Galaxy Assembly h ough Me ge s
Unco e ing D y and Non-d y Me ge s in he SDSS
Paula B ochado
Depa amen o de F´ısica e As onomia da
Faculdade de Ciˆencias da Uni e sidade do Po o
Decembe 2012
Galaxy Assembly h ough Me ge s
Unco e ing D y and Non-d y Me ge s in he SDSS
Paula B ochado
Thesis Ad iso s
D . Ja le B inchmann and D a. Ca a ina Lobo
Thesis submi ed o he Faculdade de Ciˆencias da
Uni e sidade do Po o o comple e he deg ee o Ph.D. in As onomy
Depa amen o de F´ısica e As onomia da
Faculdade de Ciˆencias da Uni e sidade do Po o
Decembe 2012
To my mom, my dad and my husband.
5

Acknowledgmen s
”No e e y hing ha can be coun ed coun s, and no e e y hing ha
coun s can be coun ed.”
W. B. Came on, 1963
Simply pu , and as oman ic as i sounds, his p ojec would no ha e been possible
wi hou he people ha I hank in he ollowing lines. They ha e been he e wi h me,
in many ways, since he e y beginning o his jou ney.
Fi s and o emos , I am deeply g a e ul o my supe iso s Ja le B inchmann and
Ca a ina Lobo om whom I ha e lea ned much mo e han I can possibly g asp.
Ca a ina has been a pa o my pa h e e since I ook my i s s eps in as onomy, as
a young unde g adua e. I hank he dedica ion and suppo h oughou hese yea s,
while I ied o ind my way h ough science.
Mo e han a supe iso , Ja le has been a men o . I eel lucky o ha e had he guidance
o such a b illian scien is and a wonde ul pe son. The in aluable lessons he has
augh me om day one, ne e se ling and always ques ioning, a e he co ne s one
o he as onome I aim o become. His endless pa ience, en husiasm and ca e when
eaching me, om he mos basic lines o code o he mos complex scien i ic analysis,
is some hing I will de ini ely nu u e o yea s o come, along wi h he sense o
wonde men he embedded in me. Thank you Ja le.
I hank my pa en s om he bo om o my hea . They ha e always, uncondi ionally,
encou aged me o ollow my d eams, especially his p ecious d eam o becoming an
as onome . They ha e augh me whe e he e is a will he e is a way, always che ishing
my sillies ideas and showing me how o make he bes ou o hem. Thank you o
you emendous suppo and abo e all o you inspi a ion.
I hank Rupes, who has been holding my hand o mo e han a decade now. He has
always been he e, challenging me, d eaming wi h me, e e y s ep o he his bumpy
7
way. I imagine ha pu ing up wi h me when I s a ed ambling abou galaxies was
some imes no easy, bu you always lis ened. I easu e how you always ound a way
o make me smile and a wo d o com o me. Thank you o being my sa e ha en.
My iends ha e been he backbone o my li e, especially in he las ew yea s. Ana
and Ped o, my pa ne s in c ime, hank you o bea ing wi h me in he mos u bulen
o imes, ne e lea ing my side. I hank you o he ca e, o he non-sense, o igh ing
agains and wi h me ne e sc a ching ou iendship and o hose memo able momen s
– whe he on s age, a a dinne able o in he middle o a o saken oad wi h ou
helme s on. My Twine Lu´ısa, wi h whom I ha e been sha ing he ups and downs o
his lea ning p ocess e e since I s a ed my academic oyage, and Jo˜ao, ha e kep
me ancho ed o wha eally ma e s in he la e s ages o his wo k. I am eage ly
looking o wa d o ul ill ou long-las ing goal o aising a glass in Times Squa e, as I
am o b eak ice wi h a spoon o many mo e win e d inks. I ha e also been lucky o
ha e made new iends du ing his ime – wi h whom would I discuss as onomy as
passiona ely as oo ball i no wi h my I alian iend Daniele? – and o ha e deepen
ies wi h old ones, seeing hem g ow inside and ou side as onomy – I belie e we will
be sha ing s o ies o e many RENA dinne s ye o come. I am uly su ounded by
beau i ul people, places and s o ies, including an angel, a wa mhea ed ci y, a hal -
smile and a wonde ully c azy amily.
I also wan o lea e a hea el hank you o Doc o Alice Ma ques. Wi h he help, I
am he e, now, p esen ing his wo k and I am inally able o look back wi h a smile.
I acknowledge CAUP, my hos ins i u ion, and I hank S e ewach Leiden o he
aluable suppo du ing my s ays in he Ne he lands. In pa icula , I would like o
hank P o esso a Te esa Lago o poin ing me owa ds he igh di ec ion e e since
I had my e y i s as onomy class and my collabo a o Me cedes Filho wi h whom
I expec o be wo king and sha ing successes in he coming yea s. I hank in a e y
special way Joana Ascenso who has been no only a g ea iend bu a eache , and
Joana Sousa, wi h whom I s a ed his expe ience a ew yea s ago and ha e seen ou
iendship g ow much mo e beyond as onomy. Along wi h Joana and Joana, I also
sha ed my o ice wi h Rui, B uno and Jo ge. I will ce ainly miss ou ime oge he , ou
con e sa ions, ou sha ed ea s and ou unbea able ce ain ies. I also hank my e y
good iend Nelma o he suppo in c i ical momen s, being a igh -hand h oughou
his ime.
Wi hou music, li e would be a mis ake, Nie zsche once w o e, and indeed music, like
as onomy, is pa o wha I am. I hank Co al de Le as da Uni e sidade do Po o
8
o allowing me o sing some o he bes music e e composed and o gi ing me g ea
momen s o joy a e a day’s wo k. I would also like o hank my good iend and
compose Daniel o one o he mos exquisi e expe iences ha came ou o his hesis:
a music made om Figu e 2.16. Inspi a ion de ini ely comes in many o ms, shapes
and colo s. Jus like galaxies.
The las ew yea s ha e been no hing sho o a olle coas e ide – hey ha e augh
me mo e abou science, mysel and o he s, han I could possibly imagine. I canno
wai o he ides ha a e s ill ye o come.
9
3.3.2 D y and non-d y me ge s ......................107
3.3.2.1 Is he e a mass cu -o ? .................109
3.3.3 Me ge ac ions ...........................111
3.3.4 Me ge a e .............................113
3.3.5 En i onmen ............................115
3.3.6 D y me ge spec al p ope ies ...................115
3.4 Galaxy assembly h ough d y me ge s ...................117
4 Me als and Gas in Me ge s 122
4.1 Me allici y in Me ge s ...........................123
4.1.1 Mass-Me allici y Rela ion .....................123
4.1.2 Measu es o gas con en in galaxies ................125
4.1.3 Using a con ol sample .......................128
4.2 In es iga ing he me als and gas in me ge s ...............130
4.2.1 The low-me allici y gas con en o me ge s ............130
4.2.2 The gas con en o me ge s .....................132
4.2.3 The gas consump ion ime-scale in me ge s ............134
4.3 Discussion ..................................136
5 P obing single me ge s 145
5.1 A subse o d y me ge s ..........................146
5.2 The quin uple d y me ge .........................146
6 Conclusions and Fu u e Wo k 150
6.1 Assembling galaxies ia me ge s ......................150
6.1.1 S a o ma ion and nuclea ac i i y ................151
6.1.2 Majo and mino me ge s .....................151
16

6.1.3 D y and non-d y me ge s ......................152
6.1.4 En i onmen ............................153
6.1.5 Gas in galaxies ...........................154
6.1.6 P obing single me ge s .......................156
6.2 Fu u e wo k .................................156
6.2.1 Building la ge me ge samples ..................156
6.2.2 Building smalle me ge samples ..................157
6.3 Final ema ks ................................157
A Volume Co ec ion 158
B An illus a i e se o me ge s 160
Re e ences 179
17
Lis o Tables
2.1 Rep esen a i e sample o s udies in he li e a u e ............ 58
3.1 F ac ion o d y me ge emnan s in di e en mass bins .........121
B.1 A se o po en ial me ge s d awn om he me ge sample. .......160
18
Lis o Figu es
1.1 The Hubble Deep Field aken wi h he Wide Field Plane a y Came a
2 om he Hubble Space Telescope .................... 25
1.2 The e olu ion o he Hubble Sequence. .................. 27
1.3 Image o he mic owa e sky c ea ed om se en yea s o WMAP da a,
a e sub ac ion o he signal o ou Galaxy. ............... 28
1.4 The galaxy dis ibu ion ob ained om spec oscopic edshi su eys
and om mock ca alogues cons uc ed om cosmological simula ions. . 30
1.5 The Schech e unc ion. ........................... 32
1.6 S a o ma ion enhancemen as a unc ion o he p ojec ed sepa a ion,
p om Li e al. (2008). .......................... 37
1.7 In e ac ing galaxies showing signs o igge ed s a o ma ion (Hubble
Space Telescope). .............................. 39
1.8 A ”Toom e” Sequence o me ge s in he SDSS, aken om he me ge
sample (2) .................................. 41
1.9 In eg a ed s ella mass densi y as a unc ion o edshi , spli by mo -
phology (as indica ed) and wi h a s ella mass cu o M∗>1011M⊙
(Bundy e al. 2005). ............................ 43
1.10 A sequence o d y me ge s d awn om he SDSS, as desc ibed in 2 . . 46
1.11 Colo image om he iple ed me ge a z= 1.27 in an Dokkum
e al. (2001) ................................. 47
2.1 The spec oscopic edshi dis ibu ion o he spec oscopic galaxy sam-
ple o SDSS DR6 .............................. 54
19
2.2 The galaxy dis ibu ion o a subse o he inal sample om whe e he
pai sample will be buil . .......................... 56
2.3 The dis ibu ion o he numbe o galaxies in g oups as a unc ion o
he a e age p ojec ed dis ance and o eloci y o se . ........... 59
2.4 An example o he SDSS DR6 galaxies’ co e age di ided in he g id
showing he i s 6 le els o subdi ision. .................. 61
2.5 An illus a ion o he s ing o galaxies issue and he duplica ion p oblem. 62
2.6 The M101 galaxy, also known as Pinwheel Galaxy, as imaged in SDSS. 63
2.7 Spec oscopic edshi dis ibu ion o he galaxies om he me ge sam-
ple in ol ed in majo me ge s. ....................... 64
2.8 Median alue o he concen a ion index pa ame e as a unc ion o
p ojec ed sepa a ion, o he galaxies in he me ge sample. ....... 67
2.9 The median o he a ia ion o he di e ence in R90 and R50 o galaxies
in me ge s and isola ed galaxies. ...................... 68
2.10 Eccen ici y his og am o he galaxies in he me ge sample. ...... 69
2.11 g- colo - edshi diag am o he galaxies in he pa en sample o
spec oscopic galaxies o he SDSS DR6. ................. 72
2.12 g- colo - edshi diag am o he galaxies in he me ge sample plo ed
o e he pa en sample o spec oscopic galaxies o he SDSS DR6. . . . 73
2.13 g- colo - edshi plane o he galaxies in he me ge sample plo ed
o e he pa en sample o spec oscopic galaxies o he SDSS DR6. . . . 74
2.14 u- colo e sus absolu e -band magni ude o all spec osopic SDSS
DR6 galaxies. ................................ 75
2.15 The g- colo dis ibu ion o SDSS DR6 galaxies abo e a ac ion o 0.68
as a unc ion o s ella mass. ........................ 77
2.16 The pa ame e as a unc ion o in o he i e a i e sea ch o a ed
egion. .................................... 79
2.17 g- colo - edshi diag am showing he uppe and lowe limi s o he
lane. ..................................... 81
20
2.18 No malized dis ibu ion o he Hα/Hβ lux a io o he galaxies inside
he colo lane. ................................ 83
2.19 Enhancemen in he Hαemission as a unc ion o p ojec ed dis ance. . 92
2.20 The dis ibu ion o he galaxies in B inchmann e al. (2004) sample in
he BPT line- a io diag am. ........................ 94
2.21 The dis ibu ion o he galaxies in he me ge sample plo ed o e he
sample o spec oscopic galaxies o he SDSS DR6, in he same BPT
line- a io diag am. ............................. 94
2.22 AGN ac ion in me ge s as a unc ion o p ojec ed dis ance. ...... 95
2.23 Composi e and s a o ming ac ion in me ge s as a unc ion o p o-
jec ed dis ance. ............................... 96
3.1 F ac ion o 1:1 o 1:2 and 1:1 o 1:3 me ge s as a unc ion o emnan
s ella mass wi hin he edshi in e al 0 < z < 0.06 (Khoch a & Silk
2009) .....................................100
3.2 To al s ella mass dis ibu ion o mino and majo me ge s. ......103
3.3 Mass dis ibu ion o galaxies in me ge s, co ec ed o 0.63. .......104
3.4 Mass dis ibu ion o galaxies in majo and mino me ge s. .......105
3.5 The mass unc ion o he SDSS DR6 plo ed o e his og ams o galaxies
in majo me ge s, in buns o SSFR. ....................106
3.6 Mass dis ibu ion o galaxies in d y and non-d y me ge s. .......107
3.7 Mass dis ibu ion o galaxies in d y and non-d y majo me ge s. . . . 108
3.8 Mass dis ibu ion o he emnan s o majo me ge s and mino , d y
and non-d y me ge s. ............................110
3.9 Numbe densi y o all me ge s, o non-d y me ge s and d y me ge s. . . 111
3.10 Me ge a es o di e en mass a ios o he me ge sample. .......114
3.11 Dis ibu ion o he numbe o neighbou s o he d y me ge sys ems. . . 116
3.12 Spec al indices abundance as a unc ion o s ella mass. ........118
21

