After the main event in gamma rays, the first

afterglow signal comes in the X – ray band. This signal shortest and strongest among

the afterglows – may begin while main GRB event is continuing. Because the late

part of the prompt emission and light curve of the burst (observed several hours

after) are coincident. The fluxes of X – ray afterglow can be expressed as; αβ

ν αν −− ttf )( (3.6)

where t is the elapsed time, since the beginning of the burst α is ~1.4 and β is ~ 0.9.

This flux is also proportional to (1 + z)β−α with the cancellation of k correction and

the temporal decay. Piran et al. (2001), 11 hours after the burst find a flux of is 5 ×

10−13 erg cm−2 s in 1 – 10 keV range using 21 BeppoSAX bursts. De Pasquale et al.

2002 found comparable results with the larger sample. They also find the X – ray

emission is brighter if GRBs have an optical counterpart. X – ray afterglow provides

a few percent of the GRB energy in general. When the beaming factor is taken into

account to determine the luminosity (Lx = fb Lx,iso), the opening angle and X – ray

luminosity are correlated (Berger et al. 2003). Some of the GRBs observed X – ray

counterparts as follows (Piran 2004); GRB 970508 (Piro et al. 1999), GRB 970828

33


(Yoshida et al. 1999), GRB 990705 (Amati et al. 2000), GRB 991216 (Pino et al.

2000), GRB 001025a (Watson et al. 2002), GRB 000214 (Antonelli et al. 2000), and

GRB 011211 (Reeves et al. 2002).

(b) Optical and IR Afterglows: ~50% of GRBs that were well localized by X – ray

observations are observed also by optical and/or IR afterglows. The optically

observed afterglows are of around 19 – 20 mag about one day after the GRB. The

signal decay and observed optical spectrum shows a power law decay as t−α (α ≈ 1.2

and show large variations around this value) and by frequency a ν−β respectively. The

redshift of the GRBs are measured with the host galaxy emission lines.

As was seen in GRB 990510 (Harrison et al. 1999, Stanek et al. 1999) many

afterglow light curves show achromatic break with the phenomenological formula

)(

*)(

**21211 )/]()/(exp[1{)/()( ααααα

νν−−− −−= ttttttftF . (3.7)

This break shows a steeper decline with α ≈ 2 and is interpreted as a jet break.

Opening angle (Rhoads 1999) and viewing angle of the jet can be estimated by the

use of this jet break formula (3.7) (Piran 2004).

Up to now, too many optical afterglows were detected. First optical afterglow

was detected from GRB 970228. Several GRBs showed red bumps that are

interpreted as evidence for a SN association. Recently, some bursts such as GRB

060218 (SN 2006aj) and GRB 030329 (SN 2003dh) have also provided evidence for

a most remarkable “during the burst” supernova signatures.

Radio Afterglow; ~%50 of the well localized bursts were detected to show 8 GHz

radio afterglow observations. The observed peak fluxes of these events were about 2

mJy. An upper limit for the undetected bursts is at the order of 0.1 mJy and 0.2 mJy

with a turn-over around. These bursts were almost all observed in X- ray band and

therefore they were first localized by X – ray observations. Rarely, radio – afterglow

bursts can be “dark” bursts (i.e., no optical counterpart) but ~ 80% of the optically

observed bursts can also have radio counterpart (Piran 2004).

34


GRB 011030 is the earliest GRB that showed a radio afterglow 0.8 days after

the initial burst event (Taylor et al. 2001). In some cases (GRBs 980329, 990123,

000926, 001018, 010222, 011030, 011121) the radio afterglows were detected at

around 1 day that is because usually bursts cannot be detected in radio band until

about 24 hours after the burst. A reverse shock and transition from the reverse shock

to a forward shock were also observed in the early phases of GRBs 990123, 990506,

991216, 980329, and 020405.

There are cases with peculiar structures, such as the radio light curve of GRB

970508 (Figure 3.6) which showed early strong fluctuations was interpreted to arise

from the scintillations and strong to weak scintillation transition decreasing the

amplitude of these fluctuations. The estimated ~ 1017 cm size of this burst at ~4

weeks after the burst provided us with the first direct observational proof of

relativistic expansion process which occurs in most (Frail et al. 1997, Piran 2004).

When the blast wave expands sub-relativistically and quasi-spherically, the

radio afterglow allows an unambiguous calorimetry for the blast wave together with

a long-lived nature. Radio light curves evolve on larger timescales. Detection of the

relativistic Newtonian transition of a spherical expansion in GRBs provide a

“calorimetric” estimation of the total energy of the ejecta (Waxman et al. 1998, Piran

2004).

35


Figure 3.6 Radio afterglow light curve of GRB 970508 with the fireball prediction comparisons (Frail et al. 2001, Piran 2004).

3.3 Spectroscopic Observations and Analysis Technique for the GRB Afterglows

Spectroscopic observations of optical afterglows require special techniques

developed and dedicated to such observations. We used the IDL (Interactive Data

Language) packages that were written by experts who work at SAO-RAS for

spectroscopic data reduction of GRB afterglows. IDL is a programming language

that is widely used in Space Science. It was developed in 1970s at the Laboratory for

Atmospheric and Space Physics (LASP) at the University of Colorado at Boulder

(http://en.wikipedia.org/wiki/IDL_programming_language). IDL packages, which

we used, were created for the long slit spectral observations and the follow up data

processing. Below, we will describe the observations and data reduction for the event

XRF 080109 / SN 2008D.

X – Ray Flash event XRF 080109 was detected during the follow up

observations of SN 2007uy by Swift satellite on Feb. 9, 2008 UT. The flash event

was not detected in the previous observations of SN 2007uy. Its position was in the

outer parts of NGC 2770 which is also the host galaxy of SN 2007uy (GCN 7159).

According to XRF spectra and light curve, the object was similar to some other

XRF/GRB events connected with supernovae. The flashing object was named as

XRF 080109 / SN 2008D.

We used the SAO – RAS 6m BTA telescope for the early spectroscopy of the

XRF 080109 / SN 2008D event. A combination of the transparent grism VPHG550G

of the operational spectral range 3500 – 7500 ÅÅ and resolution 10 Å (at FWHM –

Full Width at Half Maximum) was used as a dispersing element. Observations were

carried out during Jan. 16, 2008 starting at 21:46:22.31 UT.

The raw data were taken using the long slit to analyze the XRF 080109 / SN

2008D spectra using the mentioned IDL programs.

The standard data processing includes the following: Firstly it is necessary to

create master frames of a flat (Figure 3.7), a bias (Figure 3.8), and a comparison

spectrum by a Ne-Ar lamp (Figure 3.9). The 'spec_calib_pipeline' program in IDL

(Mbias.fits, Mflat.fits, Marc.fits) was used to create spectral calibration frames . The

http://en.wikipedia.org/wiki/IDL_programming_language

36


next step of processing is to extract the spectral calibration frames “Mbias” and Mflat

images from the object spectra using the routine 'spec_reduc'. The cleaned object

spectra labeled as 'red_obj1.fits' is shown in Figure 3.10.

Figure 3.7 The sample image of a 'flat' image

Figure 3.8 Raw data for the bias

Figure 3.9 A sample of the unprocessed arc image

37


Figure 3.10 Master flat and master bias subtracted image of the object

Now it is necessary to apply the wavelength calibration to our cleaned

spectra. This calibration is done in three steps. First step is to use 'corr_ethalon.pro' to correlate, rows of arc image with the ethalon vectors. This routine was used to

create the 'guess values' for the lines. The next step is to compute the positions of arc

lines (Fig 3.9) making use of the guess values using the routine 'ident_lines.pro' .After that 'disp_relation.pro' is used to compute the wavelength calibration

parameters. Finally the rectification of frame and calibration in wavelength is made

by 'rectify_frame.pro'. In the next step of our data reduction, the peak positions of the lines are

computed by comparing them with 'night_sky_lines.dat' . Here we try to find the best

fit with night sky lines. This is followed by the background subtraction and fitting by

using the input 2D-frame. For these calculations the routine 'skyfit.pro' is used.

If the frames are found to be shifted along the slit, it is necessary to apply

'pair_subtraction.pro' to compute a pair subtracted image. The resulting image is

shown in Figure 3.11. Finally the 2D spectrum is reduced to a 1D (one dimentional)

spectrum by use of 'extract_spec.pro'. If we have more than one spectra we can sum

these spectra using the routine 'sum_spec.pro'.

38


Figure 3.11 The image of the object with the subtracted pair. The first black line is also shown GRB spectra. It is processed with the negative version of the program Up to here, all steps can be applied also for the standard star reduction. After

this step it is necessary to determine a response function for flux calibration using the

standard star which was selected from Oke catalog. Finally, flux calibration and

extinction corrections are applied to the 1D spectrum given in Fig 3.12 (one

dimensional). For details see APPENDIX 1.

Figure 3.12 The resulting summed 1D spectrum of XRF 080109

3.4 The Code 'SYNOW' for Spectral Interpretation

39


The multiparametric code 'SYNOW' was repeatedly used for direct analysis of

the spectra of massive SNe when hydrogen showed up at the phase of radioactive

heating of the expanding envelope (Emlhamdi et al. 2006, Branch et al. 2002, Baron

et al. 2005). The algorithm of the computation of model spectra was based on the

assumptions of: spherical symmetry; homological expansion of layers, and having a

sharp photospheric boundary.

This code can be used to identify the lines, estimate the expansion velocity of

the photosphere and the interval of the velocities for the lines of each ion found in

the spectrum (Branch et al. 2001, Elmhamdi et al. 2006).

According to the shapes of the observed spectral features, the following two

cases are possible: the case in which the atmospheric layers where the spectral line

forms do not detach from the expanding photosphere (“undetached case”) and the

case in which these layers are detached from the expanding photosphere (“detached

case”). If the matter of the moving gaseous layers is located above the photosphere

(i.e. , all layers starting from the photospheric level radiate and block the radiation

from photosphere) then a line is observed both in absorption and emission as in case

of the classical P-Cyg type line. This is the case considered in this work (Figure

3.13).

40


Figure 3.13 The line profiles corresponding to the cases of the envelope layers undetached from the expanding photosphere and the layers detached from the photosphere (i.e., those located above the photosphere) when the gas expansion velocity increases proportionally to the distance from the center (V α r). The hatched regions are those that form the absorption component of the P-Cyg profile.

The parameters that are used in the SYNOW code finished through the

“in.dat” file. The format of “in.dat” file is shown in Figure 3.14.

The meanings of parameters of “in.dat” file are as follows;

• 'vphot' is the velocity of the photosphere in km/s

• 'vmax' is the upper limit of velocities in the model

• 'tbb' is the blackbody temperature of the photosphere in oK degrees.

• 'ea' is the lowest wavelength of the synthetic spectrum, in Angstroms

• 'eb' is the highest wavelength of the spectrum in Angstroms

• 'nlam' is the number of wavelength points where the spectrum is computed .

• 'taumin' is the selected minimum optical depth.

• The actually starting place of the computing spectrum. That should have

higher value than ea.

• 'pwrlaw' if this parameter is set to .true. optical depth of lines will thread

according to power law with pwrlawin.

• Pwrlawin is the index that optical depth of lines fall off like a powerlaw

• numref is the number of reference lines used in modeling

• an is the atomic number of all ions considered in the model

• ai are ionization stages of all ions considered in the model

• tau1 are corresponding optical depth of line formation of every ion

• vmine are the least velocities over the photosphere for every ion (in 1000

km/s)

• vmaxe is the highest velocity in the envelope of each ions in the 1000 km/s

units

• ve are the characteristic velocities (υe) in the used law τ (r) ~ exp (-(υ(r) / υe))

of relation between optical depth of lines of a given ion τ and υ, where υ ~ r.

• temp is the excitation temperature of the ion that is determined by assuming

Boltzmann excitation in 1000s of K.

41


Figure 3.14 The format of “in.dat” file

When you run the code “fort.11” output file will be created by SYNOW each

time it runs. You can draw the first two column of “fort.11” to see the shape of

synthetic spectrum using some graphical programs like xmgrace. The last column

in “fort.11” just shows the blackbody spectrum of the photosphere.

3.5. PEGASE

42


In this work we also used the PEGASE program to make modeling of GRB

host galaxies using the optical and infrared observations of GRB afterglows.

Detailed work and explanations about PEGASE are given in section 4.3.

High quality spectroscopic galaxy surveys are widely used to interpret the

galaxy evolution models either synthetic spectral distribution (SED) or

spectrophotometric indices and colors are fitted to observed spectrum to explore

chemical evolution of a stellar population and history of star formation. Widely

used two indices of Rose et al. (1994) and Lick indices (Worthey 1994,

Kuntschner & Davies 1998, Trager et al. 2000) are sensitive to the age or

metallicity of a stellar population and describe the strengths of spectral features.

The evolution of these indices is predicted by various models that are based on

stellar evolution (evolutionary tracks) and stellar formation history (initial mass

function, star formation rate…).

PEGASE program is used for the study of galaxies by evolutionary synthesis.

It generates high resolution spectra in the range of λλ = 400 – 680 nm. The

metallicity age and kinematics of the galaxy can then be explored by this code.

4. INTERPRETATION OF OBSERVATIONS AND RESULTS Eda SONBAŞ

43

4. INTERPRETATION OF OBSERVATIONS AND RESULTS

4.1 Observations of The Stellar – Wind Envelope Around The Massive

Supernova Progenitor XRF / GRB 060218 / SN 2006aj

4.1.1 Review of Earlier Observations

On February 18.149, 2006 Swift space observatory had recorded a peculiar

GRB event accompanied by a powerful component of the radiation from a supernova

(SN) in the observed spectra and and in its light curve. Therefore this burst was

classified as GRB 060218 / SN 2006aj indicating both a GRB and a SN. However,

the spectrum of the GRB observed was dominated by an X – ray radiation and

therefore the GRB was also classified as an X – Ray Flash (XRF). Afterwards, the

entire event was referred as XRF / GRB 060218 / SN 2006aj or XRF 060218 / SN

2006aj. We will use the latter designation most often in order to emphasize the fact

that, in this case the supernova event (SN 2006aj) began with a powerful X – ray

flash.

This XRF / GRB combination was the first event with direct observational

evidence available for the early phase of the expansion of the shock, developed as a

result of the explosion of the compact core of the supernova, and the breakout of the

same shock toward the outer boundary of the stellar-wind envelope of the supernova

progenitor. The shock was first observed in the form of the X – ray flash XRF

060218 during the first two hours. After that, when the envelope became optically

thin, the shock took the form of a powerful ultraviolet flash with maximum at the 11

hours after the detection of the GRB (Blustin 2007, Campana et al. 2006). During the

first 2800 seconds the X – ray spectrum of XRF 060218 exhibited, in addition to the

non-thermal emission typical to the afterglows of GRBs, a powerful component with

decreasing temperature and the maximum of emission gradually shifts into the UV –

optical part of the spectrum. This burst is classified as long GRB since 90% of peak

intensity level (T90) was reached at according to T90 = 2100 ± 100 s (Figure 4.1,

Campana et al. 2006). The bright source found by XRT at Swift is peaked at 985 ±


44

15 s with 100 counts/s in the 0.3 – 10 keV range. The X – ray flux behavior then

showed that it is a typical GRB afterglow. The Swift Ultraviolet/Optical Telescope

(UVOT) has also found a brightening emission after its first detection by factors up

to 5 to 10. The light curves has reached a minimum at around 200 ks (1ks = 103 s)

then it had shown a re-brightening at 700 – 800 ks interval in the optical band

(Figure 4.2).

Figure 4.1 The early afterglow light curve of XRF 060218 as observed by Swift. XRT (X – Ray Telescope) and BAT (Burst Alert Telescope) on Swift . The open squares (green) showed the BAT light curve. Observed count rate for each BAT point were converted to flux in the 15 – 150 keV band using the observed spectra. Combined BAT and XRT spectra were fit with cut-off power-law plus a blackbody. The filled circles (red) show a V-band light curve. The V flux has been multiplied by a factor of 100 for clarity. Gaps in the light curve are due to the automated periodic change of filters during the first observation of the GRB (Campana et al. 2006).


45

Figure 4.2 XRT and UVOT (Ultra Violet/Optical Telescope) light curves of

GRB060218 by Swift observations. Open black circles represent the XRT light curve (0.3 – 10 keV) and has a long, slow power law rise followed by an exponential (or steep) decay. The breaks in the light curve at about 10 ks with a power law index of 1.2 ± 0.1 to a shallower power law decay is the characteristics of typical GRB afterglows. The UVOT light curve is shown in lower panel. Different UVOT filters are given by different colors: Red : V (centered at 544 nm), green : B (439 nm), blue : U (345 nm), light blue : UVW1 (251 nm), magenta : UVW1 (217 nm) and yellow : UVW2 (188 nm) (Campana et al. 2006).

XRF / GRB 060218 is one of the nearest GRBs with a redshift of z = 0.0331.

In this respect, it can be compared with GRB 030329 / SN 2003dh, which was also

identified with a type Ic SN with a redshift z = 0.1683. Both events came to the focus

of attention, since such GRB and SN coincidences are by no means common: – They

usually occur once in 2 – 3 years (Chapman et al. 2007). The study of GRB can be

said to constitute a new phase of the study of the same massive SN, from the start of

event. Therefore the reported early spectroscopic observations are of great

importance for understanding the mechanism of the explosions of both the massive

SN itself and the GRB together with the mechanism of the formation of the burst

sources. The large collecting area of the 6-m telescope (BTA) of the Special


46

Astrophysical observatory (SAO) of the Russian Academy of Sciences (RAS) and its

most eastern location when compared to other major European observatories is of

great importance in the international program of follow up spectroscopic

observations and monitoring of the rapidly decaying optical afterglows of GRBs.

In this work we report the spectra (Figure 4.3) taken with the BTA

(Fatkhullin et al. 2006). Like in case of GRB 030329 / SN 2003dh (Sokolov et al.

2003), the spectra of XRF 060218 / SN 2006aj were among the earliest such data

taken with a high signal-to-noise ratio. Table 4.1 lists the data on quality spectra. The

high quality of the spectra allowed us to see the methods of analysis and

interpretation commonly applied to the spectra of SN Ic (Branch et al. 2001,

Elmhamdi et al. 2006).

Figure 4.3 Observed spectra of the afterglow of XRF 060218 / SN 2006aj

taken with BTA the 6-m telescope of SAO (Fatkhullin et al. 2006). The UT times and the time elapsed since the start of observations after the SN explosion is indicated with black (2.55 d after explosion) and red lines (3.55 d after explosion). Note also the emission lines of the host galaxy at z = 0.0331.


47

Table 4.1: Log of early spectra of the event XRF 060218 / SN 2006aj taken prior to February 23, 2006 with different telescopes. We indicate the time when the spectrum was taken after the XRF / GRB 060218 flash. Here Tsp is the number of days elapsed since February 18.149, 2006. We list only the spectra with high signal-to-noise ratio. The table does not include the early spectra taken by Modjaz et al. (2006) with the FLWO (Fred Lawrence Whipple Observatory) 1.5-m telescope 3.97 days after the burst due to low signal-to-noise ratio.

Campana et al. (2006) showed the X – ray flash itself, and the UV flash that

occurred about 10-11 hours after XRF 060218 and UV excess in the early spectra of

the afterglow can all be explained by the interaction of the shock from the supernova

with the stellar-wind envelope around the massive progenitor star. This is the so-

called “shock breakout” effect – the breakout of the shock through the envelope that

surrounds the collapsing core of the star.

This effect has been long known for massive SNe of different types; i.e.,

core-collapse SNe Ib/c and SNe II (Colgate 1968, Imshennik & Nadezhin 1989,

Calzavara & Matzner 2004), and it can be observed as a relatively short phase of the

SN explosion, which begins with an XRF and ends with a bright UV flash, indicating

that the shock has reached the surface of the exploding star as shown in Figure 4.2.

This process can also be viewed as the shock reaching the outer layers of the

optically thin “surface” of the stellar-wind envelope that surrounds the collapsing

core of SNe Ib and Ic. In cases of the famous SN 1993J and SN 1987A as well as in

case of XRF 060218 / SN 2006aj, this effect was observed as the early abrupt and

short peak on the optical light curve that is shown in Figure 4.4. This effect was


48

already mentioned in the very first afterglow spectra of GRB 030329 / SN 2003dh

taken 10 hours after GRB 030329 (Sokolov et al. 2003). In case of XRF 060218 / SN

2006aj the detection of the GRB at the beginning of the SN explosion, i.e., before the

shock breakout, allowed its motion in the interior of the stellar-wind envelope to be

observed during the first two hours as an X – ray flash with a thermal spectrum.

Figure 4.4 Optical (U, B, V, R, I) and Infrared (J) light curves of XRF

060218 / SN 2006aj (Jelinek et al. 2007). The first maximum corresponds to the UV flash – the shock break out. The spectra from Table 4.1 refer to the transition region in the vicinity if the minimum of the light curve. The lower arrows indicate the times when the spectra were taken with the 6-m telescope after the beginning of the SN explosion. T0 (days) corresponds to February 18.149, 2006.

It is evident from Table 4.1 that the spectra taken with the 6-m BTA telescope

correspond to the light curve minimum, i.e., we observe the transition:

• From phase (1) of the thermal radiation associated with shock breakout

to the “surface” of the stellar-wind envelope – the first peak in Figure

4,


49

• To phase (2) of the subsequent rise of the SN flux with a maximum

after 10 days (Figure 4.4), which corresponds to (non-thermal)

radiative heating of the expanding envelope of the SN resulting from

the 56Ni → 56Co → 56Fe decay.

Thus the physical conditions during our observations varied rapidly and our

work dedicated to interpreting the spectra of the transition phase. However, before

we start to compare the observed and theoretical spectra, we give the estimates of the

characteristic energies, temperatures, sizes and velocities, which directly follow from

the results of X – ray and UV observations of the XRF 060218 flash made before the

beginning of the spectroscopic observations with the 6-m BTA telescope.

4.1.2. The Energies of the X – Ray and UV Flashes

Campana et al. (2006) report the light curves in the XRT (0.3 – 10 keV) and

UVOT energy intervals. They estimated the energy radiated via gamma and X – rays

during the first two hours, when the shock breaks out through the wind. The X – ray

flash energy released in this process is of order 6 × 1049 ergs. During the next 8 – 11

hours the UV light curve shows a powerful peak – the first maximum in Figure 4.4,

which is associated with the shock breakout to the surface, or, more precisely, to the

outer boundary of the stellar-wind envelope, which becomes sufficiently transparent.

The UV flash is actually what is called the Colgate “shock breakout” (Colgate 1968).

The data reported by Campana et al. (2006) imply an estimate of 3 × 1049 ergs for the

energy released during about 28 hours in this UV flash within the Swift / UVOT

energy interval, thereby providing direct evidence suggesting that the X – ray (X –

ray radiation may explained with Bremsstrahlung in the early phase and the inverse

Compton scattering in the later phases) and UV flash are of the same nature.

4.1.3. Evolution of Temperature


50

At first, when the shock breaks through the wind envelope and energy is

released mostly in X – ray interval, the thermal spectrum corresponds fairly well to

temperatures of kT ≈ 0.17 keV, which are equivalent to 2 × 106 K. By the time of the

UV flash (11 hours after the GRB ) – the first maximum in Figure 4.4 – the

temperature decreases down to 0.03 – 0.05 keV corresponds to 350 – 850 thousand K

or even lower given the uncertainties of kT measurements made by Swift / UVOT at

that time (the UVOT energy interval covers only a part of the blackbody flux

(Blustin 2007)). After the next 22.5 hours the temperature of the thermal component

decreases down to 43000 K or to 9.19.07.3 +

−≈kT eV is shown Figure 4.5 ( Campana

2006). According to the estimates of the temperature decrease reported, the

temperature may be even lower than 10000 K by the time of our first spectroscopic

observations (2.55 days after the GRB, see Table 4.1).

4.1.4. The Size of The Stellar-Wind Envelope

The data reported by Campana et al. (2006) can also be used to estimate the

radius of the wind envelope that surrounded the Wolf-Rayet star before the

explosion. Blustin (2007) provides arguments in favor of a Wolf-Rayet star

surrounded by a dense stellar-wind envelope as the progenitor of XRF 060218 / SN

2006aj. The size of this envelope can be naturally associated with the bright UV flash

observed 11 hours after the GRB. At that time the shock, which was until then

observed only at X – ray energies, also begins to show up at optical wavelengths,

because the layers of the wind envelope above the shock become optically thin and

the shock breaks out to the “surface” (more precisely to the upper layers) of this

extended envelope associated with the progenitor star / pre SN. According to Swift /

UVOT data on the evolution of the temperature and radius of the thermal component

of the GRB / XRF 060218 afterglow, i.e., at the temperature of kT ~ (0.03 – 0.05)

keV at the time of maximum brightness (~ 104 s), the size of the wind envelope

around the pre-SN must be equal to 300 Rּס for a bolometric luminosity of (4.6 –

35.5) × 1045 erg s−1. The radius can be determined from;


51

3/1

4

≈

aTER iso

shell (4.1)

where Eiso is the equivalent isotropic power of the GRB, a is the radiation density

constant and T is the thermal temperature. The mass loss rate through the wind is

determined with the following formula the:

≈

shell

windshell

RM

Mυ& (4.2)

where υwind ~108 cm s−1 is the typical value for Wolf – Rayet stars. υshock speed is also

determined with the assumed value of ρwind ~1012 g cm−3 from: 24 3 shockwindBBaT υρ≈ (4.3)

Figure 4.5 The temperature and radius evolution of thermal soft component

of GRB 060218. Upper panel shows the temperature of the soft thermal component evolution. The joint BAT and XRT spectrum has been fitted with a blackbody component plus a power law in the first ~ 3000 s. The last green point shows the UVOT data. In the lower panel the radius of the soft thermal component evolution


52

was seen and the last green point comes from the fitting UVOT data (Campana et al. 2006).

4.1.5. Shock Velocity

While the shock moves inside the stellar-wind envelope, the radius

corresponding to the thermal component of radiation (associated with the shock

breaking out from the center) increases steadily from about 5.7 Rּס to 17 Rּס, as

implied by the same data of Swift / UVOT X – ray observations made before the

maximum of the UV flash, as reported by Campana et al. (2006). The radius of the

thermal component continues to increase up to 4700 Rּס in 1.4 d (see Figure 4.5), i.e.,

by the time of Swift/UVOT observations are made after the maximum of the UV

flash. By that time the shock has already broken out to the “surface” of the stellar-

wind envelope and the luminosity of the thermal component decreased (see Figure

4.4). We can divide the path traversed by the shock (of the radius 94.093.029.3 +

− x 1014 cm

of the thermal component at kT ≈ 9.19.07.3 +

− eV), by the time (1.4 days) to estimate the

expansion velocity of the photosphere associated with the shock by the time of

spectroscopic observations: (2.7 ± 0.8) x 109 cm s−1 (Blustin 2007). Such a velocity

is typical for massive SNe and it is comparable to the widths of the lines observed in

the spectrum of SN 2006aj (Pian et al. 2006).

In the following sections, we will show that analyses of our optical spectra of

XRF 060218 / SN 2006aj taken 2.55 and 3.55 days after the start of the SN explosion

confirm the above estimates of energies, temperatures, sizes, and velocities based on

the Swift / XRT / UVOT results of observations of shock breakout. At that stage the

contribution of the blackbody component of the shock radiation in these spectra

remained dominant as it is indicated by strong blue excesses in our spectra (Figure

4.3). However, as it is evident from the UBVR light curves (Figure 4.4), the GRB

afterglow becomes appreciably redder in five days, and this is due to the change of

the radiation mechanisms of the SN shell begins, which has been described

repeatedly in many reviews (see e.g., Imshennik & Nadezhin 1988).


53

In section 4.1.6 we will describe early spectroscopic observations of the

optical afterglow of XRF 060218 / SN 2006aj made with the 6-m BAT telescope of

the Special Astrophysical Observatory (SAO) of Russian Academy of Sciences

(RAS). In section 4.1.7 we will describe the modeling of the spectra using SYNOW

code and discuss the spectroscopic manifestations of the wind envelope in the form

of broad absorptions near the rest frame wavelengths of 5800 Å and 6100 Å, which

corresponds to absorption in the Hα line. In section 4.1.8 we will discuss:

• The apparent evolution of the signs of the Hα line according to the data for

XRF 060218 / SN 2006aj obtained with the 6-m and other telescopes,

• The manifestations of the Colgate “shock breakout” in the light curves,

spectra, sizes, and luminosities of the “common” SNe Ib/c,

• The asymmetry of the explosions of massive SNe.

4.1.6. Early Spectroscopic Observations of the Optical Afterglow of XRF 060218

/ SN 2006aj Made With the 6-m Telescope

The program of the special observations of the optical afterglow with the 6-m

telescope included the spectroscopy of the variable object discovered by Swift /

UVOT telescope and identified with the XRF 060218 event. We used the SCORPIO

focal reducer mounted in the primary focus of the BTA telescope

(http://www.sao.ru/hq/lsfvo/devices/scorpio/scorpio.html). Observations were made

in two sets: February 20, 2006 and February 21, 2006. As the dispersive element we

used a combination of a transparent grid and a VPHG550G prism with an operating

wavelength interval 3500 – 7500 Å and a resolution (FWHM) of 10Å.

We reduced our observations using a standard technique, which included bias

subtraction; flat fielding; wavelength calibration using the comparison spectrum of a

Ne-Ar lamp; atmospheric extinction correction, and absolute flux calibration using

observations of a photometric standard during every observing night. We put a bright

starlike object onto the entrance slit along with the optical transient in order to reveal

the possible short-term variability and perform night-to-night control of the absolute-

flux calibration (see Figure 4.6). We made a total of four exposures during each of

http://www.sao.ru/hq/lsfvo/devices/scorpio/scorpio.html


54

nights of 2.55 d and 3.55 d after the explosion. We found no short term variations

during our observations and therefore added all the four individual spectra in order to

improve the signal-to-noise ratio. Figure 4.3 shows the averaged spectra for each of

nights and Table 4.1 gives the average epochs for these spectra.

Then we corrected the resulting spectra (Figure 4.3) with the Galactic

extinction in accordance with the dust distribution maps (Schlegel et al. 1998) shown

in Figures 4.7 and 4.8.

The E(B-V) color excess toward the object (α2000 = 03h21m39s.683, δ2000 =

+16o52´01´´.82) is equal to 0.14. we computed the extinction in terms of the dust-

screen model in accordance with the formula Fint (λ) = Fobs(λ)100.4k(λ)E(B-V), where

Fint (λ) and Fobs(λ) are emitted (unabsorbed) and observed fluxes, respectively. We

adopted the Milky-Way extinction curve, k(λ), from Cardelli et al. (1989).

Figure 4.6 Field of the optical transient. The beginning of spectroscopic

observations with the 6-m telescope: February 20.647, 2006 (about 60 hours after the GRB); P.A. = 3o (the position angle of slit); VOT = 18.16, and (B – R)OT = 0.3

4.1.7. Comparison of the Observed Spectra of XRF 060218 / SN 2006aj With

Synthetic Spectra

To interpret the spectra obtained, we used the “SYNOW” code for computing

synthetic spectra (Branch et al. 2001). We can identify this optically thick


55

photosphere, which radiates a blackbody spectrum, with a shock when interpreting

the early spectra. This is true at least for the first days (Tsp = 1.95 and 2.55 d from

Table 4.1) after XRF 060218, when both the spectra (Figure 4.3) and photometry

(Figure 4) exhibit strong blue excess, which is due to the blackbody radiation

component from the first, short and powerful UV flash. Spectral lines are assumed to

form above this expanding photosphere as a result of resonance scattering, which

SYNOW codes treat in the Sobolev approximation (Sobolev 1958).