3.13 Top: The solid line shows he SDSS s ella mass unc ion, wi h he d y
me ge emnan mass unc ion es ima e being he solid line wi h e o
ba s. Bo om: The a io o he d y me ge emnan unc ion o he
o al s ella mass unc ion. This shows he egion o he s ella mass
unc ion whe e d y me ge s a e impo an . ................119
4.1 The median o he M-Z ela ion o galaxies in he SDSS DR6 wi hin
he edshi in e al 0.005 < z < 0.2 and o he galaxies in he me ge
sample. ....................................124
4.2 A compa ison o Σgas measu ed om HI(x-axis) e sus he Σgas mea-
su ed using he spec oscopic me hod o 37 egions in M101 ( aken
om B inchmann e al in p ep) ......................127
4.3 Dis ibu ion o gas as a unc ion o s ella mass o he edshi in e al
0.005 < z < 0.2. ..............................129
4.4 The median dis ibu ion o he di e ence in me allici y as a unc ion o
p ojec ed sepa a ion. ............................131
4.5 The median o he a ia ion o ∆ log gas as a unc ion o p ojec ed
dis ance. ...................................133
4.6 The a ia ion in Robe s ime, o gas consump ion ime-scale, as a
unc ion o p ojec ed sepa a ion. ......................135
5.1 Th ee po en ial me ging sys ems aken om he me ge sample. . . . . 147
5.2 The image, om he SDSS DR6, o he possible quin uple me ging sys em.149
6.1 The simple closed box model o he cen al egions o galaxies. . . . . 155
22
Chap e 1
In oduc ion
Ou o he many sciences encompassing he mos di e se ields o knowledge known
o us, as onomy is he only one whe e he objec o s udy is, li e ally, ou side ou
wo ld. F om as e oids o clus e s o galaxies, he wide a ie y o disciplines inside
as onomy ha e his peculia p ope y - ou o wha we see in a e y sho ins an o
he cosmic his o y, a comple e s o y is buil hoping o explain a single snapsho . The
ield o ex agalac ic as onomy is no di e en and unde s anding how he galaxies we
obse e came o be as we see hem oday is a s o y s ill o be comple ed.
A emp ing o ell his s o y is, by i sel , a daun ing ask. The o e whelming numbe
o galaxies we know by now and he as onishing di e ences be ween hem, om ed
o blue, om dwa sa elli es o gian cannibals, om young and s a o ming o old
and dead, he e a e no wo galaxies alike. Ne e heless, because we a e cu ious beings
by na u e, his has no s opped us, no will i o he nex decades, om ying o
unde s and why he galaxies a e he way hey a e, how and when we e hey o med,
wha is hei s o y as a whole and wha makes each one di e en .
One o he disco e ies made in he las ew decades is ha one o he pa hs leading
o he assembly o new galaxies is he me ge be ween al eady exis ing galaxies: he
s udy o his p ocess has shed some ligh in o heo ies conce ning galaxy o ma ion
and, simul aneously, has gi en ise o o he ques ions.
23
24 CHAPTER 1. INTRODUCTION
In his PhD p ojec , an a emp is made o answe some emaining ques ions, in-
e i ably aising u he ones, abou galaxy assembly h ough me ge s and, in pa ic-
ula , o me ge s be ween old, ed and dead galaxies.
1.1 Galaxies in he uni e se
E en when he Ea h was he cen e o he uni e se, he di e en poin s o ligh
in he celes ial sphe e we e al eady a ising cu iosi y among sky gaze s. Long be o e
Galileo made use o a elescope o es ablish ha he Milky Way is composed o a
e y la ge numbe o s a s in 1610, he And omeda galaxy and he Magellanic Clouds
we e subjec o sc u iny due o hei nebulous na u e. Wi h he de elopmen o new
elescopes, wha seemed o be a blank space be ween s a s and nebulae seen wi h
he naked eye, s a ed o be illed wi h o he s a s, o he plane s, hei moons and
e en o he nebulae. When Edwin Hubble obse ed some o hose nebulae in 1936
and ealized ha no only we e hey no pa o he Milky Way bu hey we e also
made o s a s, he classical de ini ion o galaxy was bo n and wi h i ended he deba e
ega ding he eali y o ”island uni e ses”.
Less han 20 yea s ago, he Hubble Space Telescope ga e us a glimpse o ha blank
space c ea ing he Hubble Deep Field No h (HDF, Williams e al. 1996) image, seen
in Figu e 1.1, whe e, wi h ew excep ions, e e y poin o ligh is nei he a s a no a
plane , bu in ac a galaxy when he Uni e se was 1 billion yea s old. As he eme gence
o o he powe ul ins umen s ended up p o ing, he galaxies a e indeed he building
blocks o ou uni e se and, ega dless he di ec ion we look a , hei dis ibu ion is
qui e simila .
The my iad o galaxies shown in he HDF image ange om high o low luminosi y
galaxies, om blue o ed and, e en mo e in e es ingly, om e y young galaxies in
he ea ly s ages o hei o ma ion o ully assembled galaxies. Galaxies wi h spi al
and disky s uc u es, alongside galaxies wi hou any pa icula ecognizable pa e n,
can bo h be easily iden i ied h ough hei b igh blue colo s indica ing he p esence
o newly o med s a s. La ge ed galaxies e ol ing quie ly and small galaxies wi h
hei colo s eddened by hei always g owing dis ance, a e also clea ly spo ed among
he housands o objec s. E en hough he e a e no wo galaxies alike, hese gene al
1.1. GALAXIES IN THE UNIVERSE 25
Figu e 1.1: The Hubble Deep Field aken wi h he Wide Field Plane a y Came a 2
om he Hubble Space Telescope (Williams e al. 1996). C edi : R. Williams (STScI),
he Hubble Deep Field Team and NASA
32 CHAPTER 1. INTRODUCTION
Figu e 1.5: Bes i o Equa ion 1.1 o obse ed composi e galaxy luminosi y
dis ibu ion o 13 clus e s. Filled ci cles show he e ec o including cD galaxies
in composi e. Schech e (1976)
e al. 1995), in pa icula , models consis en ly ound mo e blue galaxies han seen in
he obse a ions. By acing in de ail he me ging his o ies o an assemble o da k
ma e halos, Kau mann e al. (1993) ep oduced he obse ed ends in luminosi y,
colo , gas con en and mo phology o galaxies, bu p edic ed an excess in he B-band
luminosi y densi y o a ac o o 2. Ei he he models we e o e p edic ing he numbe
o da k ma e halos o some halos we e s ill obse a ionally unde ec able. Ano he
possible explana ion would be ha galaxy p ope ies a e no only de ined by he
la ge scale me ge o da k ma e halos, bu also by in insic p ocesses wi hin he
galaxies hemsel es con ibu ing o hei e olu ion and, as a esul , o hei place in
he luminosi y unc ion.
The i s clues owa ds his la e hypo hesis we e highligh ed wi h s udies using semi-
analy ic models o calcula e luminosi y unc ions in he K-band, which is mo e closely
ela ed o he s ella mass han he blue luminosi y used p e iously. Finding e idence
ha shows ha he numbe o massi e galaxies a z∼1 is al eady below he local
alue, he p emise o massi e galaxies passi ely e ol ing since z= 2 was e u ed and
new suppo o he idea o ha ing massi e galaxies s ill o ming nowadays eme ged
(e.g. Kau mann & Cha lo 1998;Kau mann & Haehnel 2000). In o he wo ds, he

1.2. GALAXY FORMATION AND EVOLUTION 33
p oblem would no be in he numbe o p edic ed halos. Mo eo e , Bald y e al. (2004)
ound a bimodal beha iou on he (u- ) colo s. M absolu e magni ude plane o local
(z < 0.08) galaxies. This bi-modali y is composed o blue and ed dis ibu ions and
e en mo e e ealing is he clea end o mo e massi e galaxies occupying he ed
dis ibu ion. The numbe densi y pe magni ude o he ed dis ibu ion o e akes he
blue dis ibu ion a ∼3×1010M⊙, hus indica ing ha , a b igh magni udes, he
majo i y o he galaxies a e on he ed sequence and ha he ea ly models o galaxy
o ma ion ended o p edic signi ican numbe s o massi e blue galaxies.
These esul s poin ed in he di ec ion o a heo y whe e massi e ed galaxies a e o med
h ough me ging o smalle objec s along wi h o he p ocesses ha would edden he
galaxy o he local obse ed measu emen s. P ocesses like eedback om s a s and
cen al supe -massi e black holes ha e long been hough o play a signi ican ole in
galaxy o ma ion (Dopi a 1985;Dekel & Silk 1986;Haehnel 1995;Silk 1997;Ke eˇs
e al. 2005;Di Ma eo e al. 2005). S ipping he newly assembled galaxy o i s
gas h ough eedback would ex inguish he o ma ion o new s a s and he s ella
popula ions al eady in place would con inue o age in o a ed massi e galaxy, in
ag eemen wi h Bald y e al. (2004).
In ac , ecen s udies ha e come o con i m he impo ance o hese p ocesses - se e al
s udies ha e shown using hyd odynamical simula ions ha in addi ion o a bu s o
s a o ma ion du ing a majo gas- ich me ge , winds a ise om he acc e ing cen al
supe -massi e black holes and he gas is swep away esul ing in a gas-poo emnan .
Because du ing he me ge phase idal in e ac ions ha e d i en he gas owa ds he
inne egions o he new assembled galaxy, he ac i e galac ic nuclei (AGN) ou lows,
ha eme ge sensibly 0.5 Gy a e he i s passage o he wo galaxies, emo e su icien
gas o hal s a o ma ion ac i i y and he emnan s ella popula ion e ol es apidly
o ed colo s(e.g. Di Ma eo e al. 2005;Sp ingel e al. 2005b;Bowe e al. 2006;C o on
e al. 2006). This is howe e one possible scena io and i is no e y well unde s ood
ye .
In he hie a chical con ex o ΛCDM heo y, galaxies o m i s inside da k ma e
halos and o he la ge sys ems a e o med subsequen ly by acc e ion and me ging
(e.g. Hogg 2006). S udying he me ge o galaxies - wi h galaxies being he basic
ing edien s o he so-called ”cosmic web”- is he e o e he e y i s s ep o di ec ly
co obo a e he hie a chical na u e o he ΛCDM pa adigm.
34 CHAPTER 1. INTRODUCTION
E en hough galaxy me ging is a pa h o o m new galaxies om wo (o mo e)
p ogeni o galaxies, i is con enien o sepa a e he di e en p ocesses leading o a
new galaxy, namely, he assembly o he galaxy and i s o ma ion. While he i s e m
alludes, in gene al, o he shape o he new galaxy in o which he p ogeni o galaxies
ha e been ans o med, he la e e e s o he o ma ion o i s s ella popula ion.
Rega ding he galaxies’ s ella popula ions, hey may al eady be in place wi hou
signi ican ans o ma ion du ing he me ge o hey can, indeed, be s ongly a ec ed
by he e en .
1.3 Galaxy in e ac ion and me ging
Once hough o be ”island uni e ses”, i is now widely accep ed ha mos o he
galaxies known in ou uni e se ha e unde gone some ype o in e ac ion a some poin
o hei li es. The consequences o a me ge e en a e con olled by a se ies o ac o s
om he mass a io o he galaxies o he ela i e eloci y o he encoun e and e en
hei ela i e posi ion a he ime o he i s passage - he e a e housands o obse ed
in e ac ing galaxies, in e y di e se s ages o hei encoun e s and in a wide ange o
shapes, mass a ios and ela i e eloci ies.
1.3.1 Dynamical p ocesses
G a i a ional in e ac ion be ween galaxies ends o signi ican ly a ec hei mo phol-
ogy (e.g. Toom e & Toom e 1972;Schweize 1986;Mihos 1995;Lo z e al. 2008a,b),
some imes beyond ecogni ion, and i hey end up me ging, he inal galaxy can be o
a comple ely di e en na u e om he p ogeni o s. The s udy o in e ac ing galaxies
s a ed mo e han hal a cen u y ago, wi h F i z Zwicky in his s udy o mo phological
ea u es o in e galac ic ma e (Zwicky 1953) and, o his da e, he bes known ype o
in e ac ion is be ween spi al galaxies. Due o hei iolen ans o ma ion du ing he
p ocess o in e ac ion, and e en ual me ging, he in e ac ion o spi al galaxies is no
only likely o change signi ican ly he mo phology o he galaxies, wi h he c ea ion
o b idges be ween he in e ac ing galaxies and ails due o idal o ces, bu o en
p oduces bu s s o s a o ma ion (Li e al. 2008). These eye-ca ching e en s make
his ype o me ge s sui able o subsequen obse a ions and, un il e y ecen ly, hese
1.3. GALAXY INTERACTION AND MERGING 35
we e indeed hough o be he only ypes o me ge s wo h o no e.
The deg ee o ans o ma ion o a galaxy due o a close encoun e is no easy o
es ima e. The spec um o galaxy in e ac ions is e y wide and i i includes he inal
me ge o he galaxies, is e en wide . Howe e , when exac ly a sys em o in e ac ing
galaxies ends up me ging in o a single objec , is mo e o en han no de e mined by
he eloci y a which hey i s encoun e ed each o he - in he ame o e e ence o he
mo e massi e galaxy, i he eloci y o he in alling galaxy is smalle han he escape
speed, hey become g a i a ionally bound and mo e o en han no he encoun e will
end in a me ge . Wi h espec o p oximi y,close encoun e s can lead o me ge s
i , as men ioned, he passing galaxy becomes g a i a ionally bound o i s companion.
On he o he hand, dis an , o weak, ansi s a e usually cha ac e ised by ” lybys”,
sligh ly dis up ing he galaxies due o idal e ec s, bu no necessa ily leading o a
me ge (Toom e & Toom e 1972;Pa k & Hwang 2009;Mihos e al. 1992).
Conce ning hei mass a ios, no all me ge s induce signi ican changes in he galaxy
mo phologies, as is he case o sa elli e me ge s, i.e. me ge s be ween galaxies wi h
e y di e en sizes, o masses. These me ge s a e usually deno ed as mino me ge s
and, by con as , me ge s o galaxies ha ha e ela i ely simila masses a e called
majo me ge s. In he li e a u e, he on ie o he mass a io be ween he membe s
o a me ge is usually se a 1
3, wi h mino and majo me ge s ha ing mass a ios,
espec i ely, below and abo e his limi (Lin e al. 2004;Bell e al. 2005;Bundy e al.
2007;Cio i e al. 2007;Bundy e al. 2009).
The en i onmen in which galaxies a e imme sed also de e mines he na u e o he
in e ac ion. I migh be expec ed, ha in e ac ions and me ge s would happen wi h
highe equency in mo e c owded en i onmen s, as is he case o clus e s o galaxies
whe e he g ea e numbe o galaxies could enable me ge e en s mo e easily. Howe e ,
he high eloci ies ha galaxies expe ience inside he clus e end o inhibi slow
passages and, as a consequence, po en ial me ge s (e.g. Mihos 2003). E en hough he
con ibu ions o he di e en p ocesses inside a high densi y en i onmen , and how
hey a e in e wined, a e a om ully unde s ood, some ypes o in e ac ions a e
al eady well s udied. In summa y i is possible o ca ego ize some o hose ypes o
in e ac ion and/o me ge , ha p oduce somewha signi ican changes in he galaxies,
as ollows:
•Cannibalism (e.g. Hausman & Os ike 1978;Lequeux e al. 1998) - This ype o
36 CHAPTER 1. INTRODUCTION
in e ac ion, and e en ual me ge , e e s o he phenomenon o smalle galaxies
being acc e ed by he cen al galaxy in a clus e o galaxies. As a consequence o
his p ocess, he cen al galaxy is usually he b igh es and i keeps inc easing i s
luminosi y and mass wi h he inc easing numbe o in- alling galaxies. As long
as he passing galaxy passes h ough he ou e en elope o he cen al galaxy a
a ela i ely low eloci y i will end up being cannibalized.
•Ha assmen (e.g. Moo e e al. 1996,1998;Lake & Moo e 1999;Ba azza e al.
2009) - F equen (a leas once pe Gy ) high-speed encoun e s o se e al hou-
sand km/s be ween galaxies, whe e he la ges galaxies wi h luminosi ies ypi-
cally ≥L∗g a i a ionally ac on smalle ones, dis up ing he lowe luminosi y
galaxy and emo ing mass wi hou being signi ican ly al e ed hemsel es. A e a
ew apid encoun e s he ha assed galaxy ends o ge a sphe oidal shape and, in
ac , his p ocess is hough o be esponsible o he o ma ion o dwa ellip icals
in clus e s. Moo e e al. (1996) ac ually in oked his mechanism o explain he
di e en mo phologies seen in clus e a z∼0.4, wi h a p edominance o small
disk galaxies, and a z∼0 whe e dwa ellip ical galaxies a e mo e common.
•Ram P essu e S ipping (RPS) (e.g. Abadi e al. 1999) and S angula ion (e.g.
Kawa a & Mulchaey 2008;Bekki e al. 2002) - These wo p ocesses, mos domi-
nan in gas- ich galaxies, like spi als, a e no caused by he in e ac ion wi h o he
galaxies bu wi h he in aclus e medium (ICM). None heless, hese in e ac ions
lead o spec al ans o ma ions o he a ec ed galaxies, whe e he i s signs o
he changing s ella popula ions can be de ec ed, and o mo phological ans o -
ma ions as well. In he case o RPS, he ICM exe s a p essu e on he in e s ella
medium o a galaxy mo ing h ough he clus e , s ipping aluable gas om
galac ic disk inhibi ing he o ma ion o new s a s. S angula ion, o s a a ion,
e e s o a simila p ocess bu he s a o ma ion is hal ed due o he ho gas
being emo ed om he halo o he galaxy, by he in e ac ion wi h he ICM. The
ICM emo es he gas due o he p essu e i exe s on i and he necessa y in- all
o gas om he halo o he disk, ha once cooled would ans o m in o s a s, is
quenched and he galaxy e en ually s ops i s s a o ma ion.
These dynamical p ocesses ha in e ac ing galaxies go h ough end o in oduce deep
ans o ma ions in hei physical p ope ies. In ac , one can conjec u e ha only a
e y pa icula se o ci cums ances o he in e ac ion o galaxies would no lead o
signi ican changes on all galaxies in ol ed. Encoun e s be ween galaxies ha e long
been associa ed wi h he mos ex eme p ocesses in galaxies. As men ioned abo e,
1.3. GALAXY INTERACTION AND MERGING 37
Figu e 1.6: S a o ma ion enhancemen as a unc ion o he p ojec ed sepa a ion,
p om Li e al. (2008). The galaxies in he his sample ha e been spec oscopically
con i med as galaxies in he SDSS DR4 and ha e been selec ed o be s a o ming
acco ding o he classi ica ion by Kau mann e al. (2003). This subse o galaxies
shows a clea enhancemen in hei speci ic s a o ma ion a es a small dis ances
be ween galaxies.
he igge ed s a o ma ion induced by he in e ac ion, and e en me ge , o gas- ich
galaxies, usually spi als, is one o he mos well s udied in e ac ing phenomena. Recen
s udies ha e shown ha idal o ces gene a ed in he in e ac ing galaxies a e e icien
in igge ing enhanced s a o ma ion by causing pe u ba ions in he gas (e.g. de G ijs
e al. 2003;Ba on e al. 2007;Li e al. 2008), and his enhancemen has been p o en
o be s ongly dependen on he dis ance o he neighbou galaxy, see Figu e 1.6.
Famous examples, among qui e a ew obse ed also by he Hubble Space Telescope,
o his phenomenon is he galaxy M51 (also known as Whi lpool Galaxy) and he
An ennae Galaxy, bo h shown in Figu e 1.7. The b igh blue and pink egions ma k
up in ense s a o ming ac i i y, bu while in he case o he An ennae Galaxy his
is he esul o a me ge be ween wo spi al galaxies, he Whi lpool Galaxy exhibi s
his e en as a esul o he pe u ba ions caused in i s gas by he companion galaxy
close o i . O he clea examples o in e ac ion induced s a o ma ion a e some ing
galaxies, as illus a ed in Figu e 1.7 b), easily iden i ied by hei shape cha ac e ized
by a b igh ing o ex eme s a o ma ion.