To understand this result of ours, it is important that, in SYNOW code, the

expansion velocity of the layers located above the photosphere is proportional to the

distance from each point of the expanding layer to the center (V α r).

We choose this model where layers are not detached from the photosphere to

interpret our first spectrum (Tsp = 2.55 days, see Figure 4.3). Moreover, at this time

the expanding photosphere can be identified with the shock, because our first

spectrum (with greater excess in its blue part) was taken closer to the time of the

shock breakout to the “surface” of the stellar-wind envelope, surrounding pre-SN –

the first maximum in Figure 4.4.

A layer that has been detected from the photosphere shows up in the spectrum

as a smoothed emission feature and a strongly blueshifted absorption – the “remnant”

of the P-Cyg profile – as it is shown in Figure 3.13. The atmosphere detaches when

the velocity of the layer where spectral line forms exceeds appreciably the velocity of

the photosphere. If V α r, then the absorption part of the P-Cyg profile is a result of

the shielding of the photosphere by a narrow layer, and the emission feature is

weaker with decreasing of the thickness of the layer that emits the line. Such a model

corresponds to our second spectrum (Tsp = 3.55 d), at least for HI lines, which may

form in the wind envelope. It is safe to assume that the shock has set into motion and

“detached” the uppermost layers of this extended wind envelope that surrounded the

progenitor star of the GRB / SN.

At this phase, 3.55 d after SN explosion, when the effect of the blackbody

component becomes weaker (Figure 4.4) and the expanding envelope of the SN

becomes increasingly more transparent, the shock cannot be fully identified with the

photosphere. At this time the so-called photospheric phase of the SN explosions


56

begins and its spectra were the prime object for SYNOW code to interpret

(Elmhamdi et al. 2006).

Figures 4.7 and 4.8 compare the spectra of XRF 060218 / SN 2006aj taken

with the 6-m telescope 2.55 and 3.55 d after the flash with the synthetic spectra

modeled using SYNOW code. The first spectrum taken immediately after the bright

UV flash (Figure 4.4) can be easily modeled with a rather simple set of parameters.

We set the temperature of the photosphere (tbb) used to fit the spectral energy

distribution in Figure 4.7 (with restframe wavelengths) equal to 9000 K. This also

agrees with the temperature decrease rate according to the results of Swift / XRT /

UVOT observations (Campana et al. 2006); according to these authors, the

temperature must have been lower than 10000 K by the beginning of our

spectroscopic observations. The SYNOW model for the spectrum taken 2.55 d after

the flash implies the velocity of the photosphere to be around 33000 km s−1 . This

parameter also lies within the quoted errors of the expansion velocity of the

photosphere associated with the shock ((2.7±0.8) × 104 km s−1) estimated using the

same Swift / XRT / UVOT data.

This means that by the beginning of our spectroscopic observations, about

one day after the last Swift / UVOT observations of the decaying UV flash, the shock

propagation velocity remained within the same limits. Moreover, the broad and

therefore barely visible depression in the continuum (see Figure 4.3) with the

observed wavelengths at 5900 – 6300 Å, and the barely visible flux excess in the

wavelength interval 6300 – 6900 Å can be best fitted by a broad P-Cyg profile of the

Hα line (see Figure 4.7) with the same velocity of 33000 km s−1 . In the SYNOW

code this corresponds to the undetached atmosphere case (Figure 3.13).

It is safe to assume that, at this time, a part of the stellar-wind envelope

located above the photosphere on an extended layer above it, move together with the

photosphere and we observe the acceleration or breakout of the upper layers of this

envelope. This is yet another result of the shock breakout at the “surface” of the

massive circumstellar envelope that is (almost) unmoving before the SN explosion.

According to the estimates given in the beginning of this section, (section 4.1.3) at

least at the time of the shock breakout (11 hours after GRB) this envelope had the


57

size of about 300 Rּס. To demonstrate that barely visible flux variations in Figure 4.3

in the wavelength interval 5900 – 6900 Å of the observed spectrum can indeed be a

very broad P-Cyg profile of the Hα line for velocities of about 0.3 × 1010 cm s−1, we

show in Figure 4.7, a model spectrum with a classical P-Cyg profile, but for a much

lower velocity, equal to 8000 km s−1 (in the Figure, the laboratory wavelength of the

Hα is also marked by a narrow emission of the host galaxy).

We fitted the undetached case to various synthetic spectra, but the velocity of

photosphere (Vphot) is the same for all elements and their ions (vmine) and equal to

33000 km s-1. The main computation parameters are listed in Table 4.2 for the

“undetached case” of Figure 4.7. The only difference is in the effective formation

depths of ion lines (τ / tau1), which can be chosen relatively small (τ < 0.5) for all

elements and ions to satisfactorily describe the observed spectrum. In the wavelength

interval of about 4500 – 7200 Å we find hydrogen to have the highest formation

depth τ = 0.2, whereas the other elements and their ions in the synthetic spectra are

taken into account even at lower τ. This fact may be indicative of small changes in

the relative abundances of other elements compared to that of the most abundant

element. This may be due to the fact that the contribution of the stellar-wind

envelope remains dominant in the first spectrum on Tsp = 2.55 days.

Note that we achieved the best agreement between the observed and synthetic

spectra in the part of the spectrum where observations have the highest signal-to-

noise ratio and the theoretical spectrum passes within the noise band or as close to it

as possible. In particular, the sight flux deficit left of the emission line of the host

galaxy [OII]3727 Å (in Figure 4.7) can be described by the effect of CaII with τ =

0.5 which is the greatest formation depth that we use in this paper.

Our second spectrum which is for Tsp = 3.55 d (see Figure 4.3 with observed

wavelengths) shows an absorption at 6300 Å, which we interpret as a strongly

blueshifted “remnant” of the P-Cyg profile of the Hα line. The best fit of the

synthetic spectrum corresponds to the “detached case” in Figure 3.13, where some

part of the stellar-wind envelope has already “detached” from the photosphere. That

is, the contribution of the thermal flash (Figure 4.4) rapidly decreases, the shock


58

becomes increasingly more transparent and this is the beginning of the photospheric

phase of the expansion of the SN envelope (Branch et al. 2001).

We modeled this second spectrum in the velocity interval (18000 < V <

24000) km s−1 (see Figure 4.8 with the rest frame wavelengths). The closest to the

observed spectrum is the synthetic spectrum with model parameters from Table 4.3,

where the velocity and temperature of the photosphere are equal to 18000 km s−1 and

tbb = 8200 K, respectively. The synthetic spectrum allows for the contribution of

lines with small τ (< 0.3) for the ions HI, HeI, FeII, SiII, OI, CaI, CaII, TiII, NI, CI,

CII, MgI, MgII, and NaI. In Figure 4.8 we presented the positions of some of these

lines or the regions in the spectrum where the ion considered contributes

significantly to the spectrum in case of the adopted model parameters from Table 4.3.

The list of the reference lines used in SYNOW code can be found in Elmhamdi et al.

(2006), and a more detailed information can be found in the SYNOW code itself

(http://www.nhn.ou.edu/~parent/synow.html).

The broad absorption in the region of 6100 Å (Figure 4.8) can be described

by the effect of HI for the “detached case” (Figure 4.7) at τ = 0.16. The synthetic

spectrum in this wavelength interval (5700 – 6300 Å region in Figure 4.8) is

characterized by a smoothed redshifted emission and a strongly blueshifted

absorption with a minimum in the vicinity of 6100 Å – the “remnant” of the P-Cyg

profile of the Hα line.

Thus at Tsp = 3.55 d hydrogen has already detached from the photosphere and

the corresponding layer moves at a velocity of 24000 km s−1 (see Table 4.3). Here we

also included all ions with small τ < 0.3, but the observed spectrum cannot be

described with the same values for all other parameters as it was the case in Table 4.2

(the “undetached case”). The layers where most the lines of other elements form,

move either with the same velocities as the photosphere (18000 km s−1 ; OI, CaI,

CaII, TiII, NI, and NaI lines), or with the same velocity as hydrogen (24000 km s−1;

HI, HeI, FeII, SiII, CI, CII, MgI, and MgII lines). The characteristic velocities

(parameter ve in Table 4.3), which determine the characteristic thickness of the

layers occupied by each element, also have to be chosen differently for different

elements in order to describe best the observed spectrum. It is safe to assume that this

http://www.nhn.ou.edu/~parent/synow.html


59

spectrum is more affected by the chemical composition of the expanding envelope of

the SN. This chemical composition has changed as a result of the evolution and

explosion of the highly evolved massive core of the progenitor star.

Figure 4.7 The rest frame (z = 0) spectrum of the afterglow of XRF / GRB

060218 / SN 2006aj taken with the 6 – m telescope, 2.55 d after the explosion and corrected for Galactic extinction. Shown is its fit for the “undetached case” by synthetic spectra with the velocity of the photosphere (Vphot) equal to 33000 km s-1 for all elements and their ions (vmine) – the smooth lines (which differ only in the blue part of the spectrum at λ < 4000 Å ). The main parameters for the computation of the synthetic spectrum shown by the thick black line are listed in Table 4.2. Here HI indicates the P-Cyg profile of the Hα line at Vphot = 33000 km s−1. As an example of the Hα P-Cyg profile the dashed line shows the model spectrum for the velocity of the photosphere equal to 8000 km s-1.

Table 4.2 The model parameters set for the computations in the case where the layers of the elements and ions are undetached from the photosphere (“the undetached case”). The parameters correspond to the synthetic spectrum shown by the thick black line in Figure 4.7.


60

Figure 4.8 The restframe (z = 0) spectrum of the afterglow of SN 2006aj /

XRF 060218 taken with the 6 – m telescope 3.55 d after the explosion and corrected for Galactic extinction. The synthetic spectra are shown by smooth lines. Shown are the position of the spectral lines of some ions and blends of their lines in the regions of the spectrum where the ion in question contributes significantly to the spectrum for the case of the adopted model parameters. The thick black line shows the synthetic spectrum computed with the parameters from Table 4.3 when the absorption at 6100 Å can be described by the dominant effect of HI for the “detached case”. It is a strongly blueshifted part of the P-Cyg profile of the Hα line for the expansion velocity of the detached HI layer equal to 24000 km s-1.


61

Table 4.3: Parameters of SYNOW model for the case where the layers of some elements in the expanding envelope of the SN have detached from the photosphere (the “detached case” in Figure 3.13). The parameters correspond to one of the synthetic spectra shown in Figure 4.8 by a thick black line (see text).

4.1.8. Results and Discussion

4.1.8.1. Evolution of the Spectra of XRF 060218 / SN 2006aj and Other Massive

SNe

The broad and low-contrast feature in the spectrum of XRF 060218 / SN

2006aj with a minimum at the rest frame wavelength of 6100 Å (at z = 0) can be seen

in all early spectra – both our data and the spectra taken with other telescopes since

February 21.70, 2006 (see Table 4.1). The closest one in time (February 21.93) to our

second spectrum from Table 4.1 is the spectrum taken with the NOT (Nordic Optical

Telescope) telescope (Figure 4.9, Sollerman et al. 2006), which also exhibits a

minimum at the same wavelength near 6100 Å. The same feature can also be seen in

ESO Lick spectrum taken on February 22.159; however the signal noise of this

spectrum is appreciably lower (it is the second spectrum in Figure 1 in Mazzali et

al.(2006)) than those of the two spectra mentioned above. According to the data

shown in Figure 4.10 (Mazzali et al. 2006), the feature evidently evolves beginning

with the very first VLT spectrum and becomes deeper in the spectrum taken on VLT

on February 23.026, i.e., about five days after XRF 060218. Mazzali et al. (2006)

point out that, later spectrum taken after the minimum on the light curve (see Figure


62

4.4) exhibit increasingly strong effect of the ever stronger broad absorption

associated with SiII and located near 6000 Å. However, the narrow minimum at 6100

Å can be seen even on the VLT (Very Large Telescope) spectra taken in March. This

is similar to what can also be seen in the spectra of some SN Ib (Parrent 2007),

where the narrow absorption associated with Hα also shows up and evolves in the

spectra taken at the beginning of a long rise toward the brightness maximum typical

to SN Ib and Ic and similar to the rise shown in Figure 4.4.

Figure 4.9 The flux-calibrated and de-reddened spectra of SN 2006aj obtained

by NOT (Sollerman et al. 2006)

Our spectra mentioned above also confirm the beginning of the evolution of the

spectral feature at 6100 Å, which can then be seen for XRF 060218 / SN 2006aj

according to the data obtained with other telescopes. However, we also believe that

the first spectrum taken on February 20.70 with the 6-m telescope also shows the

same feature – we thus interpret the broad and barely visible depression of the


63

continuum in the wavelength interval 5600 – 6600 Å (5800 – 6800 Å – the observed

wavelengths in Figure 3 for Tsp = 2.55 days) like the P-Cyg profile of the Hα line for

the velocities of 33000 km s−1. The same small deviation of the continuum can also

be seen on the first VLT spectrum taken on February 21.041 (see Figure 4.10) eight

hours after our first spectrum. As for the spectrum taken with MDM 14.5 hours

before our spectrum (see Table 4.1), it also shows a weak deviation of the continuum

at 5700 Å, which can be seen despite the low signal-to-noise ratio (Mirabal et al.

2006). In this case the same broad P-Cyg profile of Hα line with velocities 30000 km

s−1 may compare with the identification suggested by the authors. Thus, with all early

observations from Table 4.1 taken into account, we can conclude that, we indeed

observe the evolution of optical spectra – the transition from the phase of the Colgate

“shock breakout” effect and the associated powerful (thermal) UV flash to the

spectra of the phase of the SN light-curve rise, which corresponds to radioactive

(nonthermal) heating resulting from 56Ni → 56Co → 56Fe decay.


64

Figure 4.10 VLT and LICK spectra ant its synthetic fit with the Monte-Carlo

spectrum synthesis code of SN 2006aj (Mazzali et al. 2006)

This must be the main difference and novelty of our approach toward the

interpretation of early spectra of XRF 060218 / SN 2006aj compared to the approach

employed by Mazzali et al (2006), who identify the VLT / LICK spectra of this SN

and consider the theoretical spectra of radioactive heating. Moreover, even the

authors of SYNOW code (Branch et al. 2001, Elmhamdi et al. 2006, Branch et al.

2002, Baron et al. 2005), who computed their synthetic spectra with signs of the Hα


65

line when analyzing the spectra of common SN Ib-c (“stripped-envelope SNe ”), do

not include a significant contribution of the thermal radiation from the UV flash at all

early phases and, in the cases where such a contribution is taken into account, the

corresponding computations are not performed for such high expansion velocities as

those observed for XRF 060218 / SN 2006aj.

Mazzali et al. (2006) modeled the evolution of all the ESO Lick and ESO VLT

spectra of SN 2006aj (Pian et al. 2006) up to March 10, 2006 (20 days after XRF

060218) using Monte-Carlo method and a refined program of the synthesis of SN

spectra based on the same assumptions as SYNOW code, but with the account taken

of the model distributions of density and temperature in envelope above the

photosphere, and with a radiative transfer with line transitions and electron scattering

(see Branch et al. 2006, Lucy 1999, Mazzali et al. 1993 for details of the method).

All the characteristic features of SN Ic become increasingly stronger in the

spectra before the broad maximum on the light curves of SN 2006aj (about 10 days

after the GRB in Figure 4.4) and after this maximum. The spectra are modeled for

the velocity interval (20000 < V < 30000) km s−1. The strongest features in the

spectra mentioned by the above authors are FeII, TiII, and (at later stages) CaII (<

4500 Å), FeIII and FeII (near 5000 Å), SiII (near 6000 Å), OI (near 7500 Å), and

CaII (near 8000 Å).

We take this entire list into account while interpreting our early spectra using

SYNOW code (see Figure 4.8) in almost the same velocity interval. However, we

also include the contributions from the HI and HeI lines. It is evident from the

computations of Mazalli et al. (2006) that the above authors did not consider the

evident signs of the Hα line for the earliest spectra take into account only the

increasingly stronger effect of SiII near 6000 Å (see Figure 4.10). At the same time,

we pointed out above that, the absorption feature with a minimum at 6100 Å can be

traced in the observed spectra at least up to March 4. It is the most conspicuous

feature near 6000 Å in the early spectrum taken on February 23 throughout the entire

wavelength interval from 5000 Å to 8500 Å, where the Monte-Carlo method reveals

no absorption at 6100 Å.


66

4.1.8.2. Hydrogen Signatures in The Spectra of Massive SNe

Hydrogen is also common in the spectra of SN Ib and Ic (Ib-c). In particular

the SYNOW code mentioned above revealed the signs of hydrogen and evolution of

the blueshifted Hα line when analyzing a series of optical spectra for common

massive SN Ic and Ib. Note that, in this analysis, special attention was given to the

traces of hydrogen in observations of such stripped-envelope SNe (i.e., SN Ib-c),

although, according to the formal definition of these SN types, they have no evident

and immediately apparent hydrogen lines in their optical spectra. SN Ib-c are usually

analyzed in terms of the model of gravitational collapse of massive and stripped-

envelope SN Ib-c, which have lost their envelope before the collapse; carbon-oxygen

cores, and, signs of this envelope must always show up in the spectra of such SNe.

However, hydrogen can be more or less confidently identified only in sufficiently

early spectra of SN Ib-c, as shown by Branch et al. (2001), Baron et al. (2005),

Branch et al. (2006), Elmhamdi et al. (2006), and Parent et al. (2007).

According to a possible competing hypothesis for describing the spectral

feature with a minimum near 6100 Å, it can also be interpreted as the absorption

component of the P-Cyg profile of the CII 6580 Å line (Brach et al. 2006). However,

if this entire extended (about 300 Rּס) stellar-wind envelope has first shown up as an

X-ray flash and then as an UV flash in the afterglow of XRF 060218 / SN 2006aj, at

least the unevolved part of the envelope, (which is at the stellar-wind stage of the

evolution of the massive progenitor star and formed long before the SN explosion)

must be associated with neutral hydrogen. The fact that hydrogen always shows the

highest velocity contrast between the filled HI layer and the photosphere compared

to other elements in the spectra of SN Ib-c, also supports the hypothesis that these

very layers (that are associated with the wind envelope)

are the first to start moving as a result of the SN explosion (Branch et al. 2001,

Elmhamdi et al. 2006), and Branch et al. 2002). We can also estimate the mass of

this part of the envelope that has begun to move (the masses of gas in Hα or in the HI

layer), using the equation from the paper by Elmhamdi et al. (2006) as follows:


67

)()1038.2()(29.3 234

594.093.0 ατ HtvMM d

−Θ

+− ×≅ (4.4)

where the time after the explosion, td, is in days; v4 is the velocity of the layer in the

units of 104 km s−1, and τ (Hα) is the formation depth of the Hα line. In this equation

derived for the Sobolev optical depth in the expanding envelope (Elmhamdi et al.

2006, Castor 1970), we use the velocity 24000 km s−1, which refers only to the

moving part of the HI layer (see the data in Table 4.3). It also corresponds to the

“detached case” in Figure 3.13. As a result the mass of the HI layer is of order of 6 ×

10−4 Mּס, and its distance from the center by that time (3.55 days after the SN

explosion) is no less than 7.36 × 1014 cm.

4.1.8.3. The Colgate “Shock – Breakout” Effect in XRF 060218 / SN 2006aj and

Other SNe – the Light Curves, Spectra, Luminosities, and Sizes

SN Ib and Ic have been observed since long and their most likely progenitors

are believed to be Wolf-Rayet stars surrounded by a more or less dense wind

envelope, which is the result of evolution of massive star. The shock that develops

during the explosion of the evolved core of the star traverses the envelope and

produces a bright and short X-ray and UV flashes, which may last several hours. The

duration of the flash depends on how massive and extended was the wind envelope

that surrounded the progenitor star before the SN explosion. The interaction between

the shock produced by the SN explosion and the envelope (the “shock breakout”

effect also known as the “Colgate effect”) has been predicted long ago (Colgate

1968, Bisnovatyi-Kogan et al. 1975, Blinnikov et al. 2002). In particular, Calzavara

and Matzner (2004) had dedicated their paper to future systematic observations of

this effect. However, before the XRF 060218 / SN 2006aj event, the effect in

question could have been fully observed only for a small number of massive SNe.

Due to the compact size of the blue supergiant (20 – 30 Rּס), a very short

effect of the interaction of the shock was observed (albeit not from the very

beginning) in the famous SN 1987A of Type II (Imshennik and Nazedhin 1988). The

shock interaction effect was observed over a long – two-to-three weeks – time


68

interval in a very extended (of the order of 1015 cm) envelope of the Type I in

supernova SN 1994W – see, e.g., Chugai et al. (2004). In another SN Ib-c the effect

was observed only at its very end before the subsequent rise of the SN flux,

corresponding to the 56Ni → 56Co → 56Fe radioactive heating – this was the case of

SN 1999ex (Stritzinger et al. 2002). Similar “remnants” of this effect can be seen (in

the R band) even for SN 1998bw, usually associated with GRB 980425 (Galama et

al. 1998).

The Case of SN 1993J in M81: One can say that SN 1993J was observed

almost at the very beginning of the explosion. We say “almost”, because in this case

the time of the very beginning of the SN explosion is known to within 12 hours and

not to within several seconds as in case of the XRF / GRB SN. This SN was first

classified by its spectrum as a type II SN because of hydrogen lines appear in its

early spectra (Figure 4.11). However, after a certain time this SN changed its type to

Ib (Flippenko et al. 1993), when hydrogen ceased to be confidently seen in the

spectra. In Figure 4.11 the early spectra of SN 1993J show a strong UV excess,

which is characteristic of the shock interaction, and the smoothed continuum with

almost no lines beyond 5000 Å, which resembles the spectrum of XRF 060218 / SN

2006aj in Figure 4.7, although in case of SN 2006aj the expansion velocities are

significantly higher (see below comments about the asymmetry of explosions of SNe

identified with GRBs). SN 1993J also had an unusual light curve with the luminosity

increased rapidly up to the first maximum, and then abruptly decreased during

several days, and then slowly increases again over the next two weeks – a behavior

similar to that shown in Figure 4.4 for XRF 060218 / SN 2006aj. The light curve of

SN 1993J has been since long modeled by several groups (Nomoto et al. 1993,

Young et al. 1995, Shigeyama et al. 1994). These authors pointed out that such a

behavior of the light curve of SN 1993J can be explained by the interaction of the

shock with an extended (about 300 Rּס) hydrogen envelope of mass 1 Mּס around the

progenitor star. Note that the luminosity achieved at the summit of the first (fast)

maximum, which lasts only about 4 – 5 hours, may be as high as 1045 erg s−1, and a

full energy release of the order of 5 × 1049 erg.


69

Figure 4.11 The earliest optical spectrum of SN 1993J (see text). We adopted

this spectrum from SUSPECT – The Online Supernova Spectrum Database (http://bruford.nhn.ou.edu/suspect/, Richardson et al. 2002).

It is not surprising that approximately same envelopes can also explain the

flash due to shock breakout, in case of XRF / GRB 060218 / SN 2006aj. In this case

the total energy released by the SN is of the same order about, 1049 erg. However, in

case of SN 1993J there was no gamma-ray burst and everything seems to be

explained by the asymmetry of the SN explosion (see section 4.1.8.4).

The fact that GRB / XRF 060218 / SN 2006aj is yet another case where the

Colgate effect could have been observed in pure form and since the very beginning

of the SN explosion can be understood as a hint suggesting that the gamma-ray burst

may be the first signal in gamma rays, which is indicative of the beginning of the

collapse of the massive core, followed by the entire process – the explosion of a

massive SN: the burst may be followed by an X – ray and then a powerful UV flash.

In this case a search for all kinds of manifestations of wind envelopes around

massive progenitor stars in early spectra and in photometry of the XRF / GRB

http://bruford.nhn.ou.edu/suspect/


70

afterglow may become the key point in optical observations of transient sources

associated with XRF / GRB, because the interaction of the SN shock with these

envelopes is the first optical event after the beginning of the core collapse of the

massive core of star, and here we can expect manifestations of new effects associated

with, e.g., the asymmetry of the explosion.

4.1.8.4. Asymmetry of Type Ib and Ic Supernova Explosions

The XRF 060218 / SN 2006aj explosion was indeed a classical XRF event

(Campana et al. 2006, Heise et al. 2004). The fact that in cases of common and

nearby SNe the explosion starts not with a gamma ray burst that can be naturally

explained by the asymmetric, axissymmetric, or bipolar (with the formation of jets)

explosion of massive SNe. One of the popular hypotheses (Soderberg et al. 2005) is

based on the assumption that in case of an XRF type flash the observer is located

outside the cone where for some reasons, the bulk of gamma ray radiation in

concentrated.

The farther is the observer from the axis of the SN explosion, the more X –

ray photons are recorded from the burst: The GRB pass into an X – ray rich GRB

(like GRB 030329) and become XRF (Sokolov et al. 2006). If the SN is observed at

an angle close to 90 degree to the explosion axis, no gamma – ray burst is observed

and only an X – ray flash is seen followed by a powerful UV flash caused by the

interaction of the shock with the envelope surrounding the pro-SN, as it was the case

of SN 1993J.

Thus, if the SN is observed from a direction that is close to the equator of the

explosion (and this is the most likely situation) and if the massive collapsing core of

the star is surrounded by a sufficiently dense stellar wind envelope, then, the effect

considered should be observed only at X – ray and gamma-ray energies. In this case

the contribution of the afterglow of the gamma-ray burst to the light curve of a

“common” SN may be insignificant. One way or another, but it must be significantly

smaller than in case of a classical GRB observed close to the axis of the SN

explosion (the most likely situation). In this connection, Filippenko et al. (2006)


71

recently pointed out that a significantly asymmetric explosion may be a characteristic

feature of massive SNe of all types, although it is still unclear whether the

mechanism that generates the gamma ray burst is responsible for the explosion of the

star.

4.2. SYNOW Modeling of XRF 080109 / SN 2008D

On January 9.57, 2008, X – Ray Telescope (XRT) on board Swift detected a

weak XRF with a SN (SN 2008D) radiation component in the galaxy NGC 2770

(with a distance of 27 Mpc). Type Ib-c core collapse supernova SN 2008D is

identified with the bright X – ray transient which is attributed to the shock breakout

emission. Therefore the event was named as XRF 080109 / SN 2008D. The shock-

breakout effect was observed as the early abrupt and short peak on the optical light

curve (see Figure 4.12, Mazzali et al. 2008).

The program of the special observations of the optical afterglow with the 6-m

telescope included the spectroscopy of the variable object discovered by Swift

observatory and identified with XRF 080109. We used SCORPIO focal reducer

mounted in the primary focus of the 6-m telescope (Afanasiev & Moiseev 2005). The

XRF 080109 was observed on January 16, 2008 (Figure 4.13) and February 6, 2008

(Figure 4.14). As the dispersive element we used a VPHG550G grism with an

operating wavelength interval 3500 – 7500 Å and resolution (FWHM) of 10 Å.

We reduced our spectroscopic data using standard technique, which included

the subtraction of electronic zero, correction for flat field, wavelength calibration

with the help of comparison spectrum of a Ne-Ar lamp, correction for atmosphere

extinction and calibration by an absolute flux with the use of observations of a

spectrophotometric standard at every night.

Then observational spectra were corrected for galactic extinction according to

the dust distribution maps of Schlegel, Finkbeiner, & Davis (1998).


72

Figure 4.12 The light curve of SN 2008D. The shape is similar to the light

curves of other GRB / SN (Sollerman et al. 2006).


73

Figure 4.13 Observations of SN 2008D with BTA / SCORPIO on Jan. 16.05, 2008

Figure 4.14 Observations of SN 2008D with BTA / SCORPIO on Feb. 6.18, 2008


74

The value of E(B-V)_host in the direction of the XRF 080109 (α2000 =

09h09m30s.621, δ2000 = +33o08´20´´.16) is equal to 0.6 (Modjaz et al. 2008). We

computed the extinction in terms of the dust-screen model in accordance with the

formula

Fint (λ) = Fobs(λ)100.4k(λ)E(B-V) (4.5)

where Fint (λ) and Fobs(λ) are emitted (unabsorbed) and observed fluxes,

respectively. The Milky Way extinction, k(λ), in the XRF080109 direction is

negligible with E(B-V)_MW = 0.022. The final spectra of XRF 080109 / SN 2008D

is shown in Figure 4.15.

Figure 4.15 BTA / SCORPIO spectra of SN 2008D

Physical conditions in the envelope of this SN were modeled with the

parameterized SYNOW code similar to XRF / GRB 060218 / SN 2006aj. Main


75

absorption features noticeable in both spectra (Figure 4.16 and 4.17) as P-Cyg

profiles of lines He I, Fe II, O I, and presumably, H I (the feature at 6200 Å). Models

with H I are shown as black thick lines; brown lines are the observed spectra. The

models containing Si II (λ_rest = 6347 Å) instead of H I must to have a velocity

vmine lower than vphot, the photosphere velocity, to fit the observed spectrum.

Vphot is equal to vmine of Fe lines. Si II model parameters are: tau1 = 0.005, vmine

= 23.0 km/s, ve = 10.0 for the first spectrum and tau1 = 0.5, vmine = 8.5 and ve=5.0

for the second (red thin lines at the Figures 4.16 and 4.17). The absorption near 6200

Å can be also explained by presence of CII with parameters tau1 = 0.005, vmine =

24.0, ve = 10.0 and tau1 = 0.0008, vmine = 16.0, ve = 3.0 for the first and second

spectra respectively (blue thin lines at the Figures 4.16 and 4.17).

Figure 4.16 The BTA spectrum of SN 2008D obtained on Jan. 16 (6.49 d after explosion) in wavelength range 3700 – 7500 Å. Main absorption features noticeable in both spectra (this and Feb. 6 spectrum, see Figure 18) were interpreted with the help of the SYNOW code as the P-Cyg profiles of the lines HeI, FeII, HI, and OI. All vmine velocities are given in thousands km s-1.


76

Figure 4.17: The BTA spectrum of SN 2008D obtained on Feb. 6 (since 27.62 d after explosion) in the same wavelength range as previous.

The absorption minimum near 6100 Å is well fitted by Hα with vmine = 23.0

km/s, tau1 = 0.4 and vmine = 15.0 km/s, tau1 = 0.5 for the first and second spectrum

respectively (see Table 4.4 and 4.5). Beside this absorption feature, we have also fit

other alternative lines (as Si II λrest = 6347 Å and C II λ_rest = 6580 Å) to Hα, which

are acceptable and previously introduced in literature for similar features showed in

Type Ib-c SNe (Elmhamdi et al. 2006, Valenti et al. 2008). As was seen in the

Figures 4.16 and 4.17 Si II λrest = 6347 Å lines can be excluded for the fitting of SN

2008D. So, CII λrest = 6580 Å remains a possible alternative for Hα. Therefore, the

absorption features close to 6100 Å are probably due to the contamination by Hα and

CII (Elmhamdi et al. 2006, Valenti et al. 2008).

Table 4.4 Parameters for the spectrum of SN 2008D obtained 6.48 days after explosion. The model parameters correspond to the black thick line at the Figure 4.16.