38 CHAPTER 1. INTRODUCTION
1.3.2 T ans o ma ion o galaxies ia me ging
The ans o ma ion o galaxies h ough me ge s has long been heo ized o be o g ea
impo ance in galaxy e olu ion. Typically, du ing a me ge e en , a leas one o he
galaxies in ol ed is likely o be comple ely ans o med and since galaxies a e hough
o su e a me ge e en du ing hei li e spans i is easonable o assume ha he
impo ance o in e ac ions in galaxy e olu ion is a om ma ginal.
Me ge s in ol ing gas- ich spi al galaxies a e by a he ones ha mos d ama ically
change he s uc u e o he galaxies in ol ed. Majo me ge s o gas- ich, disk galaxies
lead o he c ea ion o b idges and idal ails, inducing pe u ba ions in he gas and
c ea ing si es o s a o ma ion. As simula ions ha e epea edly shown, he emnan o
a majo spi al me ge has p ope ies consis en wi h hose o an ellip ical galaxy, as he
dis ibu ion o s a s and a ailable gas se les a ound he cen e o he al eady me ged
nuclei and i s s ella popula ions e ol e (e.g. Ba nes & He nquis 1996;Neg opon e
& Whi e 1983;Ba nes 1988;Ba nes & He nquis 1992;Sp ingel e al. 2005a;Richa d
e al. 2010;L´opez-Sanjuan e al. 2010). A ecen s udy by Hopkins e al. (2009b)
in es iga ed he impo ance o gas ( he key ing edien in spi al in e ac ions) on he
mo phological changes in me ge s and showed ha me ge s be ween gas- ich low-mass
galaxies end o p oduce bulgeless sys ems. On he o he hand, mino me ge s be ween
spi al galaxies and small sa elli e galaxies do no end o signi ican ly dis up he spi al
galaxy bu in u n con ibu e o he inc ease o he emnan bulge (Eliche-Mo al e al.
2006).
Ano he pa icula ype o ans o ma ion in ol ing spi al galaxies is he head-on
collision o a ela i ely small galaxy wi h he cen e o a spi al galaxy, p oducing he
ing s uc u es men ioned in Sec ion 1.3.1 and shown in Figu e 1.7. These in e ac ions
also depend on he ela i e eloci y o he galaxies o de ine hem as a me ging sys em,
as men ioned in Sec ion 1.3.1. I hey do me ge, simula ions o hese collisions ha e
shown ha hei la e s ages a e compa ible wi h he p ope ies o gian low su ace
b igh ness galaxies (GLSBs), cha ac e ized by hei unusual la ge and la discs, whose
o igin emain unce ain (Mapelli & Moo e 2009).
1.3. GALAXY INTERACTION AND MERGING 39
Figu e 1.7: In e ac ing galaxies showing signs o igge ed s a o ma ion. Top Le a)
NGC 4038/4039 (An ennae galaxy), image 0.3’ wide; Top Righ b) AM 0644-741 ing
galaxy, image 2’ wide; Bo om c) M51 (Whi lpool galaxy), image 9.6’ wide. These
images a e a composi e o sepa a e exposu es aken wi h di e en il e s - Johnson B
(blue), Johnson V (g een), Hα(pink) and nea -in a ed ( ed). The b igh blue and
pink colo s - clea ly seen in he An ennae galaxy, in he spi al a ms o he Whi lpool
galaxy and in he ing o image c) - indica e egions o ac i e s a o ma ion, since
he blue colo s esul om he e y massi e, and sho li ed, s a s and he pink colo s
om he exci ed hyd ogen a oms in he gas su ounding hese si es o s a o ma ion.
C edi : NASA, ESA and he Hubble He i age Team (STScI/AURA) - ESA/Hubble
Collabo a ion
40 CHAPTER 1. INTRODUCTION
Ei he way, he ans o ma ion o galaxies h oughou a me ge e en is a complex
p ocess and, e en hough imescales o a comple e me ge a e ela i ely small, 0.5
Gy o galaxies sepa a ed up o 20 h−1kpc (e.g. Lo z e al. 2008b;Pa on e al. 1997),
each o he se e al s ages al e s signi ican ly he in ol ed galaxies.
Mo e han 30 yea s ago, Toom e (Toom e 1977) de ined a sequence o pe u bed
me ging galaxies whe e he main bodies o he galaxies, wi h p ominen ails, appea ed
o be close. Adop ing he same p inciple, an illus a i e Toom e’s Sequence o he
galaxies in he me ge sample used in his wo k, aken om he SDSS as desc ibed
in 2, is shown in Figu e 1.8. All o hese sys ems a e in ob ious in e ac ion and hei
mo phologies a e isibly, and se e ely, al e ed. No e how he las ile (bo om- igh )
al eady esembles an ellip ical galaxy, e en wi h he s a o ming pocke s and some
no iceable idal ea u es.
Nowadays, i is widely accep ed ha ellip ical galaxies may be assembled h ough
me ge s o spi al galaxies, wi h hei pe u bed gas apidly and massi ely u ned in o
s a s and he me ging o he galaxies des oying hei disky s uc u e a ou ing he
c ea ion o a sphe oidal shaped new galaxy.
1.3.2.1 The case o Milky Way
Ou own Galaxy has mo e han a dozen neighbou sa elli es (e.g. S iga i e al.
2008). Among hem a e he well known La ge and Small Magellanic Clouds and
he Sagi a ius dwa sphe oidal galaxy (Sg dSph) known o be he mos d ama ic
in e ac ion case wi h he Milky Way. The Sg dSph is esponsible o some s ella
s eams obse ed in he Milky Way halo, bo h in he sou he n and no he n Galac ic
hemisphe es, indica ing a leas one passage h ough he Galaxy. Howe e , i s
hough o ha e happened wi hou signi ican dis up ion o he Milky Way’s disk
(Beloku o e al. 2006;Niede s e-Os hol e al. 2010), a e y ecen s udy by Pu cell
e al. (2011) sugges s, h ough simula ions, ha he esponse o he Milky Way o he
in all o he Sg dSph esul s in he o ma ion o spi al a ms, also in luencing he cen al
ba . In ecen yea s, se e al s udies ha e shown he p esence o idal s eams and o e
densi ies in he Milky Way disk and halo, sugges ing mino in e ac ions wi h se e al
sa elli e galaxies in he Local G oup (Newbe g e al. 2002;Ju i´c e al. 2008;Newbe g
e al. 2009). In ac , he ba -shaped co e o he Galaxy has been p oposed o be he
esul o a me ge e en ha happened 1.9 Gy ago (Minche e al. 2009). These
1.3. GALAXY INTERACTION AND MERGING 41
Figu e 1.8: A ”Toom e” Sequence o me ge s in he SDSS, aken om he me ge
sample (2). F om op-le o bo om- igh , he me ge s a e so ed acco ding o he
descending p ojec ed dis ances o hei membe s. Scale is a ying in his composi ion
om ∼50 kpc o ∼75 kpc on he side.
48 CHAPTER 1. INTRODUCTION
s a - o ming galaxies a low edshi s o build he massi e non-s a - o ming ea ly- ype
galaxies (Bell e al. 2004;Naab & Os ike 2009). Mo eo e , obse a ional s udies
ocusing on me ge s o ea ly- ype galaxies ca ied ou in he las decade showed ha
he s ella mass densi y in luminous ed galaxies has inc eased by a ac o o &2 o e
he edshi ange 0 < z < 1, possibly explained by me ge s o ed, bulge-domina ed,
galaxies ( an Dokkum 2005).
O e all, e idence seems o suppo he scena io o high luminosi y ellip ical galaxies
being he emnan s o me ge s be ween ea ly- ype galaxies (Naab e al. 2006;Bell
e al. 2006). Howe e , his is by no means an accep ed ac , o ins ance, Masjedi
e al. (2006) conclude ha d y me ging o ellip ical galaxies is o ma ginal impo ance
om hei s udy o he co ela ion unc ion o SDSS luminous ed galaxies (LRGs).
Likewise, in hei s udy o he mass unc ions in he COSMOS su ey, Sca la a e al.
(2007) a gue ha he massi e end o he mass unc ion was in place by z= 1 and
hey ind no e idence o signi ican e olu ion. These con as ing esul s on he ecen
assembly his o y o he mos massi e galaxies in he uni e se a e likely in pa due o
he a e na u e o massi e galaxies, no allowing s a is ically signi ican s udies, and
in pa o di e en classi ica ion schemes used.
1.4.1 Me ge , pai o g oup?
In his wo k we make a b oad use o he wo d me ge when e e ing o he po en ial
me ge candida es ound as ou lined in Chap e 2. In ac , we can no say wi h
ce ain y ha all o he po en ial me ge s, mos ly composed by wo elemen s - hence
he ex ensi e use o he wo d pai s - in pa icula o he ones a highe p ojec ed
dis ances and eloci y o se s. As men ioned be o e, he me ging p ocess depends no
only on he e e -changing dis ance be ween he galaxies bu also on hei ela i e
eloci ies and, e.g., mass a io. Howe e , o simplici y, we will e e o po en ial
me ge candida es, i.e., he pai o g oup ound o be wi hin ou ini ial condi ions o
me ging, simply as me ge s, unless s a ed o he wise.
Also, mos o hese sys ems a e expec ed, and will u n ou o be, made up o wo
galaxies in i s majo i y bu some a e made up o mo e han wo elemen s and he wo d
pai would be misleading. E en hough in he li e a u e one can ind wide use o he
wo d pai when e e ing o a p obable me ge , he ex en o he me ge sample buil

1.5. THESIS OUTLINE 49
du ing he cou se o his wo k in oduced he need o e e o he po en ial me ge s
as g oups, since some o hese sys ems ha e indeed mo e han wo elemen s.
The e o e, we ha e made use o he wo d me ge o e e o he sys em - ei he a pai
o g oup - and no o he ac ual me ging s a us. We use his wo d h oughou his
p ojec unless, o cla i y he wo ds pai o g oup a e be e sui ed.
1.5 Thesis ou line
The main goal o his hesis is o unde s and he o e all ole o d y me ge s in he
assembly o massi e galaxies and econcile he di e en conclusions on he impo ance
o his ype o me ge s. To ha end, a sample o po en ial g oups o me ging galaxies
wi h spec oscopic obse a ions om he SDSS DR6 is cons uc ed, building he la ges
sample o d y me ge s o da e. Due o hei a e na u e, only a la ge se o d y me ge s
allows unde s anding hei con ibu ion o he assembly o new galaxies. In ac , he
access o spec oscopic in o ma ion makes he me ge sample unp eceden ed bo h in
numbe o po en ial me ge s and a ailable in insic da a o hese objec s.
How he me ge sample is de ined cons i u es a c i ical pa o his wo k and is ully
de ailed in Chap e 2. This Chap e also desc ibes he compa ison samples d awn om
he pa en sample, i.e. he SDSS DR6, and he gene al p ope ies ound in he sample
o po en ial me ge s. Chap e 3is dedica ed o he p ope ies o d y me ge s and hei
ole in he scena ios o galaxy assembly. The gas con en o galaxies h oughou he
me ge p ocess is explo ed in 4. The ollowing Chap e p esen s some indi idual d y-
me ge sys ems, om he me ge sample, in mo e de ail. Finally, Chap e 6con ains
he summa y o he conclusions om his wo k, las ema ks and some commen s on
he wo k ye o be done.
Chap e 2
Da a, Sample Selec ions and
Me ge Classi ica ion
The in e ac ion and me ging o galaxies is indeed a ema kable e en known, and
s udied, o se e al decades. Howe e , o unde s and how galaxies in e ac , and wha
exac ly happens du ing he in e ac ion and how much i a ec s he o e all assembly
and e olu ion o galaxies, one should in es iga e indi idual me ge s in de ail, whe e he
mos in insic aspec s o he me ge phenomenon can be explo ed bu also, whene e
possible, la ge se s o me ging galaxies in o de o ha e a b oade pe spec i e o
comp ehend he ole o his phenomenon and ge s a is ics in a signi ican way. In any
hie a chical model o s uc u e o ma ion, such as he cu en ly a ou ed ΛCDM,
me ge s a e ubiqui ous, bu dec easing in equency wi h cosmic ime, and in he local
uni e se hey a e ound o be qui e a e. So, in o de o s udy hei impac on galaxy
assembly, we would p e e a la ge sample o me ge s since obse a ions o me ge s
in he local Uni e se long ocused on small samples s udied in de ail a he han a
s a is ical s udy.
The eme gence o la ge dedica ed su eys, pa icula ly in he las decade, e.g. CFRS
(Lilly e al. 1995), SDSS (Yo k e al. 2000), 2dFGRS (Colless e al. 2003), has
p o ided subs an ial amoun s o da a, especially o he nea by uni e se. Samples
o galaxies po en ially unde going a me ge can be buil and s udied, e en a di e en
50
2.1. THE SLOAN DIGITAL SKY SURVEY 51
wa eleng hs, hus con ibu ing o inc ease ou knowledge o he p ocesses behind
galaxy assembly and e olu ion.
This Chap e desc ibes he da a used in his p ojec , including he o iginal su ey om
whe e i was d awn, he choices made in o de o assemble one o he la ges samples
o da e o galaxies po en ially unde going a d y me ge and some esul s conce ning
he o e all p ope ies o he me ge sample.
2.1 The Sloan Digi al Sky Su ey
The Sloan Digi al Sky Su ey p ojec (SDSS; Yo k e al. 2000)) is one o he majo
successes o mode n as onomy. Being he la ges su ey o he ex agalac ic sky hus
a unde aken, i co e s mo e han a qua e o he sky, ob aining pho ome y in
i e di e en bands (u,g, ,i and z) o 357 million unique objec s and spec oscopic
in o ma ion o mo e han 1 million objec s, in i s las Da a Release, DR7 (Abazajian
e al. 2009)1.
The SDSS su ey used a dedica ed 2.5 m elescope (Gunn e al. 2006), loca ed a
Apache Poin Obse a o y in New Mexico, in he Uni ed S a ed o Ame ica. Since
i i s s a ed o co e he sky i had wo phases o obse a ions, SDSS-I (2000-
2005) and SDSS-II (2005-2008), and du ing hose eigh yea s has p oduced se en da a
eleases(e.g Abazajian e al. 2009;Adelman-McCa hy e al. 2008,2006), he inal one
- Da a Release 7 (DR7) - was made a ailable in Oc obe 2008 (Abazajian e al. 2009).
The elescope is composed o wo ins umen s: o he imaging i had a d i -scan CCD
came a (Gunn e al. 1998) and a pai o spec og aphs, each wi h 320 op ical ibe s.
The pho ome y oo p in a ea co e ed 11663 squa e deg ees o he sky, wi h he o al
spec oscopic a ea co e ed being, in DR7, 9380 squa e deg ees. The spec oscopy da a
was ob ained wi hin a wa eleng h ange o 3800-9200 ˚
A and in a single un up o 640
spec a can be measu ed wi h each ibe ha ing a diame e o 3”.
As desc ibed in S ough on e al. (2002), he imaging da a was ga he ed i s and, a e
as ome ic and pho ome ic calib a ions ha e been applied, some selec ed objec s a e
lagged o spec oscopic ollow-up. They ha e hei spec a ex ac ed and calib a ed,
1h p://www.sdss.o g/d 7
52 CHAPTER 2. THE MERGER SAMPLE
and hei edshi in o ma ion and spec al ype de i ed in a la e s age. Based on
he pho ome y, a numbe o possible spec oscopic samples a e iden i ied: Main
Galaxy Sample, Luminous Red Galaxies (LRG), Quasa s and S a s, among o he s.
The Main Galaxy Sample comp ises objec s which a e selec ed o ha e an -band
Pe osian magni ude P≤17.77 a e co ec ion o o eg ound Galac ic ex inc ion.
In addi ion he e is a -band Pe osian hal -ligh su ace b igh ness cu a µ50 ≤24.5
mag a csec−2, as de ailed in S auss e al. (2002) bu we will ocus on ( ai ly) high
su ace b igh ness objec s he e so his la e cu is no o impo ance o ou discussion
he e. The Luminous Red Galaxies sample is acqui ed h ough selec ed cu s in he (g- ,
-i, ) colo -colo -magni ude cube and co ec ed o Galac ic ex inc ion. The de ails
o selec ion o he LRG sample a e desc ibed in Eisens ein e al. (2001), howe e his
sample o galaxies is no ele an o he p esen wo k.
The SDSS da a a e sepa a ed in o a da a a chi e se e DAS2and a ca alogue a chi e
se e . The o me s o es, among o he hings, he FITS images, spec a and objec
de ec ions. The spec oscopic da a e e ed o a e based on analysis o he FITS
spec a om he DAS analysed as desc ibed in T emon i e al. (2004) and B inchmann
e al. (2004). The Ca alog A chi e Se e Jobs Sys em, o CasJobs3, allows o sea ch
h ough he SDSS da abase, using SQL. The Pho oObjAll and he SpecObjAll ables
con ain, espec i ely, he cha ac e is ics o each pho ome ic objec and he measu ed
pa ame e s o he spec um o he objec s 4.
The spec oscopic pipelines (Schlegel e al. 1998;SubbaRao e al. 2002) p oduce
a classi ica ion o he spec a s o ed in he specClass pa ame e . The ca ego ies
used a e: quasa , high- edshi quasa , s a , la e- ype s a , galaxy o unknown. Fo
galaxies, in pa icula , a quan i a i e spec oscopic classi ica ion is o e ed by he
eClass pa ame e . This is de ined as in Connolly e al. (1995) o be he a io o
he i s wo eigencoe icien s o a gi en spec um. Connolly e al. (1995) ind i o be
a use ul spec oscopic classi ie , in e ms o ea ly- o la e- ype galaxies, al hough as
discussed by Yip e al. (2004) i is no a comple e desc ip ion o he galaxy spec a.
In he p esen wo k, due o he scheduling o he wo k plan, he ca alog o spec o-
scopically con i med galaxies used in he i s s age was he DR4 elease (Adelman-
2h p://das.sdss.o g
3h p://casjobs.sdss.o g
4h p://cas.sdss.o g
2.1. THE SLOAN DIGITAL SKY SURVEY 53
McCa hy e al. 2006), wi h 567 486 galaxies. In a second s age, i was upda ed wi h
he DR6 elease (Adelman-McCa hy e al. 2008), con aining 796 912 galaxies. This
da a elease has an imaging oo p in a ea o 9583 squa e deg ees and a spec oscopic
a ea o 7425 squa e deg ees, and he pho ome ic ca alog encloses mo e han 287
million unique objec s.
E en hough his s udy is mos ly ocused on he spec oscopic ca alogue, we also make
use o he pho ome ic sample om he SDSS DR6, o ensu e spa ial comple eness.
The spec oscopic a ge ing in he SDSS is based on a sophis ica ed iling algo i hm
desc ibed in Blan on e al. (2003) and o imp o e spec oscopic co e age in dense
egions hese a e isi ed mul iple imes. Tha no wi hs anding he e is some spa ial
incomple eness, pa icula ly likely o la e-s age me ge s, and we complemen he
spec oscopic ca alogue wi h da a om he pho ome ic ca alogue o add ess his
incomple eness. Howe e he e will o cou se be some incomple eness a e y la e
s ages in he me ging p ocess when he galaxies ha e used almos comple ely. This can
o some ex en be add essed by looking a ain mo phological ea u es ( an Dokkum
2005;Da g e al. 2010). This echnical issue is o pa icula impo ance i one is ying
o look o la e s age me ge s. I he galaxies a e e y close oge he hey would no
be conside ed because he ibe usually pinpoin s he cen al egion o he galaxy, i
wo galaxies happen o ha e hei cen al egions e y close, as i is he case in e y
la e s age me ge s, hese would be missed due o he es ic ions in ibe placemen .
He e we will no aim o ollow me ge s o comple e usion and hus a oid his issue.
The edshi dis ibu ion o he SDSS galaxy sample is shown in Figu e 2.1, whe e i
is possible o g asp he ex en o edshi s co e ed by he DR6 elease o SDSS. E en
hough he e a e measu emen s o edshi s up o z=1, he peak is clea ly a ound
z≃0.1, co esponding o he main galaxy sample. The second peak, occu ing o a
edshi close o 0.35 is he esul o he p esence o he LRG sample.
Since his s udy ocuses on he low- edshi uni e se, he i s condi ion applied o he
main sample o spec oscopic galaxies om he SDSS was o se up an uppe limi o
he edshi . The main galaxy sample consis s p ima ily o galaxies wi h z≤0.2 and
we adop his as ou uppe edshi cu -o .