77

Table 4.5 Parameters for the spectrum of SN 2008D obtained 27.62 days after explosion. The model parameters correspond to the red thick line at the Figure 4.17.

4.2.1. Velocities of Photosphere and Envelopes for both SN 2006aj and SN

2008D

Using the best fitting SYNOW models of two spectra of SN 2006aj (2.55 d

and 3.55 d after the explosion) and two spectra of SN 2008D (6.49 d and 27.62 d

after explosion) the expansion velocities of the photosphere and hydrogen and

helium envelopes for each observation epoch are estimated and shown in Table 4.6.

Table 4.6 Expansion velocities of photospheres (vphot), hydrogen (VH), and helium (VHe) envelopes of SN 2006aj and SN 2008D, relative to time since the burst. The values in brackets is the time from the classic maximum connected with the radioactive decay 56N → 56Co → 56Fe. The classic maximum for the SN 2006aj determined as 10.4 d after the explosion (using V band light curve, Sollerman et al. 2006), for SN 2008D – 19 d after the explosion (Mazzali et al. 2008).


78

We compare these velocities with the empirical law of expanding velocities

of photospheres and envelopes of 11 Type Ib SNe, which were implied in the paper

Branch et al. 2002. Comparisons are illustrated as Figure 4.18 and 4.19. Possible

deviation of velocities of SN 2006aj from the empirical power law connecting the

large differences between the time of classical maximum of this supernova and the

analogus moments of Type Ib SNe.

Figure 4.18 Velocity at the photosphere, as inferred from FeII lines, is

plotted against the time after maximum light. The line is a power-law fit to the data, with SN 1998dt at 32 days (open circle) excluded ( Branch et al. 2002). Squares (SN 2008D) and diamonds (SN 2006aj) are photospheric velocities, inferred from our spectra.


79

Figure 4.19 Minimum velocity of the HeI lines (filled squares when

undetached, filled diamonds when detached) and the minimum velocity of the hydrogen lines (open circles; always detached) are plotted against time after maximum light. The curve is the power-law fit to the velocity at the photosphere, from Figure 4.18. Filled circles represent the minimum velocity of the HeI lines for SN 2008D, crosses refer to the minimum velocity of the hydrogen lines for SN 2008D, and triangles mark the minimum velocity of the HeI and hydrogen lines for SN 2006aj. Other data are from Branch et al. 2002.

4.3. Modeling Wide-Band Spectra of GRB 021004 and GRB 060218 Host

Galaxies

Observations of host galaxies have two purposes. First, emission lines in

spectra of host galaxies and absorptions in spectra of optical afterglows caused by

extinction in interstellar medium of the host galaxy allows us to determine the

distance to source of GRB, which is necessary for a complete astrophysical

interpretation of this phenomenon. Second, the study of properties of host galaxies is

another important tool for clarification of nature of progenitors, which is

undoubtedly one of the most important questions for GRB's nature. More

observational facts indicate a relation between GRBs and evolution of massive stars:


80

• Observed fluxes in the blue part of spectra of host galaxies allow

assuming a high star forming rate.

• A substantial internal extinction in host galaxies means a high

concentration of gas and dust which is typical for regions of formation of

massive stars.

• Distribution of gamma-ray locations relative to centers of host galaxies in

the image plane coincide statistically with distribution of massive stars (in

the cases when the morphology of galaxies could be traced).

• Observation of intensive lines of iron in X – ray afterglows of some

GRBs indicates the presence of a dense gas enriched with heavy elements

surrounding regions of GRB formation.

The main aim of the work is to study physical characteristic of host galaxies

by modeling their spectra and comparing with data of wide-band photometry.

Physical parameters of galaxies were estimated as internal extinction, mass and age

of stellar population. In this work we use two values of initial metalicity (Zּס and 0.1

Zּס) and three variants of internal extinction law: that of the Milky Way (MW,

Cardelli et al. 1989), the Small Magellan Cloud (SMC, Prevot et al. 1984), and the

average extinction law for local galaxies with star forming burst (SB, Calzetti et al.

2000).

4.3.1. The Host Galaxy of GRB 021004

On October 4.504, 2004 the space observatory HETE detected a source in the

band 25 – 400 keV which was identified afterwards as a long gamma-ray burst GRB

021004 (http://gcn.gsfc.nasa.gov/other/2380.hete). Coordinates of the object after

precise location are α2000 = 00h26m54s.68, δ2000 = +18°55′41″.6.

The field of GRB 021004 was studied on November 30 and December 5,

2004 (at the epoch when contribution of the optical component could be neglected)

in four filters B, V, Rc, Ic with BTA SCORPIO device (the CCD TK1024, the pixel

size 0´´.289). Observations were carried out according to the program of study of

GRB afterglows.

http://gcn.gsfc.nasa.gov/other/2380.hete


81

The CCD response curve in combination with filter transmission curves and

device optics implemented photometric system close to the standard Johnson-Kron-

Cousins system (Bessel 1990). Observational conditions were photometric and the

seeing was ≈1″.3.

The processing of data of the GRB 021004 field obtained with the BTA under

the program of the study of GRBs (Sokolov, V. V and Fatkhullin, T.A.) was

standard; it included subtraction of electronic zero, correction for the flat field,

elimination of traces of space particles and subtraction of traces of interference

pattern (for Rc and Ic bands). The values of stellar magnitudes and their errors in each

band (m ± dm), obtained after photometry and calibration by standards, were

corrected for the galaxy extinction according to maps of dust distribution (Schlegel et

al. 1998) and the extinction law of the Milky Way (Cardelli et al. 1989) (MWext

corrections and corrected stellar magnitudes m ± dm corr. are given in Table 4.7).

For the convenience of representation of modeling results, the stellar magnitudes (m

± dm corr.) were converted from the Vega photometric system into the AB system

(with known ABoff coefficients for each filter), and then into the flux density F and

flux density error dF (expressed in μJy) (Fukugita et al. 1995).

F = 10-0.4·(m+ABoff-MWext+48.568)+29 (4.6)

dF = 0.4 · ln(10) · F · dm (4.7)

Table 4.7 Photometry of the GRB 021004 host galaxy. m±dm are stellar magnitudes in the Vega system without correction for extinction; MWext is the Galactic extinction in the direction of a host galaxy; m±dm corr. are stellar magnitudes in bands with accounting the Galactic extinction; ABoff are coefficients of conversion from the Vega system to the AB system; F±dF are flux densities in each band.


82

Filter m+dm MW_ext m+dm corr AB_off F+dF, mJy B 24.55+0.11 0.31 24.24+0.11 -0.08 0.803+0.081 V 24.35+0.11 0.24 24.11+0.11 0.04 0.818+0.083

R_c 24.36+0.18 0.19 24.17+0.18 0.23 0.650+0.108 I_c 24.00+0.30 0.14 23.86+0.30 0.49 0.706+0.195 J 23.15+/-0.38 0.07 23.08+/-0.38 0.90 0.919+/-0.322

The remaining J filter data necessary for modeling were taken additionally from

available literature; the J photometry (de Ugarte Postigo et al. 2005), redshift of the

host galaxy z = 2.33 (Fynbo et al. 2005) and flux in the Lyα 1216 Å line;

(2.46±0.50)×10-16 erg s−1 cm−2 (Möller et al. 2002).

4.3.2. The Host Galaxy of GRB 060218

On February 18.149, 2006 the space platform SWIFT detected a source in the

band 15 – 50 keV identified as XRF 060218

(http://www.mpe.mpg.de/~jcg/grb060218.html). Coordinates of the source after

specification are α2000 = 03h21m39s.683, δ2000 = +16°52′01″.82.

Data of UBVRI photometry and redshift of the host galaxy were taken from

the Sollerman et al. (2006) (correction for the Galaxy extinction was already made).

Additionally, as a parameter necessary for modeling we used the integral flux in the

doublet [OII] 3727 Å/3729 Å: (196.33 ± 7.42) × 10−17 erg s−1 cm−2 (Wiersema et al.

2006, UVES spectrum). Analogously to the case of the GRB 021004 host galaxy,

stellar magnitudes of this galaxy were converted in flux density (the conversion

process is given in Table 4.8).

Table 4.8 Photometry of the GRB 060218 host galaxy. m±dm are stellar magnitudes in bands with accounting the Galactic extinction ABoff are coefficients of conversion from the Vega system to the AB system; F±dF are flux densities in each band.

http://www.mpe.mpg.de/~jcg/grb060218.html


83

Filter m+dm corr. AB_off F+dF, mJy U 20.45+/-0.15 0.74 12.49+/-1.73 B 20.46+/-0.07 -0.08 26.32+/-1.70 V 20.19+/-0.04 0.04 30.14+/-1.11 R 19.86+0.03 0.23 34.36+/-0.95 I 19.47+/-0.06 0.49 38.78+/-2.14

4.3.3. Modeling of Spectra

The modeling was made with the software package PEGASE (Projet d’Étude

des GAlaxies par Synthèse Évolutive

ftp://ftp.iap.fr/pub/from_users/fiov/PEGASE/PEGASE.2/) (Fioc & Rocca –

Volmerange 1997). The work implemented two method of representation of a model

spectrum: the one component (see, e.g. Fioc & Rocca – Volmerange (1997) ) and

two-component models (see, Sokolov et al. 2001). A fundamental idea of the two

component model is that the model galaxy spectrum is represented not by a single

spectral energy distribution, but by sum of two components: the first one is the young

population of galaxy, i.e. It simulates emission of star forming regions (hereinafter

referred to as a burst component), and the second one is emission of the old stellar

population. The computation algorithm can be presented as the following sequence

of steps:

• Computation of an array of model energy distributions with PEGASE

package. The input parameters are; initial metallicity and scenario of star

forming. We accepted the Salpeter initial mass function defined within the

stellar mass range from 0.1 Mּס to 120 Mּס , and two values of metalicity equal

to the solar metallicity (Zּס) and one tenth of the (0.1 Zּס). As a scenario of star

forming we took a simple instantaneous burst one and a more complicates

scenario of exponential fading (SFR ~ e-t/τ). The latter one demands setting a

so called characteristic fading time τ during which the star forming rate falls

e times. In this work we determined the range of τ values from 10 to 20000

million years with a detailed array of 18 values. Fluxes in nebular lines and

ftp://ftp.iap.fr/pub/from_users/fiov/PEGASE/PEGASE.2/


84

continuum were calculated in the PEGASE package for all model spectra.

Age values of model spectra were remained equal to PEGASE default ones,

i.e. From 0 to 20 billion years; they form an array of 69 values. Then

calculation results were selected from the input file of PEGASE into separate

corresponding to each value of age, metallicity and the parameter τ. Spectra

were represented as text files of the two-column format: the first column is

wavelength, the second column is monochromatic luminosity.

l The PEGASE software computes a spectral energy distribution at the red

shift z = 0, i.e. represents results expressed in monochromatic luminosities,

whereas from observations we get the spectral energy distribution. That is

why, to compare models with observations it is necessary to know distance,

i.e. it is necessary to set cosmological parameters. We took the following

ones: H0 = 65 km s−1 Mpc−1, ΩM = 0.3 and ΩΛ = 0.7.

Since the PEGASE software presents monochromatic luminosities relative to

mass of the Sun, it is necessary to set also an array of mass values. We accepted a

logarithm array with the step 0.1 within the range from 6.0 to 11.0, i.e. mass of

modeled galaxies are within 106 – 1011 Mּס. It should be kept in mind that this is

luminous mass (stars, gas) without regard to dark matter. The luminosity values of

output spectra obtained in the previous item were multiplied by mass and the

expression 1/(4πR2) where R is photometric distance corresponding to red shift of a

modeled galaxy. Wavelengths were converted to the observer’s frame of reference,

i.e. they were multiplied by the factor (1+z). The same procedure was also fulfilled

for luminosities of calculated nebular lines.

The burst component was determined from the known observational flux in

the doublet [OII] 3727Å/3729Å (for the GRB 060218 host galaxy) and in the Lyα

line 1216 Å (for the GRB 021004 host galaxy). I.e. a set of modeled continuum

spectra of the burst component was generated, for which the flux in the used line was

within 1σ from the observed one. The choice of the line [OII] 3727Å/3729Å was

determined by the fact that it is free from influence of stellar absorption. In case of

the GRB 021004 host galaxy this doublet is shifted by ≈ 12400 Å. The red feature


85

was not observed spectroscopically, therefore, as an alternative, we used the line Lyα

1216 Å which is the most noticeable in the galaxy spectrum (λobs ≈ 4050 Å).

So, two libraries of preliminary spectra (of the burst component and the old

population) were compiled for each modeled galaxy as text files corresponding to

each value of input parameters (the star forming type, age, metalicity, τ, mass).

l The main aim of the suggested method is the estimate of internal extinction

in host galaxies. In this work three types of laws were selected. The first one

corresponds to the extinction law in our Galaxy (Cardelli et al. 1989) (the

MW law); the second is the law obtained empirically from a sample of

spectra of galaxies with star forming in the local Universe (Calzetti et al.

2000) (the SB law); the third is the extinction law for the Small Magellan

cloud (Prevot et al. 1984) (the SMC law). Curves of these three laws are

shown in Figure 4.20. The laws can be parametrized well in a wide spectral

range by the relation RV ≡ AV / E(B-V). In the suggested method the

following values of RV were used:

3.1 for the MW law,

4.05 for the SB law, and

2.72 for the SMC law.

Extinction was superimposed only on the burst component, proceeding form

the fact that extinction in regions of active star forming is stronger than on average

in a galaxy. For simplicity, a model of screen for dust distribution (the dust-screen

model) was accepted. Then the mathematical expression of reddening takes the form

Fobs(λ) = Fint(λ)10−0.4·Aλ, where Fobs(λ) and Fint(λ) are an observed and proper

(without extinction) fluxes respectively, Aλ = k(λ) E(B-V) is extinction at the

wavelength λ. Then spectra of the burst component and old population were added.

The obtained sum is a finished model spectrum of galaxy.

• Then the procedure of fitting of finished model spectra to observed wide-

band ones was carried out. The finished model spectra obtained in the

previous step were integrated with transmission curves of filters used in

observations; then the value of statistics χ2 was calculated:


86

∑−

= 2

2,mod,2 )(

i

ieliobs ffσ

χ (4.8)

where fobs and fmodel are observed and model fluxes in the i band, σi is a

measurement error of the i flux.

l Comparison with local star-forming galaxies allows us obtaining the range of

E(B – V) values. Then, depending on the extinction curve type, it is possible

to obtain the range of AV. Breaking this range into an array of values and

fulfilling the calculation procedure, one can obtain the value AV

corresponding to minimum of χ2.

4.3.4. Results of Modeling

The modeling has shown a good convergence of model spectra to

observational data; in the best cases χ2/d.o.f. is equal to 0.158 and 1.158 for the GRB

021004 and GRB 060218 host galaxies respectively. Results and conclusion for each

galaxy are given in Tables 4.9, 4.10 and Figures 4.21, 4.22 for GRB 021004 and

Tables 4.11, 4.12 and Figures 4.23, 4.24 for GRB 060218.

The results of modeling testifies a negligible extinction for both galaxies: AV

= 0.47 (E(B-V)=0.15) in the best model for GRB 021004 and AV = 0.31 (E(B-

V)=0.08) for GRB 060218. The age of main stellar population determined from

modeling is within the range 0-3 and 70-140 million years for the burst component

and old population respectively. For the GRB 060218 host galaxy two components

with the ages 300-1600 million years and 15-20 billion years are selected for the

burst component and old population respectively. Other results of modeling is the

estimate of mass of visible stellar population M ≈ 5×1010Mּס for the GRB 021004

host galaxy and M ≈ 3×108Mּס for GRB 060218.

Estimates of internal extinction in host galaxies can be used in modeling and

interpretation of spectra of GRB afterglows and spectra of supernovae observed in

these galaxies (Valenti et al. 2008).