54 CHAPTER 2. THE MERGER SAMPLE
Figu e 2.1: The spec oscopic edshi dis ibu ion (in g ey) o he spec oscopic
galaxy sample o SDSS DR6, including he main galaxy sample and he LRG sample.
The black his og am e e s o he edshi dis ibu ion o he galaxy sample wi h he
cu s men ioned in Sec ion 2.1.1.
2.1. THE SLOAN DIGITAL SKY SURVEY 55
2.1.1 Fi s cons ain s on he main galaxy sample om he
SDSS
To accoun o spa ial incomple eness he SDSS e u ns o he same pa o he sky
mo e han once i possible. The e a e wo main classes o duplica e obse a ions: one,
which is easy o deal wi h is he e-obse a ion a essen ially he same posi ion o an
objec . This is done as pa o he quali y assu ance in he SDSS. The o he ype o
duplica e occu s because he pho ome ic objec de ec ion code, pa icula ly he ea ly
e sion, had a endency o spli la ge, nea by galaxies in o se e al pa s which we e
aken o be indi idual objec s and hence a ge ed simul aneously o spec oscopic
ollow-up (e.g. Blan on e al. 2005). These duplica e obse a ions a e no as easy o
iden i y as he e-obse a ions o he same objec s. Howe e , his is only a p oblem a
low edshi so a low- edshi cu helps educing he se iousness o his issue.
Bea ing his in mind, and o minimize he need o isual inspec ion, while keeping
he main pa en sample o galaxies as un ouched as possible, and o a oid he e ec
o signi ican de ia ions om he local Hubble low, we adop a lowe edshi limi o
z= 0.005.
The o e all quali y o SDSS da a is high, ne e heless, in he case o he objec s
wi h spec oscopic obse a ions, a scienceP ima y lag was de ined o assu e he
quali y o he da a. To be a scienceP ima y an objec has o ul ill a se o i e
condi ions, elying pu ely on spec oscopic and geome ic conside a ions, ega ding
he spec oscopic obse a ion o a gi en objec . In such a case, he objec will ha e
scienceP ima y = 1. E en hough he da abase includes all spec oscopic objec s,
unde specObjAll, i also con ains a sepa a e able o objec s wi h scienceP ima y
= 1 s o ed in he specObj able. This condi ion o scienceP ima y = 1 was also a
equi emen in he selec ion o galaxies making up he pa en sample om which he
pai sample would be ex ac ed.
•Objec s wi hin a spec oscopic edshi in e al o 0.005 < z < 0.2
•Objec s lagged as science p ima y, i.e. wi h scienceP ima y = 1.
The inal pa en sample, om which he pai sample is buil , con ains 549 572 galaxies
wi h he edshi dis ibu ion shown in Figu e 2.1. The dis ibu ion o a subse o his
sample is shown in Figu e 2.2.
56 CHAPTER 2. THE MERGER SAMPLE
Figu e 2.2: The galaxy dis ibu ion o a subse o he inal sample om whe e he
pai sample will be buil . The galaxies in his plo a e wi hin he edshi in e al
men ioned in his sec ion, howe e , o simplici y, only galaxies wi h s ella masses
log(M∗/M⊙)>11 and a an absolu e declina ion below 10 deg ees a e plo ed.
2.2. SAMPLE SELECTION 57
2.2 Sample Selec ion
I is possible o ind se e al di e en pai c i e ia ac oss he li e a u e, as summa ized in
Table 2.2. F om mo phological signa u es iden i ied di ec ly in he in e ac ing sys em
(e.g. an Dokkum 2005;Bell e al. 2006) o measu emen s o p oximi y be ween he
galaxies ha a e po en ially in e ac ing, he me hods a y conside ably. Conce ning
he galaxies sepa a ion in he plane o he sky, i.e. p ojec ed sepa a ion, he no m
has been o de ine 20 h−1kpc as he maximum dis ance o a close pai . Acco ding
o heo e ical s udies, hese pai s a e expec ed o me ge wi hin 0.5 Gy (e.g. Ba nes
1988;Pa on e al. 1997,2000).
In his wo k, aking ad an age o he SDSS DR6 main spec oscopic sample o galaxies,
wo o mo e galaxies a e conside ed a po en ial me ge i hey me a se o p oximi y
condi ions. A pai , o g oup, o galaxies is conside ed o be a me ge candida e i he
ollowing condi ions a e me :
•P ojec ed sepa a ion in he plane o he sky - each neighbou galaxy has o
be wi hin a co-mo ing sepa a ion de ined by a ci cle o adius p≤30 h−1kpc,
consis en wi h o he wo ks.
•Veloci y o se along line o sigh - he di e ence be ween he adial eloci y o
he neighbou galaxy and he sea ched galaxy should no be mo e han sep ≤
800 km/s, equi alen o a di e ence in edshi o ∆z≈0.00267.
The eloci y o se limi is la ge wha is han commonly used in he li e a u e and will
possibly include some non-me ging pai s. The a e age eloci y o se o he membe s o
each g oup is illus a ed in he lowe panel o Figu e 2.3. Abo e a eloci y sepa a ion
o 600 km/s he numbe o g oups emains app oxima ely he same, wi hou signi ican
changes up o a eloci y di e ence o app oxima ely 1200 km/s. Following he same
conse a i e app oach, when compa ed wi h wha is used in he li e a u e, we op ed
o se he limi o he eloci y o se a 800 km/s.
The dis ibu ion o he galaxies po en ially unde going a me ge acco ding o ou
c i e ia, as a unc ion o p ojec ed sepa a ion and eloci y o se , is shown in Figu e
2.3.
64 CHAPTER 2. THE MERGER SAMPLE
Figu e 2.7: Spec oscopic edshi dis ibu ion o he galaxies om he me ge sample
in ol ed in majo me ge s.
The edshi dis ibu ion o he me ge sample is shown in Figu e 2.7. The lowes
edshi in he me ge sample is 0.005014, he highes is 0.199080 and he median
edshi is 0.066.
In he ollowing chap e we will ocus ou a en ion on gas-poo , i.e., d y, me ge s.
He e we will s udy he ull popula ion o me ging galaxies, bo h o s udy he inc ease
in hei s a o ma ion ac i i y and AGN con en , and o gi e a con ex o he es o
he hesis.
The ocus o cha ac e izing he me ge sample is, a his poin , wo olded:
•Classi ying he di e en ypes o me ge s in he sample. In pa icula ega ding:
–Gas con en , and/o s a o ma ion ac i i y, o he galaxies in he me ge
–Numbe o galaxies in ol ed in each me ge
–S ella mass a io be ween galaxies in ol ed in he me ge
•Examining he me ge sample wi h espec o pho ome ic ea u es only. By

2.2. SAMPLE SELECTION 65
analysing only he pho ome ic p ope ies, ha in o ma ion can la e be used
when adding he galaxies om he SDSS ha do no ha e any spec oscopic
da a.
I should be s essed a his s age ha e en hough he aim o his s udy is o
in es iga e he o e all p ope ies o me ge s be ween gas-poo galaxies, one should
emphasise he con inuous na u e o he possible di e en ypes o me ge s and indeed
he subjec i i y inhe en in designa ing a me ge ”d y” o ”we ”. Mo eo e , we make
use o he comple e sample, i.e. d y and non-d y me ge s, o s udy AGN and s a
o ma ion ac i i ies, as p esen ed in Sec ion 2.5, he las sec ion o his Chap e .
2.2.3.1 Pho ome ic p ope ies
Classi ying galaxies on he basis o hei pho ome ic p ope ies is a challenging p ob-
lem. In ac , e en he simple Hubble classi ica ion scheme o galaxies p esen s some
complexi y o iden i y a galaxy. Due o ou ocus on d y me ge s we a e pa icula ly
in e es ed in classi ica ion echniques ha can be used o sepa a e ea ly- ype galaxies
om o he galaxies and do no pu emphasis on how he classi ica ion wo ks on
la e- ype galaxies. A s aigh o wa d app oach is o selec galaxies by hei colo -
mo phology ela ion, since ea ly- ype galaxies end o be ed and bulge-domina ed,
ansla ing in o pho ome ic p ope ies like colo and concen a ion. Nex , hese
pa ame e s will be p esen ed and he esul s o applying hem o ou sample will
be analysed.
Concen a ion index A s anda d way o de ine ea ly- ype galaxies acco ding o
hei mo phology is he concen a ion index, C, which dis inguishes ellip icals om
spi als based on hei adially a e aged ligh dis ibu ions. Following S a e a e al.
(2001) and Shimasaku e al. (2001), he concen a ion index is de ined o be he a io
o he adii con aining 90% and 50% o he Pe osian lux in he -band, as s a ed in
equa ion 2.1.
C =R90
R50
,(2.1)
whe e C is he concen a ion index calcula ed o he -band and R90 and R50 a e
he -band Pe osian magni udes o R90 and R50, espec i ely.
66 CHAPTER 2. THE MERGER SAMPLE
The SDSS de e mines he magni udes in a modi ied Pe osian sys em, measu ing
galaxy luxes wi hin a ci cula ape u e which adius is de ined by he shape o
he azimu hally a e aged ligh p o ile (e.g. S ough on e al. 2002). The pho ome ic
pipeline calcula es he (ci cula ised) adii con aining, espec i ely, 50% (R50) and 90%
(R90) o he ligh o he i e ug iz bands.
S a e a e al. (2001) show ha he concen a ion index co ela es wi h galaxy ype,
specially o sepa a e ea ly- om la e- ypes. Cen ally concen a ed ellip icals a e
expec ed o ha e la ge concen a ion indices han spi als. Mo eo e , hey es ima e a
cu -o index o C = 2.63 o sepa a e ea ly- ype galaxies, wi h C >2.63, and la e-
ype galaxies o C <2.63.
Ini ially we conse a i ely conside ed se ing he limi o he concen a ion index o
C >2.8. This would appea o be a good c i e ion o sepa a e ea ly- and la e-
ype isola ed galaxies. Howe e we wished o check whe he i was eliable also o
close pai s o galaxies. When dealing wi h po en ially me ging sys ems, depending on
he sepa a ion be ween he elemen s o an in e ac ing g oup, he concen a ion index
p o ed no o be eliable since he adii, in pa icula R90, can be s ongly a ec ed by
he p esence o a neighbou , i can be seen in Figu e 2.8.
The concen a ion index o he galaxies in he me ge sample a i icially d ops o
lowe alues o galaxies ha ha e a neighbou close han 10h−1kpc, un easonably
indica ing hese galaxies would p e e en ially ha e less concen a ed ligh p o iles and
demons a ing ha he ligh concen a ion o galaxies is only a sui able quan i a i e
measu e o mo phology when he objec s a e isola ed. This ea u e can be u he
analysed in Figu e 2.9 whe e i is illus a ed he di e ence be ween R90 and R50 o
hos galaxies when compa ed wi h a se o isola ed galaxies wi h simila mass and
edshi ( ∆(z) ≤0.01 and ∆(Log(M∗/M⊙)≤0.1). The di e ence in bo h R90 and R50
shows an inc ease o hese quan i ies wi h dec easing p ojec ed sepa a ion, indica ing
ha he Pe osian lux measu ed in he wo egions is al e ed o galaxies in me ge s,
wi h hos s galaxies p esen ing somewha highe alues. The ou e adius, R90, is mo e
a ec ed han he inne R50 adius, indica ing ha he ou ski s o galaxies become
a ec ed soone han he mo e cen al egions. This g ea e di e ence o R90 when
compa ed o R50 esul s in he dec ease o he concen a ion index seen in Figu e 2.8,
make his c i e ion pa icula ly weak when applied o galaxies in me ge s.
2.2. SAMPLE SELECTION 67
Figu e 2.8: Median alue o he concen a ion index pa ame e as a unc ion o
p ojec ed sepa a ion, o he galaxies in he me ge sample. The uppe and lowe
dashed lines a e he 75 h and 25 h pe cen iles, espec i ely. The ho izon al line
indica es he C =2.63 limi .
68 CHAPTER 2. THE MERGER SAMPLE
Figu e 2.9: The median o he a ia ion o he di e ence in R90 and R50 o galaxies in
me ge s and isola ed galaxies. The shaded egions indica e he 68% con idence in e al
ob ained wi h boo s apping echniques. The subsc ip nindica es he 90, i.e. R90
ep esen ed in ligh g ey, o 50, i.e. R50, ep esen ed in da k g ey.
2.2. SAMPLE SELECTION 69
Figu e 2.10: Eccen ici y his og am o he galaxies in he me ge sample. O e -
plo ed, o e e ence, he e ical line a E = 0.7 indica es he en a i e limi ini ially
ied o es ablish a sepa a ion be ween di e en mo phological classes (sphe oids’
eccen ici ies would lie abo e his limi ).
Eccen ici y Ano he possible pho ome ic p ope y ha could, o a i s app oxi-
ma ion, success ully sepa a e sphe oidal om edge-on disk galaxies, o be applied o
whe e edge-on spi al a e o en con aminan s like colo -selec ed samples o ea ly- ypes,
is he eccen ici y, as de ined in equa ion 2.2.
E =b
a
(2.2)
In his equa ion, a and b ep esen he isopho al majo and mino axis, espec i ely,
in he -band. The i s a emp o a sepa a ion be ween he wo ypes was se a
E = 0.7. Objec s wi h E >0.7 would be conside ed sphe oidal and would ep esen
a good selec ion o ea ly- ype candida es. Howe e , a s udy o he dis ibu ion o
he eccen ici y, shown in Figu e 2.10, o he galaxies in he me ge sample, e ealed
ha he eccen ici y did no signi ican ly imp o e ou abili y o dis inguish sphe oidals,
since he e was no sepa a ion be ween he wo ypes o objec s. This happens p obably
o he same eason as in he concen a ion index esul s: he pho ome y o each