87

Figure 4.20 Curve of extinction laws used in the modeling. MW law is the

extinction in our Galaxy; SB law is the extinction of local galaxies with star forming; SMC law is the extinction in the Small Magellan Cloud.


88

Figure 4.21 Result of modeling of GRB 021004. The metalicity is solar. Thin

lines are for one-component models; thick lines are for two-component models.

Figure 4.22 Results of modelling GRB 021004. The metallicity Z = 0.1 Zּס.


89

Figure 4.23 Results of modeling GRB 060218. The metallicity Z = Zּס.

Figure 4.24 Results of modelling GRB 060218. The metallicity Z = 0.1 Zּס.


90

Table 4.9 Model parameters of the GRB 021004 host galaxy corresponding to minimum of χ2 for the metallicity Z = Zּס. Two types of star forming were considered: an instant one (τ → 0) and an exponential one (τ = 10 Myr – 20 Gyr). χ2/d.o.f. is the value of χ2 divided into the number of degrees of freedom (it coincides with the number of wideband data for this galaxy); law is one of three (MW, SB and SMC) applied extinction laws (No means the absence of absorption); N is the number of components in a model (N=1: the one-component model; N=2: the two-component model).

Table 4.10 The model parameters of the GRB 021004 host galaxy corresponding to minimum of χ2 for the metallicity Z = 0.1 Zּס.


91

Table 4.11 The model parameters of the GRB 060218 host galaxy corresponding to minimum of χ2 for the metallicity Z = Zּס.

Table 4.12 The model parameters of the GRB 060218 host galaxy corresponding to minimum of χ2 for the metallicity Z = 0.1 Zּס.

4.4. Observations of SWIFT J195509+261406 in The Field of GRB 070610 with

6-m BTA Telescope

Magnetars are neutron stars with an extremely strong magnetic field (1014-15

Gauss). Magnetar are called as soft gamma-repeaters or anomalous X – ray pulsars.

Soft Gamma Repeaters (SGRs) that are X – ray stars are the rare type of gamma-ray

transient sources. These sources are bright and repeating flashes in gamma-rays

(http://solomon.as.utexas.edu/~duncan/magnetar.html, Kouveliotou et al. 1994). Up

http://solomon.as.utexas.edu/~duncan/magnetar.html


92

to now SGRs were only detected in X – rays and gamma-rays. No optical counterpart

has yet been identified. In this work SWIFT J195509+261406 is investigated with

multi-wavelength observations. Bursting activity of SWIFT J195509+261406 has

been detected at optical wavelengths and it was suggested that SWIFT

J195509+261406 could be an isolated magnetar. In this case, SWIFT

J195509+261406 can be a link between the “persistent” soft gamma-repeaters /

anomalous X – ray pulsars and dim isolated neutron stars.

GRB070610 was detected with the Swift Burst Alert Telescope on at

20:52:26 UT on June 10, 2007 (Tueler et al. 2007). Following the detection of

GRB070610 as a single peaked GRB lasting about 4.6 s and its unusual X – ray

counterpart that was named as SWIFT J195509+261406. The multiwavelength

observing campaign was mounted by Castro-Tirado et al. (2008). The data were

collected starting ~ 1 min after the burst trigger time. In the first three nights of the

observations, the source displayed strong optical flaring activity (Kasliwal et al.

2008, Stefanescu et al. 2008, de Ugarte Postigo et al. 2007). This, together The

location of the source in the Galactic plane, supported the view that the source is

hosted by the Milky Way (10) and the strong evidence is given for this in this work

(Castro Tirado et al. 2008). The optical observations of GRB070610 were also

obtained with 6-m BTA telescope (Figure 4.25). Deep Ic – band and Rc – band

imaging was carried out on the 13th, 15th and 16th of June and on the 22nd of July

2007. A faint source was detected in the images obtained in the first three epochs

(Table 4. 13).


93

Figure 4.25 Composite color image based on the observation of BTA 6-m telescope

in R band, and 8.2-m VLT (Very Large Telescope) in I band and H band.

The flares from SWIFT J195509+261406 had durations ranging from tens of

seconds to a few minutes and flux amplitudes up to about 100 times the ‘outburst’

baseline flux (or ≥ 104 times the quiescent state). After 13 June, the activity decayed

abruptly (Figure 4.26) and no further flares were seen until 22 June, when a late time,

lower brightness flare was detected in the near infrared using the 8.2 – m Very Large

Telescope. A late time observation by the XMM – Newton spacecraft ~173 days

after the burst failed to detect the source, imposing an upper limit (3σ) to any

underlying X – ray flux of Lx ≤ 3.1 x 10−14 erg cm−2s−1 (0.2 – 10 keV).


94

Figure 4.26 a, Optical detections (Ic – band magnitudes, filled circles, with

1σ error bars) are shown together with 3σ upper limits (triangles). b. Swift X – ray data (0.2 – 10 keV, filled circles, with 1 σ error bars) together with the late time 3σ limit obtained with XMM – Newton (triangle). Both light curves show strong activity during the first three days, reaching the maximum around one day after the GRB and gradually decaying after the third day until the source became undetectable.

With the information of 12CO (J = 1 – 0) spectrum towards the SWIFT

J195509+261406 source reveals a molecular cloud at ~ 30 km s−1 from Castro –

Tirado et al. (2008). That contributes ~ 50% to the total column density N(H) derived

by Swift / XRT. Therefore Castro-Tirado et al. (2008) conclude that SWIFT

J195509+261406 is located in the Galaxy and beyond this particular molecular cloud

at a kinematic distance of D ≈ 3.7 kpc from the Sun. This value consistent with ~

4kpc derived from the 'red clump' method. Hereafter, 5 kpc will be considered.

To discern the nature of the source several possibilities were explored (Castro

Tirado et al. 2008). The first is that the source resembles the 'bursting pulsar' GRO

J1744-28 (Kouveliotou et al. 1996, Sazanov et al. 2005). However, Swift/BAT has

not recorded any other gamma-ray burst from SWIFT J195509+261406 after the

initial one. A second possibility is based on the proposed similarity to the black hole


95

candidate V4641 Sgr (Markwardth et al. 2007, Revnivtsev et al. 2002). This black

hole, orbiting an intermediate mass companion (a B9 subgiant), was suggested as the

first member of the 'fast X – ray novae' group (Uemura et al. 2002), and it has been

proposed that SWIFT J195509+261406 is a member of this class (Kasliwal et al.

2008). However, several lines of evidence indicate otherwise. First, the lack of

further detections of the baseline (non – flaring) flux during the outburst phase at

gamma-ray (Swift/BAT), milimetre and centimetre wavelengths (Kasliwal et al.

2008) implies a different physical mechanism, because considerable gamma – ray

and radio emission (the latter arising from a collimated jet) was recorded at the time

of the V4641 Sgr outbursts (Hjellming et al. 2000).

Figure 4.27 The magnitude distribution of the optical flares detected from

SWIFTJ195509+261406 in the Ic band is shown. Using all Ic band detections (with Watcher, Taut, OSN, IAC80, Mercator, BTA, VLT, FORS, NACO, BOOTES-2) of the source implied that flare fluxes are log – normally distributed as seen in the high-energy flares of SGR 1806-20 (Hurley et al. 1995) and SGR 1900+14 (Göğüş et al. 1999), supporting the claim SWIFT J195509+261406 is a new SGR, although this is not conclusive.

In this work, it is pointed out that the SWIFT source is an isolated compact

object, that is a new magnetar in our Galaxy, which displays activity like that of a

soft gamma-ray repeater (SGR) in the optical; and from which only one hard burst

was recorded in gamma-rays, near the onset of its bursting activity. If this is the case,

SWIFT J195509+261406 become the first SGR detected at optical wavelengths. This

would be supported by the burst durations (Figure 4.27) and the timing properties of

the flares.


96

The possibility of produced optical flares in SWIFT J195509+261406 is,

according to magnetar corona models (Thompson & Beloborodov 2005), that the

flares are due to coherent plasma bunches, and their luminosity depends on the

unknown bunching factor, leading easily to Lopt ≈ 1035 (D/5 kpc)2 erg s−1, as observed

at optical frequencies (Stefanescu et al. 2008).

Table 4.13: 6m – BTA observations of GRB 070610 field

Date (Start)

June 2007 UT

Filter Exposure

Time (s)

Magnitude

13.9578 Ic 120 >22.7

13.9626 Ic 120 >22.7

13.9642

Ic 120 >22.7

13.9662 Ic 120 >22.7

13.9678 Ic 120 >22.7

13.9694 Ic 120 >22.7

13.9710 Ic 120 >22.7

13.9728 Ic 120 >22.7

13.9747 Ic 120 >22.7

13.9775 Ic 120 >22.7

13.9793 Ic 120 >22.7

13.9811 Ic 120 >22.7

13.9828 Ic 120 >22.7

15.9669 Ic 120 >22.8

15.9720 Ic 120 >22.8

15.9742 Ic 120 >22.8

15.9769 Ic 120 >22.8

15.9787 Ic 120 >22.8

15.9806 Ic 120 >22.8

15.9824 Ic 120 >22.8


97

15.9841 Ic 120 >22.8

15.9859 Ic 120 >22.8

15.9877 Ic 120 >22.8

15.9916 Ic 120 >22.8

15.9933 Ic 120 >22.8

15.9953 Ic 120 >22.8

15.9970 Ic 120 >22.8

16.9140 Rc 180 >23.7

16.9170 Rc 180 >23.7

16.9200 Rc 180 >23.7

16.9230 Rc 180 >23.7

16.9250 Rc 180 >23.7

16.9270 Rc 180 >23.7

16.9300 Rc 180 >23.7

16.9330 Rc 180 >23.7

16.9350 Rc 180 >23.7

16.9380 Rc 180 >23.7

16.9410 Rc 180 >23.7

16.9440 Rc 180 >23.7

16.9470 Rc 180 >23.7

16.9500 Rc 180 >23.7

22.9530 July Ic 120 >22.3

12.8500 Oct Ic 3600 >22.5

5. CONCLUSIONS Eda SONBAŞ

98

5. CONCLUSIONS

GRB physics have progressed since the discovery of explosions in 1967. The

understanding of these explosions seems will keep scientists busy for some more

time to clarify all aspects. Especially the afterglow emissions from these bursts need

to be further explored to understand their nature and the physical conditions in their

environments.

In this work, additional arguments have been given in favor of the stellar wind

origin of the so called ‘shock breakout effect’ expected from core-collapse

supernovae observed by Swift/XRT/UVOT on XRF/GRB 060218 (Campana et al.

2006). In our optical spectra of afterglow of XRF/ GRB 060218, we detected

features which can be interpreted as a blueshifted Hα line. Similar early spectral

observations were made by ESO Lick, ESO VLT and NOT telescopes in the same

period (Pian et al. 2006, Sollerman et al. 2006); they reported no hydrogen line

detection, however. First hydrogen line observation from this object was then made

by our group using the BTA observations. This line was taken as a direct sign of a

relic wind envelope around a core-collapse progenitor star.

Results of observations and the modeling of two BTA spectra obtained 2.55d

and 3.55d after the explosion of SN 2006aj related to the X – ray flash XRF/GRB

060218 are presented below. The spectra were modeled in the Sobolev

approximation with the help of the SYNOW code (Branch et al. 2001, Elmhamdi et

al. 2006). We detected the following spectral features from the early spectra of the

Type Ic SN 2006aj :

1. the Hα P-Cyg profile for velocities of ~ 33000 km s−1 . This was a wide and

almost unnoticeable deformation of continuum in the range of ≈ 5600 - 6600 Å

for rest wavelengths (z = 0) at the first epoch 2.55 d.

2. Another Hα PCyg profile in absorption blueshifted by 24000 km s−1. This was a

wide and low contrast spectral feature at ≈ 6100 Å (rest wavelength) at the

second epoch 3.55d.

Actually, evolution of the same spectral features can be traced also in spectra

of SN 2006aj obtained with above-mentioned telescopes (Pian et al. 2006, Sollerman


99

et al. 2006). The hydrogen absorption features was being discussed in the literature

already. A more careful fitting by our code (SYNOW) reveal the existence of rather

shallow but definite presence of Hα in absorption. Such hydrogen features can

directly confirm the existence of a stellar-wind envelope. This was already observed

during XRF/GRB 060218 as a powerful black-body component first in X – ray

spectrum and then in the optical spectrum and interpreted as the so-called ‘shock

breakout effect’ (Campana et al. 2006). Thus, combining all early observations

carried out with BTA and other telescopes, we can interpret the data that we have

observed the evolution of optical spectra of the core-collapse SN 2006aj. This also

represents the transition from the phase of the Colgate shock breakout to spectra of

the phase of brightness increase due to the radioactive heating.

Our identification of these features with the Hα hydrogen line in early spectra

of the Type Ic SN 2006aj has been a confirmation of spectral observations of several

other Type Ic and Ib core-collapse SNe. (Branch et al. 2002, Branch et al. 2006,

Elmhamdi et al. 2006).

If the thermal component in the spectrum of GRB/XRF 060218 would be

confirmed by observations of afterglows of other bursts, then it will give a new

impulse to development of the theory of GRBs and related core-collapse SNe. The

intermediate redshift GRB/SNe are observed relatively rarely as compared to low

and high redshift events (Chapman et al. 2007). Nonetheless, they are the most

informative events (such as XRF/GRB 060218/SN 2006aj or GRB 030329/SN

2003dh) from the point of view of comprehension of relation between GRBs and

SNe. An important aspect of present study of these transient sources has been the

search for clues about wind envelopes around core-collapse progenitor stars of GRBs

both in the early spectra and in the photometry of GRB afterglows.

The signs of hydrogen in spectra of Types Ib, Ic and Ib-c SNe are not new. In

particular, the signs of hydrogen and evolution of the blueshifted Hα line were

already found (with the help of the same SYNOW code) in the analysis of time

series of optical spectra for usual core-collapse Type Ic and Ib SNe (Branch et al.

2001, Elmhamdi et al. 2006). In their analysis, special attention was given to the

traces of hydrogen in observations of these stripped envelopes of Type Ib-c SNe.


100

By formal definition, the Type Ic and Ib supernovae do not have conspicuous

lines of hydrogen in their optical spectra. These Ib and Ic SNe are usually modeled in

terms of the gravitational collapse of a massive core with rich carbon-oxygen

content. Its envelope is stripped before collapse and, apparently, signs of this

envelope must always be present in spectra as hydrogen lines.

The fact that in the usual and nearby SNe, the explosion does not begin with a

GRB is naturally explained by an asymmetric, axial-symmetric or bipolar (with

formation of jets) explosion of the core collapse SNe. Now one of the most popular

conceptions proceeds from the idea that in the case of flashes of the XRF type an

observer is out of the beam in which the most gamma-ray radiation is concentrated

(Soderberg et al. 2005). When observing at an angle close to 90o to the SN explosion

axis, no GRB is seen, and only an XRF and then a powerful UV flash could be

detected as in the case of SN 1993J.

We have also studied the core-collapse Type Ibc SN 2008D. A bright X – ray

transient XRT 080109 had accompanied to SN2008D as in the case of XRF/GRB

060218/SN 2006aj. No GRB companion was seen with the SN 2008D explosion.

The physical conditions in its envelope were also modeled with the nebular phase

observations using the SYNOW code. As the result of SYNOW modeling the

absorption features of P-Cyg profile lines He I, Fe II, O I, and H I (the feature at

6200 Å) noticeable in our spectra. These are acceptable and previously introduced in

the literature for similar features as in Type Ib-c SNe. Spectra have also been

modeled with some alternative lines Si II and C II to Hα. Si II λrest = 6347 Å lines

did not show a good fit to the spectra of SN 2008D. So, C II λrest = 6580 Å remains a

possible alternative for Hα. The absorption features close to 6100 Å are probably due

to the contamination to Hα by C II (Elmhamdi et al. 2006, Valenti et al. 2007). These

results also confirm our work on XRF/GRB 060218/SN 2006aj mentioned above.