70 CHAPTER 2. THE MERGER SAMPLE
galaxy in a pai , o g oup, is s ongly a ec ed and classi ying indi idual galaxies using
pho ome ically de i ed s uc u al pa ame e s is hinde ed by hei in e ac ing s a us.
Colo A low edshi , using colo as a de ining pa ame e o ea ly- ype iden i i-
ca ion migh nai ely iewed as s aigh o wa d, since hese galaxies a e expec ed o
ha e old s ella popula ions easily cha ac e ised by hei ed colo s (di Se ego Alighie i
e al. 2006;Renzini 2006). Howe e , using solely colo o de ec hese galaxies would
likely con amina e ou sample wi h objec s eddened by dus hus in oducing alse
posi i es: along wi h sphe oidal mo phologies wi h old s ella popula ions, he sample
would con ain dus eddened galaxies. Howe e i u ns ou ha his is a mo e use ul
echnique and we will discuss his in mo e de ail in sec ion 2.2.4.3 below.
Gi en he analysis p esen ed abo e, and o assess he con amina ion induced by colo
selec ion and cons uc a e e ence sample o me ge s, a isual classi ica ion o he 3962
po en ial me ge s was ca ied ou , ensu ing he cohe ence o he se e al c i e ia used
o dis inguish be ween di e en ypes o me ge s and o he e en s ac oss he sample.
2.2.4 Visual classi ica ion o he me ge sample
The isual classi ica ion o he me ge sample was pe o med in o de o asce ain he
bes possible quali y o ou me ge classi ica ion me hod since, as shown in Sec ion
2.2.3, a secu e classi ica ion would no be possible i using solely a se o selec ed
pho ome ic pa ame e s.
To ecap, we iden i ied pai s o galaxies using a spa ial sepa a ion c i e ion o 30 h−1kpc
and 800 km/s. F om his pa en sample we now wan o iden i y he ea ly- ype
galaxies.The classi ica ion was done o each g oup o he me ge sample conside ing
he g oup as a whole and no jus he indi idual galaxies making up he me ge . The
adop ed c i e ia consis ed in looking o he ollowing ea u es in each o he galaxies
o he me ge :
1. he p esence o a disk
2. any blue colo s
3. a spi al o clumpy pa e n
2.2. SAMPLE SELECTION 71
4. he clea p esence o a dus lane
5. a adially asymme ic dis ibu ion o ligh
The ca ego iza ion o he g oups was made in he ollowing manne :
•i all galaxies in he me ge did no show any o hese ea u es, he g oup would
be classi ied as a d y me ge .
•i any galaxy in he me ge displayed any o hese ea u es, he g oup would be
conside ed a non-d y me ge .
This sys em implici ly conside s me ge s be ween ea ly- and la e- ype galaxies, he so
called damp o mois me ge s, as being non-d y me ge s.
The isual classi ica ion was ca ied ou using he jpeg images om he SDSS DR6
websi e 5and no o he pa ame e o ool was used. In pa icula , we did no make
use o he spec oscopic da a o he galaxies in making he classi ica ion.
In p ac ice, o ca y ou he classi ica ion we i s c ea ed an HTML ile wi h access o
he JPG image o each galaxy cen e ed on ha galaxy, bu g ouped in he g oups/pai s
we had p e iously iden i ied. Thus du ing he classi ica ion we could make e e ence
o en i e g oup. The e ie ed images a e cen e ed a a gi en pai o igh ascension
(RA) and declina ion (DEC), ha e a size o 70 ×70 pixels and a e scaled o 0.2”/pix,
p oducing an image cen e ed on a unique objec . The linked images o each objec
in a g oup allowed an analysis o he indi idual galaxies in each g oup sepa a ely
bu he classi ica ion was ca ied ou as a g oup, as desc ibed abo e. An impo an
ad an age o his isual classi ica ion is he possibili y o use i o in es iga e he a e
o con amina ion in colo -selec ed d y me ge samples.
The isual classi ica ion esul ed in a sample o 890 d y me ge s and 3072 non-d y
me ge s, ansla ing in o 1801 galaxies in d y me ge s, he ea e d y galaxies, and
6227 in non-d y me ge s, he ea e non-d y galaxies. To no e ha hese e ms, o
”d y” and ”non-d y” galaxies, a e solely used because hey belong o d y and non-d y
me ge s, espec i ely, and hey do no e e o hei pa icula na u e.
5h p://www.sdss.o g/d 6
72 CHAPTER 2. THE MERGER SAMPLE
Figu e 2.11: g- colo - edshi diag am o he galaxies in he pa en sample o
spec oscopic galaxies o he SDSS DR6. The bimodali y o he galaxy popula ion is
clea ly e iden a low edshi bu owa ds highe edshi he cha ac e is ic luminosi y
shi s owa ds mo e massi e, and edde , galaxies and he blue sequence ades away.
2.2.4.1 Cha ac e is ics o he isually classi ied me ge s
A med wi h he isually classi ied sample o po en ial me ge s, we can now s udy how
ou isual classi ica ion di e s om o he echniques o me ge iden i ica ion.
The colo - edshi diag am o he galaxies in he main SDSS DR6 sample is shown
in Figu e 2.11 o e e ence and in Figu e 2.12 he classi ied galaxies a e o e plo ed.
As can be seen in Figu e 2.12, he posi ion o d y galaxies is well cons ained in he
colo - edshi diag am, placed, as expec ed, in he ed-sequence, whe eas he non-d y
galaxies show a much wide sca e and e en some o e lapping owa ds he ed-do ed
egion o he d y galaxies ( ecall ha we do no make use o an explici colou -cu
when ca ying ou he isual classi ica ion).
The same beha iou can be no iced in he colo -magni ude diag am in Figu e 2.14,
whe e u- colo is plo ed agains he -band absolu e magni ude. The dis ibu ion o
2.2. SAMPLE SELECTION 73
Figu e 2.12: g- colo - edshi diag am o he galaxies in he me ge sample (ci cles)
plo ed o e he pa en sample o spec oscopic galaxies o he SDSS DR6 (in g ay
scale). The ed illed ci cles (uppe panel) ep esen he galaxies in d y me ge s and
he blue illed ci cles (lowe panel) he galaxies in non-d y me ge s.
80 CHAPTER 2. THE MERGER SAMPLE
blue galaxies inside he lane would be less hen 50%. The closes op imized solu ion
is a comp omise be ween ha ing less con amina ion bu less ed galaxies o he exac
opposi e. Because he aim is o cons ain a egion o ed galaxies, he la e solu ion
was chosen. The se o ( , in) gi ing he op imized solu ion wi h hese condi ions is
he pai o limi s whe e he uppe one con ains 99 % o he galaxies and he lowe
limi encloses 7.5 %. In he egion wi hin hese limi s, he ac ion o blue galaxies is
52.7 % whe eas he numbe o ed galaxies inside ep esen 91.5 % o he o al numbe
o ed galaxies.
We hen apply his selec ion cu o he en i e sample o po en ial me ge s wi h
Log(M∗/M⊙)⩾9. This lowe limi in s ella mass was adop ed since he comple eness
limi in mass o an old, una enua ed galaxy a he low edshi limi , z = 0.005, o
he SDSS is Log(M∗/M⊙)≈8.6.
Wi h hese limi s o he lane and he mass cu sample o galaxies, he equa ions
de ining he ed egion a e gi en by equa ion 2.3 and equa ion 2.4 and he inal lane
is illus a ed in Figu e 2.17.
(g− )uppe = (0.594z2) + (1.622z) + 1.827 (2.3)
(g− )lowe = (−1.445z2) + (3.024z) + 0.594 (2.4)
Inside his lane, he numbe o d y me ge s is 733 (ins ead o he p e ious 890) whe eas
he numbe o non-d y me ge s d ops o 703 (when compa ed wi h he ini ial 3072).
This change in he me ge numbe s e ealed he signi icance o es ablishing a limi
when sea ching o d y me ge s. Mo eo e , he special ole o his ed egion is clea
when adding he pho ome ic sample o galaxies om he SDSS DR6. Below we will
adop hese equa ions o iden i y po en ial d y me ge s om he pho ome ic sample.
By aking his app oach, complemen ed by a subsequen isual classi ica ion, we ensu e
ha he selec ion o d y me ge s is as consis en as possible in he spec oscopic and
pho ome ic samples.
2.2.4.3 Non-d y me ge galaxies in he lane
The compa ison done abo e be ween colo s and he isual classi ica ion highligh s he
p oblems o using colou alone o selec d y me ge s, as men ioned p e iously in 2.2.3.1.

2.2. SAMPLE SELECTION 81
Figu e 2.17: g- colo - edshi diag am showing he uppe and lowe limi s (in
g een) o he lane de ined acco ding o equa ions 2.3 and 2.4. G ey poin s indica e all
galaxies, a e he mass cu , in he me ge sample, yellow poin s ep esen he galaxies
classi ied as being in d y me ge s and ci cled yellow poin s a e he galaxies in d y
me ge s and inside he lane.
82 CHAPTER 2. THE MERGER SAMPLE
We con i med ha such a choice, i.e. o using only colo as a selec ion c i e ion, would
lead o o e es ima ing he numbe o d y me ge s due o he p esence o in e lope s
( ha a e no bona ide d y me ge s) wi hin ou es ablished colou limi s. Wha is he
eason o his s ong con amina ion o he ed sequence by wha we deem non-d y
galaxies? One possible eason migh be eddening by dus - o es his we use he
ela i e s eng hs be ween he Hαand Hβemission line luxes o p o ide an es ima e
o he dus a enua ion wi hin he egion sampled by he ib e.
The a io be ween he Hαand Hβ luxes, as shown in equa ion 2.5, has been widely
used as a p oxy o dus es ima ions.
 Hα
Hβobs
= Hα
Hβ0
e−[τ(λHα)−τ(λHβ)] (2.5)
Assuming a simple a enua ion law (e.g. Cha lo & Fall 2000) o accoun o he e ec s
o dus in he ISM o he galaxies o each line,
τ∝λ−0.7(2.6)
whe e τis he op ical dep h a λwa eleng h, and a ” e e ence” alue o he in insic
ela ion be ween Hαand Hβ, he Case B ecombina ion (B ocklehu s 1971),
 Hα
Hβ0
= 2.86 (2.7)
i is possible o es ima e how much ligh has been a enua ed by dus .
The isual e alua ion o he me ge sample ca ied ou in sec ion 2.2.4 classi ied
as a non-d y g oup any sys em ha ing a leas one galaxy showing any o he i e
cha ac e is ics lis ed in ha sec ion. This leads o some o ou non-d y g oups ha ing
also ed and dead galaxies, so, o a ai compa ison wi h ou d y me ge galaxies, we
only selec ed he dus ies galaxy o each non-d y me ge (analogously o using only
he blues one o he compa isons pe o med in he p e ious sec ion). The lux a io
o Hα o Hβis plo ed in Figu e 2.18 and i shows a peak o he d y me ge sample
a 2.9, whe eas he peak o he dus ies galaxies o he non-d y me ge sample occu s
o highe alues, a ound 4, sugges ing he ex inc ion by dus is in ac s ong o he
2.2. SAMPLE SELECTION 83
Figu e 2.18: No malized dis ibu ion o he Hα/Hβ lux a io o he galaxies inside
he colo lane. The solid black his og am conce ns all galaxies in d y me ge s and he
dashed his og am e e s o he dus ies galaxies in non-d y me ge s.
84 CHAPTER 2. THE MERGER SAMPLE
non-d y me ge galaxies wi h ed colo s. Tha p edominance is p esen o e en highe
alues o Hα/Hβ, as e ealed by a compa ison be ween he ail o bo h dis ibu ions:
only one i h o he d y me ge galaxies wi h Hαand Hβemission lines all abo e
3σo he mean alue o he espec i e dis ibu ion, while in he case o he dus ies
galaxies in non-d y me ge s, he co esponding ac ion goes up o mo e han a hal
o he o al sample. This con i ms ha he con amina ion by non-d y me ge s in he
ed lane is in ac due o dus eddened galaxies.
2.3 Accoun ing o spec oscopic incomple eness -
he pho ome ic ca alogue
In o de o comple e he sample o me ge s d awn in sec ion 2.2.3, he SDSS DR6
pho ome ic ca alog o galaxies was added. The i s s age was o e ie e he comple e
pho ome ic ca alog o galaxies, classi ied as such om he Pho oP ima y h ough
pho ome ic pa ame e s. Due o he way obse a ions a e ca ied ou , he same objec
can be obse ed mo e han once and he Pho oP ima y gua an ees ha only he main
obse a ion o he objec is being conside ed. Simila ly o he spec oscopic ca alog,
he selec ed objec s include all galaxies wi h Galac ic ex inc ion-co ec ed Pe osian
magni ude b igh e han 17.77.
The o al numbe o e ie ed objec s, om he SDSS DR6 pho ome ic ca alog o
galaxies, was abo e 1.2 million. No e, howe e , ha his se o objec s also includes
galaxies wi h spec oscopic obse a ions, meaning he ac ual numbe o ”new” galaxies
o be analyzed, and la e added o comple eness, is oughly hal .
2.3.1 New sea ch o neighbou galaxies
The adop ed p ocedu e o ind new po en ial me ge s is, in essence, alike he sea ch
ca ied ou be o e using only galaxies wi h spec oscopic measu emen s. In ac ,
o assu e cohe ence and comple eness, he sea ch o new pai s o g oups is made
exclusi ely a ound galaxies wi h spec oscopic measu emen s (he ea e , spec oscopic
galaxies). Howe e , his ime, he olume a ound each galaxy o be explo ed o
neighbou s comp ises no only spec oscopic galaxies bu also galaxies wi h me ely
pho ome ic quan i ies (he ea e , pho ome ic galaxies).
2.3. THE PHOTOMETRIC CATALOGUE 85
We ecall ha he edshi in e al o he spec oscopic galaxies was se o be 0.005
< z < 0.2. The lack o a spec oscopic edshi s o he pho ome ic galaxies poses a
p oblem since he pho ome ic edshi s, al hough a ailable, do no p o ide su icien
disc imina ion because he unce ain y is la ge han he edshi sepa a ion adop ed
(e.g. Abazajian e al. 2009). To minimize he impac o his obs acle and ake
ad an age o he new se o galaxies, he ollowing me hod o sea ch o neighbou
galaxies was adap ed.
In summa y, a ew in e media e s eps we e included in he p e ious me hod o building
a olume wi h he p ojec ed sepa a ion and he spec oscopic edshi a ound each
a ge galaxy, so ha we now p oceed as ollows.
1. Apply p ojec ed dis ance c i e ion
2. Assign new edshi o neighbou pho ome ic galaxies
3. Apply p ojec ed dis ance c i e ion and eloci y o se
S ep numbe 1. consis ed in sea ching o neighbou galaxies using only he p ojec ed
dis ance c i e ion. Iden ically o he p ocedu e in sec ion 2.2.2, a ci cle o adius 30h−1
kpc, a he edshi o he spec oscopic a ge galaxy, was de ined and all neighbou
galaxies we e s o ed.
The ollowing s ep was o assign a new edshi o he pho ome ic galaxies acco ding
o he ollowing:
•I he pai , o g oup, has mo e han one spec oscopic galaxy, he pho ome ic
galaxies will be assigned he spec oscopic edshi ha is closes o he hei
pho ome ic edshi .
•I he pai , o g oup, has only one spec oscopic galaxy, he pho ome ic galaxies
will acqui e i s edshi .
I he assigned edshi di e s by mo e han 2σ om he pho ome ic edshi es ima e
he pho ome ic galaxy is excluded om belonging o he po en ial g oup. We chose
2σsince, in his case, he hypo hesis ha his objec is a he same dis ance as he