We also studied the host galaxy of some selected GRBs. Data of wide-band

photometry of the host galaxies of GRB 021004 (with redshift z = 2.33) and GRB

060218 (z = 0.033) are compared with the synthetic spectra of spectral energy

distribution (SEDs) of star forming in local galaxies. An array of model spectra was

created with the help of the software package PEGASE. Besides a simple one –


101

component SED model, we considered a two-component model which was a sum of

the spectra of the young (the burst component) and the old stars (the old component).

The purpose was to estimate physical characteristics of galaxies, namely, the internal

extinction, mass and age of stellar populations. We used two values of initial

metallicity (Z and 0.1 Z ) and three variants of the internal extinction law: that of

the Milky Way (MW), the Small Magellanic Cloud (SMC) and average extinction

law for local galaxies with star forming burst (SB). Results of modeling testify a tight

interval of extinction for both galaxies: Av = 0.47, E(B-V) = 0.15 in the best fitting

for GRB 021004 and Av = 0.31, E(B-V) = 0.08 for GRB 060218. The age of basic

stellar population determined from modeling for the GRB 021004 host galaxy was

within 0 – 3 Myr for the burst component and 70 – 140 Myr for the old population.

Two components of ages 300 – 1600 Myr for burst component and >15 (15 – 20 as

model considers) billion years (Byr) for old population were selected for the GRB

060218 host galaxy. The results were an estimate of mass of the visible stellar

population M ≈ 5 × 1010 M for the GRB 021004 host galaxy and M ≈ 3 × 108 M

for GRB 060218.

In the last section of this work, GRB 070610 has been explored with the

multiwavelength observations. This source was also named as SWIFT

J195509+261406. We have observed the field of GRB 070610 with 6m-BTA

photometrically. As a result of these observations this source was discovered as a

new magnetar in our galaxy. This is the 9th magnetar in the milkyway. This magnetar

also displays activity like that of a soft gamma repeater (SGR) in the optical region.

In this sense, SWIFT J195509+261406 becomes the first SGR detected by ground

based telescopes. Use of all contemporary Ic band detections of the source from 6m-

BTA and other telescopes have shown that frequency distribution of flux values of

flares show a log – normal distribution expected from SGRs (Göğüş et al. 1999).

This supports the claim that SWIFT J195509+261406 is a new SGR with an unusual

optical activity.

102

REFERENCES

AKERLOF, C. W. ve ark., 1999. IAU Circ., 7100, 1

AFANASIEV, V. L. And MOISEEV, A. V. 2005. Astronomy Letters, vol. 31, Issue

3, p.194-204

AMATI, L. Et al. 2000. Science, Volume 290, Issue 5493, pp. 953-955

AMATI, L. et al. 2002. A&A, 390, 81

ANTONELLI, L. A. et. al. 2000. The Astrophysical Journal, Volume 545, Issue 1,

pp. L39-L42.

BAND 1997. astro-ph/9712193

BARON, E. et al. 2005. ASPC, 342, 351

BELCZYNISKI, K. ve ark., 2001. ASP Conference Series, Vol. 229

BERGER, E. et al. 2003. Ap. J., 590, 379.

BESSEL, M. S. 1990. PASP, 102, 1181

BISNOVATYI – KOGAN, G.S. et al. 1975. Ap. Space. Sci. 35, 23

BLANDFORD, R. D. & McKEE C. F., 1976. Physics of Fluids, vol. 19, Aug. 1976,

p. 1130-1138

BLINNIKOV, S. I. et al. 2002. astro-ph/0212569

BLOOM, J. S. ve ark., 1998. The Astrophysical Journal, Volume 507, Issue 1, pp.

L25- L28.

BLOOM et al. 2001. The Astrophysical Journal, Volume 554, Issue 2, pp. 678-683

BLUSTIN, A. J. 2007. astro-ph/0701804

BRANCH, D. et al. 2001. supernovae and Gamma Ray Bursters. Edited by K. Weiler

,Lecture Notes in Physics, vol. 598, p.47-75 (astro-ph/0111573)

BRANCH, D. et al. 2002. ApJ, 566, 1005

103

BRANCH, D. et al. 2006. astro-ph/0604047

CALZAVARA, A. J. & MATZNER, C. D. 2004. MNRAS, 351, 694C

CALZETTI, D. et al. 2000. Astrophysical Journal, 533, 682

CAMPANA et al. 2005. The Astrophysical Journal, 625: L23 - L26

CAMPANA et al. 2006. Nature vol. 441

CARDELLI, J. A., CLAYTON, G. C. & MATHIS, J. S. 1989. ApJ, 345, 245

CASTOR, J. I. 1970 MNRAS, 149, 111

CASTRO Tirado et al. 2008. Nature, Volume 455, Issue 7212, pp. 506-509

CAVALLO, G.; REES, M. J. 1978. Royal Astronomical Society, Monthly Notices,

vol. 183, p. 359-365

CHAPMAN et al. 2007. astro-ph/arXiv:0708.2106

CHUGAI, N. et al. 2004. astro-ph/0405369

CHEVALIER, R. A. 2004. IAUS, 218, 159C

COLGATE, S. A. 1968. Canadian Journal of Physics, 46, 476

CONNORS, A.; HUETER, GEOFFREY J., 1998. Astrophysical Journal v.501,

p.307

COSTA, E. et al., 1997. IAU Circ., 6576, 1

COSTA, E. et al. 1997. Nature, 387, 783

DE UGARTE POSTIGO, A. et al. 2005. Astronomy & Astrophysics, 443, 841

DE UGARTE POSTIGO, A. et al. 2007. GCN Circ. 6501

DINGUS, B. L. et al., 1997. AIP Conference Proceedings 410, p. 1592

DJORGOVSKI, S. G., BLOOM, J. S., KULKARNI, S. R., 2003. The Astrophysical

Journal, 591, p. 13 – 16.

EICHLER, D. et al., 1989. Nature, vol. 340, July 13, 1989, p. 126-128.

ELMHAMDI et al. 2006. A&A 450, 305

104

ENSMAN, L. & BURROVS, A. 1992. Astrophysical Journal, Part 1 (ISSN

0004-637X), vol. 393, no. 2, July 10, 1992, p. 742-755

EVANS, W. D. and et. all., 1979. Science, vol. 205, p. 119- 121

FATKHULLIN, T. A. et al. 2006, GCN 4809, 1

FICHTEL, C. E. and et all., 1975. Proceedings From The 14th International Cosmic

Ray Conference, vol.1., p.29

FIOC, M. & ROCCA-VOLMERANGE, B. 1997. Astron. Astrophys. 326, 950

FISHMAN, G. J. et al. 1994. The Astrophysical Journal Supplement Series, vol. 92,

no. 1, p. 229-283

FLIPPENKO, A. et al. 1993. ApJ, 415L, 103

FLIPPENKO, A. et al. 2006. astro-ph/0603297

FRAIL, D. A. ve ark., 1997. Nature, 389, 261-263

FRAIL, D. A. ve ark., 1999. The Astrophysical Journal, Volume 525, Issue 2, pp.

L81-L84

FRAIL, D. A. et al. 2001. Ap. J. Letter, 562, L55.

FRONTERA et al. 1999. Astronomy & Astrophysics Suplement Ser. 138, 399-400

FRYER, C. L. 1999. The Astrophysical Journal, Volume 522, Issue 1, pp. 413-418

FRYER, C. L.; WOOSLEY, S. E.; HARTMAN, D. H., 1999. Astrophysical Journal,

526, 152 – 177

FUKUGITA, M., SHIMASAKU, K., ICHIKAWA, T. 1995. PASP, 107, 945

FYNBO J. P. U. et al. 2005. Astrophys. J. 633, 317

GALAMA, T. J. et al., 1997. IAU Circ., 6574

GALAMA, T. J. et al., 1998. Nature, 395, 670-672

GEHRELS, N. et al., 2002. Scientific American, 287, part no 6, 52-59

105

GEHRELS, N. et al., 2005. Nature, 437, 851

GHISELLINI, G., 2001. astro-ph/0111584

GOODMAN, J., 1986. Astrophysical Journal, Part 2, vol. 308, Sept. 15, 1986, p.

L47-L50

GORBATSKY, V. G. & MININ, I. N. 1963. “Non-stationary stars”

GÖĞUŞ, E. et al. 1999. The Astrophysical Journal, Volume 526, Issue 2, pp. L93-

L96

GROOT et al. 1997a. IAU Circ., 6584

GROOT et al. 1997b. IAU Circ., 6588

HARRISON, F. E. et al. 1999. Ap. J. 599, 123.

HEGER, A. 1998perm.conf.....H

HEISE, J. and in’t Zand, Jean 2004. Nuclear Physics B Proceedings Supplements,

Volume 132, p. 263-270.

HEISE, J. et al. 2001. in Proc. of the Int. Workshop, astro-ph/0111246

HJELLMING, R. M. et al. 2000. Astrophs. J. 544, 977 - 992

HJORTH, J. et al. 2003. Nature, 423 - 847

HURLEY, K. 1992. Sky & Telescope (ISSN 0037-6604), vol. 84, no. 6, p. 631-636.

HURLEY, K. et al., 1994. American Astronomical Society, 184th AAS Meeting,

#15.01; Bulletin of The American Astronomical Society, vol.26, p. 881

HURLEY, K. et al., 1995. Astrophys. Spac. Sci. 231, 81 - 84

HURLEY, K. et al., 2003 GCN 2439

IMSHENNIK, V. S. & NADEZHIN, D. K. 1988. Uspekhi Fizicheskikh Nauk, 156,

521

JANKA, H. et al. 2007. Phys. Rep. 442, 38 - 74

106

JELINEK et al. 2007. (not published yet).

KASLIWAL, M. M. et al. 2008. Astrophys. J. 678, 1127 - 1135

KATZ, J. L., 1994. Gamma-Ray Bursts, Proceedings of the 2nd Workshop held in

Huntsville, Alabama, AIP Conference Proceedings Vol. 307, 1994,

p.529

KATZ, J. L. & CANEL, L. M., 1996. Astrophysical Journal v.471, p.915

KIPPEN, R. M. ve ark., 1998. Advances in Space Research, Volume 22, Issue 7,

p. 1097-1100

KLEBESADEL, R. W.; STRONG, I. B.; OLSON, R. A., 1973. Bulletin of The

American Astronomical Society Vol.5, p.322

KOUVELIOTOU, C., 1981. Muenchen, Technische Universitaet, Fakultaet fuer

Physic, Doktor der Naturwissenschaften Dissertation, 1981, 206 p

KOUVELIOTOU, C. ve ark., 1993. Astrophysical Journal, Part 2 - Letters (ISSN

0004-637X), vol. 413, no. 2, p. L101-L104

KOUVELIOTOU, C. et al., 1994. Nature 368, 125 – 127

KOUVELIOTOU, C. et al., 1996. Nature 379, 799 – 801

KOUVELIOTOU, C. Et al., 1997. IAU Circular 6660: GRB 970508

KOUVELIOTOU, C. et al. 2004. The Astrophysical Journal, 608:872 – 882

KÖNİGL, A. 2003. astro-ph/0302110v1

KULKARNI, S. R., et al. 1999. Nature, 398, 389-394

KUNTSCHNER, H. & DAVIES, R. L. 1998. MNRAS, 295, L29.

KURT, V. G. 2005, Nuevo Cimento 28, 521

KWAI, N. et al. 2006. Nature, vol. 440, Iss. 7081, pp. 184 - 186

LAMB, D. Q., ve ark. 2004. New Astronomy Reviews, Volume 48, Issue 5-6, p.

107

423- 430

LEONARD, D. C. et al. 2006. Nature, Volume 440, Issue 7083, pp. 505-507

Li – Xin Li, 2008. Astro-ph/0808.1184v1

LUCY, L. B. 1999. A&A, 245, 211

MACFADYEN, A. I.; WOOSLEY, S. E., 1999. Astrophysical Journal, v. 524, p. 262

–289

MARKWARDTH, C. et al. 2007. Astron. Tel. 1102

MASSIMILIANO et al. 2007. astro – ph/ 0510566

MATZ, S. M. et al., 1999. GRB Coordinates Network,231,1

MAZETS, E. P.; GOLENETSKI, S. V., 1981. Soviet Scientific Reviews, Section E:

Astrophysics and Space Physics Reviews, vol. 1, 1981, p. 205-266.

Translation

MAZZALI, P.A. et al. 1993, A&A, 279, 447

MAZZALI, P. A. et al. 2005. Sci...308.1284M (arXiv:astro-ph/0505199v1)

MAZZALI, P.A. et al. 2006, Nature, 442, 1018

MAZZALI, P. A. et al. 2008. Science, Volume 321, Issue 5893, pp. 1185-

MAEDA, et al. 2008. Science, Volume 319, Issue 5867, pp. 1220

MESZAROS, P.; REES, M. J., 1993. Astrophysical Journal Letters v.418, p.L59

MESZAROS, P.; REES, M. J., 1997. Astrophysical Journal v.476, p.232

MESZAROS, P., 1999. Astronomy and Astrophysics Supplement, v.138, p.533-536

MESZAROS, P., 2002. Annu. Rev. Astron. Astrophys. 40 & astro-ph/0111170

MESZAROS, P., 2003. Nature, Volume 423, Issue 6942, pp. 809-810

METZGER, M. R., ve ark. 1997. Nature, 387, 879

MIRABAL, N. et al. 2006. ApJ 643, L99

MODJAZ, M. et al. 2006. ApJ, 645l, 21

108

MODJAZ, M. et al. 2008. The Astrophysical Journal, Volume 687, Issue 1, pp. L9-

L12

MORAN et al. 2004. ESASP.552..653M

MOSKVITIN et al. 2008. Gamma-ray bursts and practical cosmology. 2008pc2..

conf..228M

MÖLLER, P. et al. 2002. Astron. Astrophys. 396, L21

MURAKAMI, T. ve ark., 1988. Nature (ISSN 0028 - 0836), vol 335, sept. 15, 1988,

p. 234,235

NOMOTO, K. et al. 1993. Nature, 364, 507

ÖZEL, M. E., 1982. PhD thesis.

PACZYNSKI, B., 1986. Astrophysical Journal, Part 2, vol. 308, Sept. 15, 1986, p.

L43- L46

PACZYNSKI, B.; RHOADS, J. E. 1993. Astrophysical Journal Letters v.418, p.L5

PACZYNSKI, B., 1998. ApJ. Lett. 494, 45

PARRENT et al. 2007. astro-ph/0701198

PEDERSEN, H. et al. 1998. Astrophysical Journal v.496, p.311.

PEDICHINI et al. 1997. IAU Circ., 6635

PIAN, E. et al. 2006. Nature 442, 1011

PIAN, E. & MAZZALI, P. A., 2006. Chinese Journal of Astronomy and

Astrophysics, Supplement, Volume 6, Issue S1, pp. 335-341

PIRAN, T., 1998. Proceedings of the Yamada Conference XLIX on Black Holes and

High Energy Astrophysics. Frontiers Science Series No. 23, p. 217

PIRAN, T. 1999. Physics Reports, Volume 314, Issue 6, p. 575-667

PIRAN, T., 2001. Bulletin of the American Astronomical Society, Vol. 33, p.833

109

PIRAN, T. 2004. astro-ph/0405503

PIRO, L. Et al. 1999. Astron. & Astrophys. Supp., 138, 431

PREVOT, M. L. et al. 1984. Astron. Astrophysics, 132, 389

RACUSIN et al. 2008. Nature, Volume 455, Issue 7210, pp. 183-188

REEVES, J. N. et al. 2002. Nature, 416, 512.