86 CHAPTER 2. THE MERGER SAMPLE
nea by objec appea s unlikely. He e σis he unce ain y es ima e o he pho ome ic
edshi .
The las s ep consis s in ep oducing wha was al eady done in sec ion 2.2.2, his ime
using p ojec ed sepa a ion and eloci y o se , since e e y galaxy now has a alid,
and con olled, edshi . No e, ha o cou se wi h he app oach we canno make
gene al s a emen s abou he eloci y sepa a ion o he membe s o he g oup because
a numbe o hem will ha e ze o eloci y sepa a ion. No e also ha while pho ome ic
es ima es a e inaccu a e o galaxies wi hou clea ea u es in hei spec a, hey a e
easonably good o ea ly- ype galaxies which is wha we need o ou d y me ge
sample.
By cons uc ion, cons ain s on he p ojec ed dis ance we e al eady ensu ed. In
addi ion we equi e ha he edshi o he spec oscopic galaxy and he pho ome ic
edshi ag ee wi hin 2σ. In o he wo ds ha |zspec −zpho o| ≤ 2∗σz, whe e σzis he
es ima ed unce ain y on he pho ome ic edshi . Wi h his wo c i e ia o p ojec ed
dis ance and eloci y o se , he o al numbe o galaxies in he new sample o me ge s
is o he o de o 3 ×104.
Ha ing a sample o po en ial me ge s wi h bo h pho ome ic and spec oscopic edshi
in o ma ion, allowed us o explo e he e ec i eness o building a me ge sample using
only pho ome ic edshi s. To ha end, we ied o e ie e he me ge sample
men ioned in Sec ion 2.2.3 using now he pho ome ic edshi s o he galaxies. The
p ojec ed dis ance be ween galaxies was al eady assu ed and, in addi ion, he eloci y
o se was calcula ed using he pho ome ic edshi s o he galaxies wi hin a 1σin e al
o hei pho ome ic edshi s. Using his c i e ion, only 1891 g oups we e eco e ed,
in a clea con as wi h he 3962 g oups p e iously ound - less han 50% o he
g oups we e e ie ed. E en hough pho ome ic edshi s ha e been widely used o
in es iga e po en ial me ging sys ems, his di e ence shows he weakness behind his
ype o edshi s, mainly when dealing wi h galaxies ha a e so close oge he whe e
he accu acy o he measu emen s can become c ucial.
2.3.2 A comple e me ge sample
The new sea ch o neighbou s including galaxies wi h pho ome ic da a only, in-
c eased, as expec ed, he numbe o po en ial me ge s. The numbe o galaxies in
2.3. THE PHOTOMETRIC CATALOGUE 87
me ge s is now 36905 in a o al o 15101 pai s o g oups ( o be con as ed wi h he
numbe s o Sec ion 2.2.3, namely 8028 galaxies a anged in a o al o 3962 g oups).
No e ha his inc ease in he numbe o me ge s is g ea e han he inc ease in he
o al numbe o galaxies, 3962 g oups o 15101 om adding he galaxies wi hou
spec oscopic obse a ions. This means ha by no conside ing galaxies wi hou
spec oscopic measu emen s, a signi ican numbe o po en ial me ge s was being
neglec ed due o he p oblem o ibe placemen .
As men ioned be o e, one o he main goals o ha ing de ined a lane in a colo - edshi
diag am whe e, acco ding o ou c i e ia, he galaxies cons i u ing d y me ge s should
all, is o make use o ha egion o ind new po en ial d y me ge s, his ime using
he SDSS DR6 pho ome ic ca alog. This colo cu in he me ge sample is aimed a
iden i ying d y me ge s and, as done be o e, each g oup is conside ed a candida e d y
me ge i all galaxies o ha g oup a e inside he lane, de ined p e iously by equa ions
2.3 and 2.4. This colo cu was p eceded by a cu o in he -band whe e objec s
wi h ex inc ion co ec ed -band magni ude < 14.5 we e disca ded since hey we e
p obably s a s and we e assumed o ha e been misclassi ied as galaxies. I should be
s essed a his poin ha using he colo lane only selec s candida e d y me ge s. The
inal s a us as a d y me ge depends, as i was he case o he spec oscopic me ge
sample, on a isual classi ica ion.
Applying he colo cu esul ed no only in a educ ion on he numbe o me ge s
bu also allowed o ocus owa ds d y me ge s alone. The numbe o g oups whe e all
galaxies a e inside he lane is 3721, app oxima ely 25 % o he o al numbe o me ge s.
In compa ison o he me ge sample wi h only spec oscopic galaxies (sec ion 2.2.4.2),
whe e he ac ion o d y me ge s o e he o al numbe o me ge s is sligh ly o e 18
%, his inc ease shows ha e en hough he lane allowed o signi ican ly ocus on d y
me ge s, i is s ill no su icien o de ine one. To ha inc ease con ibu e:
• he con amina ion om non-d y me ge s - a his s age, he e is no sepa a ion
be ween d y and non-d y me ge s om a isual classi ica ion.
• he assigned edshi s - a ibu ing a edshi may o ce he inclusion o galaxies in
he lane, i.e. in he colo - edshi diag am. A pho ome ic galaxy ini ially lying
ou side he lane acco ding o i s pho ome ic edshi , may be shi ed along he
88 CHAPTER 2. THE MERGER SAMPLE
edshi axis, e en ually alling wi hin he lane once assigned he edshi o i s
spec oscopic neighbou .
• he lack o a mass cu - all objec s a e conside ed ega dless o hei s ella mass,
since, a his s age, he e is no in o ma ion on he s ella mass o he pho ome ic
galaxies.
The ollowing s age, isually classi ying he me ge s, is analogous o wha was done
wi h he spec oscopic me ge sample and will allow o, once again, sepa a e d y om
non-d y me ge s using he c i e ia desc ibed in sec ion 2.2.4.
Due o he abo e men ioned p oblems o conside ing galaxies wi hou spec oscopic
in o ma ion, i is no possible o ep oduce en i ely he me hod, and analysis, o he
me ge sample con aining pu e pho ome ic galaxies. As we ha e seen, he use o
pho ome ic edshi can lead o unsa is ac o y esul s o indi idual me ge s. The
main ad an age o isual classi ying he new sample o me ge s is s a is ical, allowing
us o de i e a co ec ion ac o o apply o he spec oscopic sample.
2.3.3 Visual classi ica ion o he pho ome ic me ge sample
Visual classi ica ion was ca ied ou o all he g oups inside he lane, ollowing he
same me hodology desc ibed in sec ion 2.2.4. The numbe o new d y me ge s is o
461, ha will add o he 733 p e iously ob ained wi h he spec oscopic sample alone.
The e o e, a o al 1194 g oups which is an inc ease o 63% ela i e o he p e ious
numbe o 733. This implies o ins ance ha using he spec oscopic sample would
eco e only abou hal o he eal numbe . This is clea ly o impo ance when you
wan o calcula e mass unc ions o ins ance. No e howe e ha ou analysis does
supp ess any likely dependence on local densi y. This could ha e some impo ance
a he e y mos massi e end o he mass unc ion as he e y mos massi e galaxies
eside in e y high densi y egions.
We should no e ha he e a e also some o he conce ns ela ed o his - wha ex-
ac ly cons i u es a new d y me ge was also aken in o conside a ion. By adding
he pho ome ic galaxies, i was expec ed ha me ge s al eady ” ound” when he
spec oscopic me ge sample was buil , would ha e new neighbou s. Indeed, du ing
he isual inspec ion, ha was e i ied o be case o o e a hund ed d y me ge s.
Howe e , i was also no iced ha some d y me ge s we e showing alse new neighbou s
2.4. SAMPLES OF ISOLATED GALAXIES 89
- he pho ome ic coun e pa o an al eady exis ing spec oscopic neighbou was being
ea ed as a new elemen o he g oup despi e ou minimum lowe limi o p ojec ed
sepa a ion men ioned in Sec ion 2.3.1 . Using hose me ge s would in oduce alse
posi i es in he me ge s wi h N > 2 elemen s - e.g. a g oup wi h 2 galaxies would
ha e 3 elemen s wi h one o he galaxies p esen ing wo ma ches, one ela i e o he
spec oscopic da a and ano he one o he pho ome ic da a. The e o e, because his
p oblem was ound in a ela i ely small amoun o he me ge s when compa ed wi h
he o al numbe , and hose me ge s would be ep esen ed in he spec oscopic me ge
sample in any way, hey we e no conside ed as new.
2.4 Samples o isola ed galaxies
The bene i s o building an isola ed sample om he same pa en su ey a e many- old
- no only he pho ome ic and spec oscopic echniques and algo i hms a e applied in
he same manne , bu i also add esses an ape u e p oblem in oduced by he di e en
edshi s a which he galaxies a e. The egion o he galaxy ha he 3” ibe s sample
is, na u ally, a unc ion o edshi .
The e o e, a majo ad an age o using a la ge su ey such as he SDSS is he abili y o
build di e en se s o galaxies using he same pa en sample ( he SDSS DR6) and being
able o, in his case, compa e o he me ge sample as de ined in p e ious sec ions.
Du ing he cou se o his wo k, a sample o galaxies wi hou nea by neighbou s was
c ea ed using he p ocedu e desc ibed in Sec ion 2.2.2. When he sea ch o neighbou s
was ca ied ou among he 549 572 galaxies (see Sec ion 2.1.1), a ” lag” was used o
ma k he galaxies wi h neighbou s, which would make up he me ge sample, and
wi hou neighbou s, which ha e been named ”isola ed”. Consequen ly, an isola ed
galaxy would be any galaxy wi hou a companion galaxy wi hin a adius p≤30 h−1kpc
and a eloci y o se o sep ≤800 km/s.
The ma ched sample o isola ed galaxies was assembled using galaxies lagged as ha ing
no neighbou s, as p e iously men ioned, and wi h simila cha ac e is ics o he galaxies
in he me ge sample, namely, spec oscopic edshi and s ella mass. Fo each galaxy
in he me ge sample, a se o isola ed galaxies was ga he ed sa is ying he ollowing
condi ions:
96 CHAPTER 2. THE MERGER SAMPLE
Figu e 2.23: Composi e ( op) and s a o ming (bo om) ac ion in me ge s as a
unc ion o p ojec ed dis ance. The e o ba s a e de e mined using boo s apping
e-sampling echniques.

Chap e 3
P ope ies o D y Me ge s
E e since me ge s be ween gas-poo galaxies ha e been obse ed, in pa icula a
z <2, a new pa h o massi e galaxy assembly has been unco e ed. Howe e , he exac
ole o hese me ge s in building massi e galaxies is s ill unde g ea deba e and e en
hough some open ques ions ha e been se led - he e is now e idence ha galaxies
e ol ing passi ely since z=2 is no he only pa h o build he massi e ellip icals seen
in p esen -day - o he ques ions ha e been aised: how do d y me ge s di e om
non-d y me ge s in building up new galaxies? Do end-p oduc s o low edshi d y
me ge s show compa ible p ope ies wi h he local popula ion o ellip ical galaxies?
How many massi e galaxies in he local Uni e se ha e been assembled ia d y me ging?
This Chap e con ains he esul s o he analysis o he me ge sample buil om he
SDSS DR6 desc ibed in Chap e 2, in oducing a e y la ge sample o me ge s be ween
gas-poo galaxies. We s udy he p ope ies o his ype o me ge s and discuss i s
ele ance o he build-up o he high-mass end o he galaxy mass unc ion.
Th oughou we assume s anda d cosmology ΩM=0.3, ΩΛ=0.7, h=0.7.
97
98 CHAPTER 3. PROPERTIES OF DRY MERGERS
3.1 E olu ion o he s ella popula ion wi h ed-
shi
Being he building blocks o he s uc u e o he uni e se, galaxies a e howe e p ima -
ily de ined by hei s ella con en . The e o e, he dis ibu ion o galaxies wi h s ella
mass is a undamen al cons ain on models o galaxy o ma ion and e olu ion. This
was al eady emphasised in a classic pape by Cowie e al. (1996) who s udied he
p ope ies o galaxies as a unc ion o K-band luminosi y and coined he concep o
downsizing - s a s in mo e massi e galaxies end o ha e been o med ea lie and o e a
sho e pe iod o ime. Subsequen ly a la ge numbe o s udies ha e ollowed o explo e
he e olu ion o he s ella mass unc ion wi h edshi (e.g. Ma chesini e al. 2009)
and es ablished a signi ican g ow h in he s ella mass in galaxies wi h dec easing
edshi , amoun ing o a ac o o 2-4 om z= 1.5−2 o he p esen .
Going o sligh ly lowe edshi s he e is e idence ha much o he s ella mass in
massi e galaxies was assembled by z∼1 (e.g. Bundy e al. 2007;Cima i 2007;Cowie
& Ba ge 2008;Cima i 2009) and wi h a weak e olu ion un il z∼0. This compa es
well wi h he e idence om s ella popula ions in local massi e galaxies which indica e
ha he s ella con en o local massi e ellip ical galaxies was p edominan ly o med
be o e z∼1 (e.g. Thomas e al. 2005) al hough he p ocess o assembly o he s ella
mass and o ma ion o he s a s migh no occu a he same ime.
In gene al he majo i y o s ella mass is locked up in ed and old galaxies (B inchmann
& Ellis 2000;Hogg e al. 2002) also poin ing o an ea ly o ma ion epoch o hei s a s.
This has led some o conclude ha he massi e end o he mass unc ion was in place
al eady a z∼1 as a gued o ins ance by Sca la a e al. (2007) in hei s udy o
he mass unc ions in he COSMOS su ey. While he e idence ce ainly poin o a
majo i y o he massi e end o he mass unc ion being in place by z∼1, he exac
pe cen age is open o discussion, wi h Ab aham e al. (2007) a guing o abou 80%
based on he Gemini Deep Deep Su ey (GDDS, Ab aham e al.) while Pozze i e al.
(2007) inding abou 60% o he VIMOS Ve y Deep Su ey (VVDS; Le Fe e e al.
2003). Thus he e ce ainly appea s o be a need o iden i y channels o he assembly
o massi e galaxies a ela i ely la e epochs.
E idence seems o indica e an inc ease in he o al s ella mass enclosed in massi e
3.1. EVOLUTION OF THE STELLAR POPULATION WITH REDSHIFT 99
galaxies since z= 1 (e.g. Renzini 2006). This has equen ly been aken as a sign o a
ans o ma ion o blue s a o ming galaxies a z=1 in o he old objec s obse ed a he
p esen epoch (e.g. B inchmann & Ellis 2000;Bundy e al. 2007,2009) al hough Fabe
e al. (2007) ha e a gued ha he numbe densi y o blue galaxies has emained nea ly
cons an since z= 1, whe eas he numbe densi y o massi e ed galaxies has inc eased
indica ing ha some ac ion o ea ly- ype galaxies could ha e been assembled a la e
imes.
While he common unde s anding has o some ime been ha ea ly- ype galaxies
o m h ough massi e me ge s, mo e ecen wo k has pain ed a mo e nuanced pic u e.
Fo ins ance Naab e al. (2006) a gued ha a scena io whe e all ea ly- ype galaxies
o igina e om majo me ge s o gas- ich spi al galaxies is unable o ep oduce he
obse ed p ope ies o e y massi e ellip ical galaxies. Simila ly Bell e al. (2005)
a gued ha he p ope ies o high luminosi y ellip icals a e no compa ible wi h an
assembly ia a p ocess o me ging o spi als wi h signi ican associa ed s a o ma ion,
since he s ella popula ions o hese p ogeni o galaxies would ha e he w ong age
and abundance cha ac e is ics.
A possible way o econcile hese esul s is o pos ula e ha some o hese massi e
galaxies ha e been assembled h ough gas-poo me ge s whe e ew new s a s a e
o med. This p ocess o galaxy me ging wi h no signi ican associa ed s a o ma ion,
dubbed d y me ging (e.g. an Dokkum 2005;Bell e al. 2005), is also consis en wi h
he lack o e y massi e s a - o ming galaxies a low edshi s o build he massi e
non-s a - o ming ea ly- ype galaxies (Bell e al. 2004;Naab & Os ike 2009).
A p e ious s udy by Khoch a & Silk (2009) sugges ed he exis ence o a shu -o
mass scale (Figu e 3.1) whe e d y me ge s would be p edominan o he build up o
massi e galaxies. Using semi-analy ic models, hey ound ha he o e all numbe o
d y me ge s is enhanced in ela ion o he numbe densi y o non-d y me ge s a a mass
scale o M∗,c ∼6.3×1010M⊙, and a gue ha d y me ge s a e a ou ed in building up
massi e galaxies since galaxies a M∗,c a e no longe o ming s a s and will only g ow
h ough me ge s. The exis ence o such a mass cu -o is s ill unde deba e and only
su eys wi h a signi ican numbe o galaxies, such as he one p esen ed in his hesis,
can shed some ligh o his heo y.
100 CHAPTER 3. PROPERTIES OF DRY MERGERS
Figu e 3.1: F ac ion o 1:1 o 1:2 (solid line) and 1:1 o 1:3 (do –dashed line) me ge s
as a unc ion o emnan s ella mass wi hin he edshi in e al 0 < z < 0.06. Inse
g aph: numbe densi y o all me ge s (solid line) and o only we me ge s as a unc ion
o emnan mass wi hin he same edshi in e al. The black e ical dashed line
indica es M∗,c (Khoch a & Silk 2009)
3.2 Me ge s be ween gas-poo galaxies
Dissipa ionless me ge s we e hough o be o li le impo ance, in pa because massi e
galaxies no mally eside in dense egions whe e ela i e eloci ies ypically a e oo high
o make me ge s a likely occu ence. In ac he i s examples o d y me ge s we e
ound in clus e s (e.g. an Dokkum e al. 2001). I was subsequen ly ealised ha
while a e, he dissipa ionless me ge o massi e galaxies migh help explain pa o
he assembly o he mos massi e galaxies.
Despi e being an appealing channel o he o ma ion o e y massi e galaxies, he
obse a ional e idence o d y me ge s is a bi con lic ing wi h some au ho s a guing
o d y me ge s as a signi ican p ocess o he assembly o massi e galaxies (Naab e al.
2006;Bell e al. 2006;Hopkins e al. 2009a;Liu e al. 2009;Eliche-Mo al e al. 2010),
while o he s ind i o be o ma ginal impo ance, such as Masjedi e al. (2006) based
on hei s udy o he co ela ion unc ion o SDSS luminous ed galaxies (LRGs). Some
ecen s udies ha e ied o disen angle he impo ance o d y and we me ge s - o en
implemen ed as ” ed” e sus ”blue” me ge s - as a unc ion o look-back ime. Fo
3.3. MERGER PROPERTIES 101
ins ance Bundy e al. (2009) ind ha he ela i e impo ance o d y majo me ge s
inc eases ela i e o ha o we me ge s a mo e ecen imes. A simila esul is
e ealed by Chou e al. (2011) whe e he numbe ac ion o d y me ge s inc eases
wi h dec easing edshi . This poin s o d y me ge s being an impo an p ocess in
he build-up o massi e galaxies a he p esen epoch.
None heless he e a e signi ican disc epancies be ween he di e en me ge es ima es
in he li e a u e and he echniques used also di e a he signi ican ly be ween di -
e en au ho s. Fu he mo e he e o s men ioned abo e a e mos ly ocused on majo
me ge s, whe e he de ini ion o majo does a y bu is ypically some hing like me ge s
wi h mass- a ios o 1:4 o la ge , and hese migh p esen only pa o he pic u e. This
was ecen ly highligh ed by L´opez-Sanjuan e al. (2010) who emphasised he need o
include mino me ge s o explain he g ow h o ellip icals a low edshi . This is also
consis en wi h he wo k o Bundy e al. (2009) who ound ha he obse ed majo
me ge s do no explain he mass g ow h o sphe oids a z < 1, al hough bo h esul s
do, o some ex en , depend on he unce ain ime-scale o galaxy me ge s. In ac he e
is much wo k o be done in disen angling he ime-scales o he me ge s hemsel es
o de e mine he place o d y me ge s (e.g. Bundy e al. 2009;De P op is e al. 2010;
Chou e al. 2011;Tojei o & Pe ci al 2011)
A undamen al di icul y o s udies o he o ma ion and e olu ion o massi e galaxies
is howe e ha hey a e s ongly clus e ed. This means ha la ge olumes mus be
sampled o educe he impo ance o cosmic a iance. Ano he eason o disc epancies
among di e en s udies is ha he classi ica ion o galaxies as d y me ge s di e s
signi ican ly be ween au ho s wi h limi ed checks on he sensi i i y o he esul s.
3.3 Me ge p ope ies
As men ioned p e iously, ha ing spec oscopic in o ma ion is a majo ad an age o in-
es iga e he ole o me ge s in he build up o galaxies. The spec oscopic in o ma ion
allows no only o mo e accu a ely es ablish a me ging s a us o he galaxies h ough
hei spec oscopic edshi , bu i also allows es ima ion o masses o he s ella and gas
componen s, chemical composi ion, s a o ma ion and nuclea ac i i y, among o he s.

102 CHAPTER 3. PROPERTIES OF DRY MERGERS
The p ima y objec i e o his wo k is o unde s and how galaxies a e assembled ia
me ging o ea ly- ype galaxies and i s impo ance in he build up o he high-mass end
o he galaxy mass unc ion. To ha end, and wi h he a ailabili y o a la ge sample
o me ge s, we can explo e how di e en me ge p ope ies may in luence he mass
unc ion o galaxies.
3.3.1 Majo and mino me ge s
In he p ocess o assembling galaxies ia d y me ging, galaxies o compa able s ella
mass play a bigge ole han me ge s o sa elli e galaxies in o a la ge one. To analyse
he impac o hese di e en ypes o me ge s, we di ide ou sample in o mino and
majo me ge s e e ing o he a io be ween he s ella mass o he galaxies in he
g oup. A g oup is conside ed mino i he s ella mass a io o he galaxies is <1
3
and majo i he s ella mass a io is ≥1
3. In he case o g oups wi h mo e han wo
galaxies, we conside he masses o he mos massi e and second mos massi e galaxy.
We should no e, howe e , ha majo me ge s may no necessa ily be be ween galaxies
wi h a subs an ial s ella mass componen . A g oup wi h low mass galaxies will be a
majo me ge as long as hei s ella masses obey he a io men ioned abo e.
The spec oscopic me ge sample is, as a esul , composed o 1941 majo me ge s
and 2021 mino me ge s. The dis ibu ion o masses plo ed in Figu e 3.2 show he
o al mass o he me ge sample. The s ella masses o each g oup we e summed
o es ima e he inal mass o he assumed end-p oduc o he me ge . E en hough
he majo and mino me ge classi ica ion did no ake in o accoun any s ella mass
cu -o , i is clea om Figu e 3.2 ha majo me ge s a e mo e signi ican o he build-
up o massi e galaxies han mino me ge s, specially o galaxies abo e M∗∼1011 M⊙.
We no e ha he mass unc ions shown h oughou his Sec ion a e al eady co ec ed
o he 0.63 comple eness ac o , as men ioned in Sec ion 2.3.3 o Chap e 2, unless
s a ed o he wise. The e ec o his co ec ion ac o can ac ually be seen in Figu e 3.3,
whe e mass unc ions o di e en se s o galaxies a e plo ed. Using me ge s based
solely on galaxies wi h spec oscopic in o ma ion would in ac be unde es ima ing he
numbe o galaxies in me ge s.
3.3. MERGER PROPERTIES 103
Figu e 3.2: To al s ella mass dis ibu ion o mino and majo me ge s. The
his og ams illus a e he dis ibu ion o he o al mass o he me ge s. The black
his og am is o all he me ge sample, he yellow and ed his og ams a e o mino
and majo me ge s, espec i ely. Indica ed in a dashed line a e also he medians o
mino me ge s in yellow, and o majo me ge s, in ed.
104 CHAPTER 3. PROPERTIES OF DRY MERGERS
Figu e 3.3: Mass dis ibu ion o galaxies in me ge s wi h e o ba s de i ed by
boo s apping me hods. Fo e e ence, he solid black line illus a es he mass unc ion
o all galaxies in he SDSS DR6 spec oscopic sample wi hin he edshi in e al
0.005 < z < 0.2. The g een line co esponds o he mass unc ion o galaxies in he
me ge sample wi hou he co ec ion ac o o 0.63. The yellow line ep esen s he
mass unc ion o galaxies in me ge s co ec ed o 0.63 as men ioned in Sec ion 2.3.3
3.3. MERGER PROPERTIES 105
Figu e 3.4: Mass dis ibu ion o galaxies in me ge s. The solid black line including
all galaxies in me ge s and, o e e ence, he do ed line is he mass unc ion o he
galaxies in he SDSS DR6. The ed line is he mass dis ibu ion o galaxies in majo
me ge s and he blue line is he same o galaxies in mino me ge s. The e o ba s
a e calcula ed using boo s apping me hods.
The mass unc ions o galaxies in majo and mino me ge s is shown in Figu e 3.4.
Galaxies in majo me ge s sample he highe -mass bins o he mass dis ibu ion,
specially a masses abo e log M∗/M⊙= 11.4, in acco dance wi h Malle e al. (2006)
who p edic a highe incidence o majo me ge s o high-mass galaxies.
As p e iously men ioned, one majo ad an age o ha ing spec oscopic in o ma ion
on he galaxies is o s udy in de ail hei s a o ma ion ac i i y usually indica ed by
he p esence o emission lines. Applying s a o ma ion c i e ia could be, a a i s
app oxima ion, a use ul ool o sepa a e be ween galaxies in ol ed in gas-poo and
gas- ich me ge s - galaxies in d y me ge s would show i ually no s a o ma ion and
galaxies in non-d y me ge s would, on he o he hand, e eal s a o ma ion ac i i y
aking place as a esul o he in e ac ion. Figu e 3.5 shows he mass unc ion o he
SDSS DR6 spec oscopic galaxies wi hin he edshi in e al used o build he me ge
sample, i.e. 0.005 < z < 0.2, and he dis ibu ion o he galaxies in he me ge s
sample in h ee di e en bins o s a o ma ion a e pe uni s ella mass, o speci ic
112 CHAPTER 3. PROPERTIES OF DRY MERGERS
ρm=
x
X
i=0
1
Vi
max
ρg=
y
X
j=0
1
Vj
max
(3.2)
wi h ρm ep esen ing he olume co ec ed dis ibu ion o me ge s, wi h xbeing he
o al numbe o me ge s. ρgis he olume co ec ed dis ibu ion o all conside ed
galaxies, i.e., y ep esen s he o al numbe o SDSS DR6 spec oscopic sample o
galaxies inside he edshi in e al.
Ano he quan i y can be de i ed om he p e ious one ha is he olume co ec ed
galaxy me ge ac ion, he a io be ween numbe o galaxies me ging and he o al
numbe o galaxies:
gm =ρgm
ρg
(3.3)
whe e ρgm is gi en by,
ρgm =
z
X
k=0
1
Vk
max
(3.4)
and ρgm is he olume co ec ed dis ibu ion o galaxies in me ge s, wi h z he o al
numbe o galaxies me ging.
We ind a o al me ge ac ion o SDSS DR6, in he edshi in e al 0.005 < z < 0.2
o m= 0.093. Howe e , since we ha e he in o ma ion o he s ella masses o
he galaxies, we in es iga e he me ge ac ion o di e en s ella mass a ios, in
pa icula o majo and mino me ge s. We es ima e a me ge ac ion o majo
me ge s o (m)majo = 0.012 and o mino me ge s o (m)mino = 0.080, in ag eemen
wi h a highe equency o me ge s among galaxies wi h e y di e en masses wi h
ela ion o me ge s o galaxies o app oxima ely he same s ella mass (e.g. Lo z e al.
2010).
The galaxy me ge ac ion, i.e. he o al ac ion o galaxies unde going a me ge ,