REMILLARD,R., WOOD, A., SMITH, D., LEVINE, A., 1997. IAU Circ. n.6726

REVNIVTSEV, M. et al. 2002. Astron. Astrophys. 391, 1013 – 1022

RHOADS, J. E., 1999, Ap. J., 525, 737.

RICHARDSON, D. et al. 2002. AAS, 201, 5609R

RICKER, G. R. ve ark., 2003. American Astronomical Society Meeting 203, #62.01

RICKER, G. R. ve ark., 2005. American Astronomical Society Meeting 206, #29.02;

Bulletin of the American Astronomical Society, Vol. 37, p.472

ROSE, J.A. et al. 1994. AJ, 108, 2054

RYKOFF, E. S., et al. 2004. Astrophys.J. 601, 1013-1018

SAKAMOTO, T. et al. 2007. The Astrophysical Journal, Volume 669, Issue 2, pp.

1115-1129

SAZANOV, S. et al. 2005. Astron. Lett. 23, 326 - 334

SCHAEFER, B. E. et al. 1994. ApJS, 92, 285

SCHLEGEL, D. et al. 1998. ApJ, 1998, 500, 525

SHIGEYAMA, T. et al. 1994 ApJ, 420, 341

SMITH et al. 2002. ApJs 141, 415 – 428

SOBOLEV, V. V. 1958. Theoretical Astrophysics, ed. V. A. Ambartsumyan,

pergamon Press Ltd., London

110

SOBOLEV, V. V. 1960. Moving Envelopes of Stars, Harvard University Press

SOBOLEV, V. V. 1967. “The course of theoretical astrophysics”

SODERBERG, A. et al. 2005. astro-ph/0502553

SOKOLOV, V. V. et al. 2001. Astron. & Astrophys. 372, 438

SOKOLOV, V. V. et al. 2003 BSAO, 56, 5

SOKOLOV, V. V. et al. 2006, BSAO, 59, 5

SOLLERMAN, J. et al. 2006. A&A 454, 2 , 503

SOMMER, M.; MUELLER, D., 1978. Astrophysical Journal, part 2 – Letters to

Editor, vol. 222, p. L17 – L20

SPURIT, H. C. & PHINNEY, E. S., 1998. Nature, Vol. 393, p. 139

STANEK, K. Z. et al. 1999. Ap. J. Letter, 522, L39.

STEFANESCU, A. et al. 2008. Nature, Volume 455, Issue 7212, pp. 503-505

STRITZINGER et al. 2002. ApJ, 124, 2100

TAYLOR G. B. et al. 1997. Nature, Volume 389, Issue 6648, pp. 263-265

TAYLOR, G. B. et al. 2001. GRB Circular Network, 1122, 1.

THIMM, G. J. et al. 2003 MNRAS 238, 15

THOMPSON, C. & BELOBORODOV, A. M. 2005. Astrophys. J. 634, 565 – 569

TRAGER, S. C. et al. 2000. AJ., 120, 165.

TUELER, J. et al. 2007. GCN Circ. 6491

UEMURA, M. et al. 2002. Publ. Astron. Soc. Jpn 54, 95 - 101

VALENTI, S. et al. 2008. MNRAS, 383, 1485

VANDERSPEK, R. ve ark., 1999. Astronomy and Astrophysics Supplement, v.138,

p.565-566

VAN PARADIJS, J. ve ark., 1997. Nature, 386, 686-689

111

VIANELLO et al. 2009. Astronomy and Astrophysics, Volume 495, Issue 3, 2009,

pp.1005-1032

VIETRI, M., 1997. Astrophysical Journal Letters v.488, p.L105

VIETRI, M. and STELLA, L. 1998. The Astrophysical Journal, Volume 507, Issue

1, pp. L45-L48.

WATSON, D. et. al. 2002. Astron. & Astrophys. 393, L1.

WAXMAN, E., 1997. Astrophysical Journal Letters v.489, p.L33

WAXMAN, E. et. al., 1998. Ap. J., 497, 288.

WAXMAN, E., 2000. The Astrophysical Journal Supplement Series, Volume 127,

Issue 2, p. 519-526

WEILER, K. W., ve ark. 2002. Annual Review of Astronomy and Astrophysics, Vol.

40, p. 387-438

WIERSEMA, K. et al. 2006. Astron. Astrophys. 464, 529

WINKLER, C., et al. 1995. Astronomy and Astrophysics, v.302, p.765.

WOOSLEY, S. E., 1993. Bulletin of the American Astronomical Society, Vol. 25,

p.894

WOOSLEY, S. E., 1993. The Astrophysical Journal, 405, 275 - 277

WOOSLEY, S. E. et al., 2002. RvMP, vol. 74, Issue 4, pp. 1015-1071

WOOSLEY, S. E. & HEGER, A. 2006. Astrophys. J., 637, 914 – 921

WORTHEY, G. 1994, ApJS, 95, 107

YOSHIDA, A. and et all., 1989. Astronomical Society of Japan, Publications (ISSN

0004- 6264), vol.41 no.3, 1989, p. 509 – 518

YOSHIDA, A. et al. 1997. IAU Circ., 6593

YOSHIDA, A. Et al. 1999. Astron. & Astrophys. Supp., 138, 433

YOST et al. 2003. The Astrophysical Journal, 597:459

112

YOUNG, T. R. et al. 1995. ApJ, 449L, 51

YUAN et al. 2009. astro-ph/0707.3132v1

ZHANG, B.; MESZAROS, P. 2004. Int.J.Mod.Phys. A19 2385-2472

ZHANG, W.; FRYER, C. L., 2001. The Astrophysical Journal, Volume 550, Issue 1,

pp. 357-367

ZHANG, W. & MESZAROS 2003. Astro-ph/0311321v1

ZHANG, B. et al. 2006. ApJ 642:354




http://www.chandra.harvard.edu

http://www-glast.sonoma.edu/

http://w0.sao.ru/hq/grb/astronomy-6m.html

http://news.nationalgeographic.com/news/2008/06/080612-supernova.html)

http://glast.pi.infn.it/Nicola/GRB_Model/node2.html






http://www.chandra.harvard.edu

http://www-glast.sonoma.edu/

http://w0.sao.ru/hq/grb/astronomy-6m.html

http://news.nationalgeographic.com/news/2008/06/080612-supernova.html

http://glast.pi.infn.it/Nicola/GRB_Model/node2.html



113

CIRRICULUM VITAE

I was born in Adana in 15/07/1980. After completing the first, middle and

high school educations in Adana, I have registered to the Çukurova University

Physics Department in 1997. The first year I took English Preparatory course. I

graduated from the Pyhsics Department in 2002. In the same year, I have started

M.Sc. education in Cukurova University, Institute of Basic And Applied Sciences. I

completed M.Sc. education the high energy astrophysics in 2004. Then I started

Ph.D. program in Cukurova University Institute of Basic And Applied Sciences. I

have been worked in Special Astrophysical Observatory of Russian Academy of

Sciences during my Ph.D. studies in 2006 - 2009 .

114

APPENDIX - 1

1. Data Reduction of SN 2008D (6 February 2008 BTA spectrum)

1.1.Master bias, master flat, master arc:

1. Bias

bs = pf ( )

S5351011.FTS

S5351012.FTS

S5351013.FTS

S5351014.FTS

S5351015.FTS

S5351016.FTS

S5351017.FTS

S5351018.FTS

S5351019.FTS

S5351020.FTS

2. Flat

fl = pf ( )

S5350737.FTS

S5350738.FTS

3. Neon

aa = pf ( )

S5350736.FTS

4. Averaged frames

Spec_calib_pipeline, bs, fl, aa, ‘Mbias.fits’, ‘Mflat.fits’, ‘Marc.fits’,

TRIM_AREA = [19, 0, 2066, 1023], LOG_FILE = ‘calib.log’

1.2. Reducton of object’s spectra: minus master – bias, divide to master flat.

1. Select object frames

Obj = pf ( )

S5350734.FTS

115

S5350735.FTS

2. Make the file-names

red = ‘red_obj’ + STRING (INDGEN (2) + 1, FORMAT =

‘(I2.2)’)+’.fits’

3. First reduction

spec_reduc, obj, red, ‘Mbias.fits’, ‘Mflat.fits’, TRIM_AREA = [19, 0,

2066, 1023]

1.3. Wavelength calibration with He-Ne-Ar-lamp.

1. Calibration of master arc frame

Corr_ethalon, ‘Marc.fits’, ‘ethalonVPHG550G.fits’, ‘guess.dat’, 30

2. Check the lines detection

ident_lines, ‘Marc.fits’, ‘lines.dat’, ‘guess.dat’, AVER_BOX=20

window, retain=2

plot_ident, ‘lines.dat’

3. Dispersion curve

Disp_relation, ‘lines.dat’, ‘rect.fits’, 0.5, [3,3]

4. Wavelength calibration of object frames

Rectify_frame, ‘red_obj01.fits’, ‘lin_red_obj01.fits’, ‘rect.fits’

Rectfy_frame, ‘red_obj02.fits’, ‘lin_red_obj01.fits’, ‘rect.fits’

5. Check the calibration of object frames

Window, xsize=1124, ysize=840, xpos=200, retain=2

İntegrate_spec, ‘lin_red_obj01.fits’, ‘tmp.fits’, [300, 498]

İntegrate_spec, ‘lin_red_obj02.fits’, ‘tmp2.fits’, [549,850]

Sum_spec, [‘tmp1.fits’, ‘tmp2.fits’], ‘tmp3.fits’

Line_wave, ‘tmp3.fits’, 0

1.4. Cosmic Hits

İnpt = ‘lin_red_obj01.fits’

Outpt = ‘cosm_lin_red_obj01.fits’

Cosmics, inpt, outpt, GAIN = 0.486, NS=10, box=5

İnpt= ‘lin_red_obj02.fits’

Outpt = ‘cosm_lin_red_obj02.fits’

116

Cosmics, inpt, outpt, GAIN = 0.486, NS=10, box=5

1.5. Sky lines removing

Skyfit, ‘cosm_lin_red_obj01.fits’, ‘sky01.fits’, YOBJ=[350, 646], RES_FRAME=

‘sk_cosm_lin_red_obj01.fits’, DEGREE=3

Skyfit, ‘cosm_lin_red_obj02.fits’, ‘sky02.fits’, YOBJ=[350, 646], RES_FRAME=

‘sk_cosm_lin_red_obj02.fits’, DEGREE=3

1.6.Removing the host galaxy contribution

1. Select the boundaries of area of the host galaxy for plot it’s profile

on each lambda. Just watch the movie and select where object is

located and which part of host profile that you take.

İnput_frame1= ‘sk_cosm_lin_red_obj01.fits’

İnput_image1 = readfits (input_frame1, /SILENT)

Sz1=size (input_image1)

a = indgen (sz1[2])

window, xsize=1124, ysize=840, xpos=200, retain=2

for i=0, sz1[1] – 1, 1 do begin & plot, a, input_image1 [i, *],

xrange=[400, 580], yrange=[0, 500], xtitle= ‘Y position (along the

slit)’, ytitle= ‘counts’, charsize=2 & wait, 0.02 & print, i &

endfor

as the result 500 – 544 range the area pf galaxy profile interpolation.

512 – 532 is the area of object.

2. Make the profile and removing the galaxy.

Skyfit, ‘sk_cosm_lin_red_obj01.fits’, ‘gall_01.fits’, YOBJ=[512,

532], FIT_REGION=[500, 544],

RES_FRAME=’gall_sk_cosm_lin_red_obj01.fits’, DEGREE = 2

Skyfit, ‘sk_cosm_lin_red_obj02.fits’, ‘gall_02.fits’, YOBJ=[512,

532], FIT_REGION=[500, 544],

RES_FRAME=’gall_sk_cosm_lin_red_obj02.fits’, DEGREE = 2

3. Check

117

İnput_frame1 = ‘sk_cosm_lin_red_obj01.fits’

İnput_image1 = readfits (input_frame1, SILENT)


a= indgen (sz1[2])

input_frame2 = ‘gall_01.fits’

input_image2 = readfits (input_frame2,/SILENT)

sz2 = size (input_image2)

b=indgen (sz2[2])


for i=0, sz1[1] – 1, 1 do begin & plot, a, input_image1 [i,*],

xrange=[400,580], yrange = [0,500], xtitle=’Y position (along the

slit)’, ytitle= ‘Counts’, charsize=2 & oplot, b+500,

input_image2[i,*], color=255 & wait, 0.02 & print, i & endfor

İnput_frame1 = ‘sk_cosm_lin_red_obj02.fits’

İnput_image1 = readfits (input_frame1, SILENT)


a= indgen (sz1[2])

input_frame2 = ‘gall_02.fits’

input_image2 = readfits (input_frame2,/SILENT)

sz2 = size (input_image2)

b=indgen (sz2[2])


for i=0, sz1[1] – 1, 1 do begin & plot, a, input_image1 [i,*],

xrange=[400,580], yrange = [0,500], xtitle=’Y position (along the

slit)’, ytitle= ‘Counts’, charsize=2 & oplot, b+500,

input_image2[i,*], color=255 & wait, 0.02 & print, i & endfor

118

1.7. Extraction and sum of the object spectra

Extract_spec, ‘gall_sk_cosm_lin_red_obj01.fits’, ‘extract_spec01.fits’, [976, 22],

‘sky01.fits’, ERR_FRAME = ‘err_spec01.fits’

Extract_spec, ‘gall_sk_cosm_lin_red_obj02.fits’, ‘extract_spec02.fits’, [976, 22],

‘sky02.fits’, ERR_FRAME = ‘err_spec02.fits’

1.8.Averaging of the spectra of object

plot_spec, ‘extract_spec01.fits’, yrange=[0, 2e4], color=255

plot_spec, ‘extract_spec02.fits’, yrange=[0, 2e4], /noerase, color=150

sp = [‘extract_spec01.fits’, ‘extract_spec02.fits’]

sum_spec, sp, ‘extract_spec_12.fits’, /MED, /AVER

plot_spec, ‘extract_spec_12.fits’, yrange = [0 2e4], /NOERASE

1.9. Atmospheric extinction

Extin_correction, ‘extract_spec_12.fits’, ‘atm_extract_spec_12.fits’,

‘/users/eda/PROGRAMS/CALIBRATION/atm_sao.dat’

Plot_spec, ‘atm_extract_spec_12.fits’, yrange = [0, 2e4], /NOERASE, color=255

Similar calculation process are also done for the standard star spectra. After

atmospheric extinction correction response curve is created as follows;

119

Response, ‘atm_SPEC_std_12.fits’, ‘atm_SPEC_std_12.fits’, ‘fg191b2b.dat’,

‘std_response.fits’, RATIO=rr

Plot_spec, ‘std_response.fits’

Oplot, rr[*,0], rr[*,1], color=[255]

FLUX CALIBRATION and GALAXY EXTINCTION

Flux_calibration, ‘atm_extract_spec_12.fits’, ‘fl_ atm_extract_spec_12.fits’,

‘std_response.fits’

Window, xsize=1124, ysize=840, xpos=200, retain=2

Plot_spec, ‘fl_atm_extract_spec_12.fits’, xrange=[3200, 7800], yrange=[0, 1e-15],

color=255

Extinc_correction, ‘fl_atm_extract_spec_12.fits’, ‘Gal_fl_atm_extract_spec_12.fits’,

‘cardelli_curve.dat’, 0.022350, /GALACTIC

Plot_sspec, ‘Gal_fl_atm_extract_spec_12.fits’, xrange=[3200, 7800], yrange=[0, 1e-

15], /NOERASE

Spec2ascii, ‘Gal_fl_atm_extract_spec_12.fits’, ‘Gal_fl_atm_extract_spec_12.dat’, 0

INSTITUTE OF NATURAL AND APPLIED SCIENCES ...period of this study. First of all I would like to...

Documents

Transcript of INSTITUTE OF NATURAL AND APPLIED SCIENCES ...period of this study. First of all I would like to...