3.3. MERGER PROPERTIES 113
gm is 0.031.
We no e ha in ou sample o spec oscopic me ge s, he p esence o sys ems ha ing
N galaxies is comple ely domina ed o g oups wi h N=2, wi h mo e han 97% o he
sys ems ha ing wo elemen s.
3.3.4 Me ge a e
To es ablish a me ge a e is no an easy ask since one can only es ima e a imescale
o he me ge s, assuming (a) he sys ems will me ge and (b) hey will me ge a a
a e age imescale. Assuming a Tmas an a e age imescale o he me ge o ake
place, we can simply de ine a me ge a e as
Rmg = m
Tmg
(3.5)
Se e al s udies (e.g. Lin e al. 2004,Ma eus 2008), use Tmg ∼0.5 Gy ob ained
om N-body simula ions. He e, we adop he me hod o Pa on e al. (2000) and we
es ablish a me ge imescale o ou sample, ollowing (Binney & T emaine 1987). We
es ima e Tmg h ough he dynamical ic ion imescale T ic gi en by
T ic =2.64 ×105 2 c
Mln Λ (3.6)
whe e is he ini ial physical sepa a ion be ween galaxies in kpc, cis he ci cula
eloci y in km/s, Mis he mass in M⊙and T ic is in Gy . We ake he assump ions o
Pa on e al. (2000) and assume ci cula o bi s and a da k ma e densi y p o ile gi en
by ρ∝ −2. The ln Λ is he Coulomb loga i hm and is es ima ed o be ln Λ ∼2. We
ake ad an age o he a ailabili y o s ella masses and we es ima e he me ge a es
o di e en mass bins, de e mining a mean physical sepa a ion ¯ p, a mean eloci y
o se ¯ cand a mean s ella mass ¯
M∗, o each me ge and o each mass bin.
In acco dance o he beha io o he me ge ac ion, he me ge a e shows an inc ease
owa ds mino me ge s, o me ge s be ween galaxies wi h e y di e en s ella masses.
The a e age me ge a e o he me ge sample is 0.025 Gy −1 o an a e age T ic, ha
we ake o be he me ge imescale Tmg, o 0.36 Gy . We no e, howe e , his calcula ion
114 CHAPTER 3. PROPERTIES OF DRY MERGERS
Figu e 3.10: Me ge a es o di e en mass a ios o he me ge sample. The mass
a io conside s he a io be ween he mos massi e and he second mos massi e
galaxies, in he case o me ge s wi h N>2.
3.3. MERGER PROPERTIES 115
is an app oxima e alue aking in o accoun he cha ac e is ics o he sample and i
spans o e a wide ange o alues o T ic, e en when conside ing di e en mass bins.
3.3.5 En i onmen
E en hough he se o condi ions men ioned in Sec ion 2.2 ha e been used o ind close
pai s o g oups, and hus a po en ial me ge , he e ec s o en i onmen in which hese
po en ial me ge s a e imme sed canno be neglec ed. Du ing he isual inspec ion,
some o hese sys ems we e ound in c owded en i onmen s and he e o e may no
me ge.
To add ess his issue, we adop he c i e ia ou lined by Ba on e al. (2007) o es ima e
he numbe o d y me ge s ha a e uly isola ed and, he e o e, will mos likely me ge.
In o de o s udy igge ed s a o ma ion in close pai s, Ba on e al. (2007) ha e
used N-body simula ions and semi-analy ic models o demons a e ha he majo i y
o galaxies in close pai s eside wi hin clus e - o g oup-sized halos. They showed ha
close pai s in low densi y en i onmen s, i.e. wi h only one neighbo wi hin 700 h−1
kpc, ep esen s he ideal scena io o isola e he e ec s o he in e ac ion.
Applying his c i e ion o ou d y me ge sample, combined wi h he eloci y sepa a-
ion c i e ion al eady used o 800 km/s, we ind ha 18% o hese sys ems a e uly
isola ed. The esul s on he numbe o neighbou s o he po en ial d y me ge s wi hin
his limi is illus a ed in Figu e 3.11.
Mos o hese sys ems a e in ac isola ed o ha e only one neighbou galaxy inside
he olume de ined by he abo e men ioned p ojec ed dis ance and eloci y o se .
3.3.6 D y me ge spec al p ope ies
To unde s and he ole o d y me ge s in he galaxy assembly a low edshi , i is
c i ical o s udy how he d y me ge s sys ems compa e o he galaxies ha a e no
in close pai s o g oups. Th ough spec al indica o s we con as he p ope ies o
he s ella popula ions o d y me ging sys ems o he galaxies wi hou he p esence o
116 CHAPTER 3. PROPERTIES OF DRY MERGERS
Figu e 3.11: Dis ibu ion o he numbe o neighbou s o he d y me ge sys ems.
ano he nea by galaxy. We make use o a selec ion o six spec al abso p ion indices
o analyse he wo samples o galaxies: he age sensi i e H-Balme indices HδAand
HγA, and some me allici y sensi i e indices, Mg, Mg2, Fe4531 and Fe5015. Some
s udies ha e shown ha composi e indices, such as [Mg+Fe], a e be e i ed o p obe
me allici y and ha he sum o HδAand HγAis mo e obus o p obe age han he
wo indices sepa a ely (Gallazzi e al. (2006) and e e ences he ein). Ou inal se
o spec al indices hen includes: [HδA+HγA], [Mg2Fe], Mg2, Mg and D4000, whe e
[Mg2Fe] is de ined by equa ion 3.7 om Gallazzi e al. (2006):
[Mg2Fe] = 0.6 Mg2+ 0.4 log (Fe4351 + Fe5015) (3.7)
F om he sample o d y me ge s we es ima ed a sample o d y me ge assumed end
p oduc s, o emnan s, whe e he s ella mass o he end p oduc galaxy is he sum
o he s ella masses o he galaxies in he g oup and he amoun o elemen s in he
emnan is he weighed sum, acco ding o s ella mass, o he elemen s o hei pa en
galaxies. We no e ha ou conclusions a e no al e ed when co-adding he spec a.
Then, om he sample o isola ed galaxies was d awn a se o galaxies wi h simila
s ella mass (whe e ∆ log(M∗/M⊙)≤0.10) o each me ge emnan . Na u ally,
se e al isola ed simila galaxies we e ma ched o e e y single emnan and each se o
3.4. GALAXY ASSEMBLY THROUGH DRY MERGERS 117
simila galaxies (one pe emnan ) con ained a di e en numbe o objec s. To build a
sample o ep esen a i e isola ed and simila objec s he galaxy wi h he median s ella
mass in i s se was de ined o be he ypical isola ed galaxy simila o i s emnan .
Thus, he se o isola ed and simila galaxies con ained as much elemen s as he sample
o me ge emnan s.
The a ia ion o he di e en spec al indices wi h s ella mass o he di e en samples
o galaxies is shown in Figu e 3.12.
The sample o galaxies in me ge s is in ag eemen wi h he pa en SDSS sample o
galaxies, howe e he isola ed sample shows a sus ained di e ence, o all indices,
owa ds highe alues, o lowe alues in he case o HδAand [HδA+HγA], indica ing
a non negligible ole o he en i onmen in he galaxies’ s ella popula ions. The d y
me ge emnan sample ollows his end, showing a e y good ag eemen wi h he
isola ed sample, hence e ealing ha he ea ly- ype galaxies assembled ia d y me ging
sha e he same spec al cha ac e is ics o ea ly- ype galaxies al eady in place.
3.4 Galaxy assembly h ough d y me ge s
Wi h a classi ied sample o galaxies in me ge s and an a e age me ge imescale, we
can now es ima e he con ibu ion o d y me ge s o he assembly o galaxies. To ha
end, we assume ha he emnan esul ing om he me ge o wo, o mo e, galaxies
in each g oup, has a s ella mass equal do he sum o i s p ogeni o ’s s ella masses.
This is a easonable assump ion o d y me ge s because he low gas con en in hese
sys ems means ha li le addi ional s a o ma ion will ake place.
Fo a gi en mass in e al, we can hen (1) es ima e a emnan mass unc ion o he
d y me ge sample, assuming he emnan ’s mass is app oxima ely he same as he
o al mass o he elemen s in me ge , and (2) es ima e how much o he galaxy mass
unc ion is buil ia d y me ging, since z=0.2, using bo h he SDSS DR6 s ella mass
unc ion and he d y me ge sample s ella mass unc ion.
The mass unc ion o he d y me ge emnan s is shown in he uppe panel o Figu e

118 CHAPTER 3. PROPERTIES OF DRY MERGERS
Figu e 3.12: Spec al indices abundance as a unc ion o s ella mass. Solid black line
ep esen s he sample o isola ed galaxies om SDSS DR6, as men ioned in Sec ion
2.4. Dashed line is he sample o end p oduc galaxies om d y me ge s and he
shaded egion shows he 1σcon idence in e al.
3.4. GALAXY ASSEMBLY THROUGH DRY MERGERS 119
Figu e 3.13: Top: The solid line shows he SDSS s ella mass unc ion, wi h he d y
me ge emnan mass unc ion es ima e being he solid line wi h e o ba s. Bo om:
The a io o he d y me ge emnan unc ion o he o al s ella mass unc ion. This
shows he egion o he s ella mass unc ion whe e d y me ge s a e impo an .
120 CHAPTER 3. PROPERTIES OF DRY MERGERS
3.13 wi h he o e all mass unc ion o he galaxies in he SDSS DR6, as e e ed o in
Sec ion 2.1.1 o Chap e 2. The emnan s a e in ac mo e equen owa ds highe mass
bins wi h he median o he s ella mass dis ibu ion o hese sys ems a log M∗/M⊙
= 11.36. Galaxies wi h s ella masses abo e his median alue in he SDSS DR6 galaxy
sample, limi ed o he edshi in e al 0.005 < z < 0.2 (Figu e 3.13), ep esen 3% o
he objec s, e ealing he ele ance o d y me ging in assembling e y massi e galaxies.
The numbe densi y o d y me ge s, i.e. he numbe densi y o emnan s pe Mpc3,
pe mass in e al, is gi en h ough Equa ion 3.8:
Φ(M1≤M≤M2) =
x
X
i=0
1
Vi
max(M1≤M≤M2)(3.8)
whe e Φ ep esen s he numbe densi y o galaxies wi h s ella mass M wi hin he mass
in e al [M1, M2] and x he o al numbe o emnan s, i.e. he o al numbe o d y
me ge s. By inding he a io be ween he s ella mass unc ions shown in he uppe
papel o Figu e 3.13, in each mass bin, i is possible o e ie e he ac ion o galaxies
p oduc o a me ge be ween gas-poo galaxies. The ac ion in each mass bin is lis ed
in able 3.1. Fo compa ison we include in he able an es ima e o he d y me ge
ac ion om he simple model o galaxy e olu ion p oposed by Peng e al. (2010),
on he basis o SDSS and zCOSMOS da a. In hei s udy, Peng e al. de eloped a
pu ely empi ical model o iden i y key ea u es o galaxy e olu ion and ound ha e en
hough he e ec s o d y me ging a e on a e age small o he popula ion o passi e
galaxies, hei impo ance inc eases wi h he obse ed inal mass – as we ha e shown
in Figu e 3.13 as well. To compa e hei numbe s o ou s we ha e scaled hei d y
me ge a es o ou edshi in e al assuming ha he a e o d y me ge s is cons an
wi h ime, which likely leads o an o e es ima e.
While he di e ences in me hodology be ween s udies a e oo la ge o wa an a
de ailed compa ison o he numbe s, we no e ha he end wi h mass is e y simila
and he e is a ai ly good ag eemen o e all. Tha he Peng e al. numbe s a e highe
is easonable as one would expec me ge a es o be somewha highe a ea lie imes
while we ha e assumed cons ancy.
In conclusion, he a e age ac ion o galaxies assembled h ough me ge s o gas-poo
galaxies o 10.1<log(M∗/M⊙)<11.8 is 0.028. We conclude ha o p esen day
3.4. GALAXY ASSEMBLY THROUGH DRY MERGERS 121
log M(M⊙) F ac ion Peng e al.
[10.0, 10.2] 0.000258 0.0115
[10.4, 10.6] 0.00158 0.0159
[10.9, 11.1] 0.0164 0.0332
[11.3, 11.5] 0.0615 0.0762
[11.5, 11.7] 0.0686 0.114
[11.7, 11.8] 0.0533 0.161
Table 3.1: F ac ion o d y me ge emnan s in di e en mass bins
galaxies wi h log(M∗/M⊙)>10 up o ∼7% o hei s ella mass is in ol ed in d y
me ge s, since z= 0.2, wi h galaxies o s ella mass 11.5<log(M∗/M⊙)<11.7 being
mo e p edominan ly assembled h ough his ype o me ge s.
Gi en ha he median edshi o he galaxies in d y me ge s is 0.1, which ansla es
in o 1.3Gy , hese ac ions a e hen an es ima ion o he uppe limi o he con ibu-
ion o d y me ge s o he local mass unc ion. Assuming ha , ypically, he me ge
ime-scale is si ua ed a ∼0.4Gy , hese galaxies would ha e me ged by edshi ze o.
128 CHAPTER 4. METALS AND GAS IN MERGERS
a e (SFR) is he a e a which he gas is being ans o med in o new s a s,
whe eas he speci ic s a o ma ion a e (SSFR) is he a e a which he galaxy
is o ming new s a s pe uni s ella mass.
•Robe s Time: This ime-scale(Robe s 1963;Kennicu e al. 1994) measu es
he amoun o ime ha will ake o he gas o a gi en galaxy o o m new s a s,
assuming a gi en s a o ma ion a e. The e o e, TRcan be w i en like:
TR=Mgas
SFR
This quan i y is pa icula ly use ul, since i makes use o he in o ma ion om bo h
he s a o ma ion a e and he amoun o gas s ill a ailable o o m new s a s. This
ime-scale can also be hough as a deple ion ime-scale: how long does he a ailable
gas ake o be ans o med in o s a s, howe e , as s a ed by Kennicu e al. (1994)
i cons i u es a lowe limi since i does no accoun o gas ecycling. This quan i y
is also use ul because i is likely o ha e a di e en dependence on ape u e size han
ei he gas o SFR on hei own.
4.1.3 Using a con ol sample
The mass-me allici y ela ion shown in Figu e 4.1, e eals a clea end o highe s ella
mass galaxies ha ing a highe me al con en . Clea ly a blind compa ison o galaxies
would no be a good idea and one needs o ake his a ia ion wi h s ella mass in o
accoun when doing he compa isons.
This beha iou is no iced when compa ing gas wi h he s ella mass o galaxies, as
illus a ed in Figu e 4.3. On he x-axis we plo he s ella mass and on he y-axis gas
— and each posi ion in his diag am is colou ed acco ding o he median edshi o
he galaxies ending up he e. This igu e shows he complex in e play o wo e ec s:
he immedia e e ec is ha mo e massi e galaxies ha e ela i ely less gas — a well-
known end seen in all HI su eys (e.g. Ca inella e al. 2010). Bu coupled o his
gene al decline in gas wi h s ella mass, he e is also a dependence on he amoun o
gas o he amoun o s a s wi h edshi , which is a sign o an ape u e e ec - galaxies
a lowe edshi s ha e only hei cen al egions sampled by he ibe , whe eas a
highe edshi s he ibe co e s a la ge adii, i no all galaxy. Mo eo e , gas discs
a e mo e ex ended han s ella discs. This na u ally leads o a unde es ima ion o he

4.1. METALLICITY IN MERGERS 129
Figu e 4.3: Dis ibu ion o gas as a unc ion o s ella mass o he edshi in e al
0.005 < z < 0.2. The dependence o gas in edshi in each mass bin is clea ly seen,
in pa icula o galaxies wi h s ella mass &10 M⊙whe e gas co e s a wide ange in
edshi , wi h lowe edshi galaxies p esen ing lowe alues o gas.
gas con en o galaxies a lowe edshi s and an ape u e co ec ion would be needed.
Howe e , by compa ing he me ge sample wi h he con ol sample, and since hey a e
bo h ob ained a simila edshi s, we expec ha we will mos ly be able o emo e
his e ec .
As men ioned be o e, he e a e majo ad an ages in using a sample o isola ed galaxies
d awn om he same su ey as he me ge sample. In his case, he isola ed sample
was buil based on he pa ame e s o he sample o galaxies making up he me ge
sample and, as desc ibed in sec ion 2.2.3, only galaxies simila in s ella mass and
edshi we e selec ed, o each me ge galaxy. The mos impo an bene i o his
me hod is ha he ela i e measu emen s, be ween a gi en galaxy in a me ge and
he se o i s simila galaxies, allows us o di ec ly compa e iden ical galaxies and
ob aining obus esul s when analysing he e ec s o a neighbou galaxy.
130 CHAPTER 4. METALS AND GAS IN MERGERS
4.2 In es iga ing he me als and gas in me ge s
4.2.1 The low-me allici y gas con en o me ge s
In o de o in es iga e whe he he in e ac ion be ween galaxies in luences hei gas
dynamics, he e we explo e whe he he me al abundance in he inne egions o
galaxies in he me ge sample changes wi h he sepa a ion o he galaxies. Ea lie
s udies (e.g. Za i sky 1993) gene ally seemed o ind ha spi al galaxies ha e ela i ely
s eep me allici y g adien s. Howe e mo e ecen ly Mo an e al. (2010) ha e ound ha
gas-phase me allici ies show a ela i ely la p o ile wi h inc easing adius, wi h a d op
in he me allici y o he ou e egions. While some galaxies ha e s ong me allici y
g adien s, i appea s ha as a popula ion hey end o ha e ela i ely la me allici y
p o iles. Obse a ions indica e ha , ega dless o hei inne p o iles, galaxies ha e
low-me allici y gas in he ou e egions, hus ha ese oi o me al poo gas a ou
in he galaxies can be mixed up du ing me ge s.
By compa ing ou me ge sample o isola ed galaxies, we can in es iga e in pa icula
he gas in low owa ds cen al egions — is he low-me allici y gas om he ou ski s o
he galaxy being channelled o he cen e du ing a me ge ? To answe his ques ion we
limi he edshi in e al o galaxies in he ange 0.02 < z < 0.08 in o de o sample
he cen al egion o he galaxies — highe edshi galaxies, as p e iously men ioned,
would ha e la ge egions sampled while he low edshi cu -o is o educe he se e i y
o ape u e e ec s. Fo his analysis he e we e a o al o 1687 galaxies om he me ge
sample. Ac oss his edshi ange he ape u e o he SDSS ib e sub ends om ∼1.2
kpc o ∼4.5 kpc
The esul ob ained by compa ing he me allici y abundance o he galaxies in he
me ge sample wi h he sample o isola ed galaxies, i.e. ∆ log(O/H), is shown in
Figu e 4.4. In he Figu e we plo he di e ence be ween he me allici y o galaxies in
me ge s and he me allici y o hei espec i e isola ed coun e pa s.
Fo me ging galaxies, an inc ease in he me allici y is clea ly seen o p ojec ed sepa-
a ions below 10 kpc - as he galaxies ge close du ing he me ge e en , he cen al
egions o hese galaxies a e mo e me al en iched when compa ed wi h hei isola ed
coun e pa s. This esul poin s in he di ec ion o no signi ican in low o low-
4.2. INVESTIGATING THE METALS AND GAS IN MERGERS 131
Figu e 4.4: The median dis ibu ion o he di e ence in me allici y as a unc ion o
p ojec ed sepa a ion. The lowe and uppe shaded a eas ep esen he 68% and 84%
pe cen iles, espec i ely, as ob ained by boo s ap e-sampling, wi h n=999.
132 CHAPTER 4. METALS AND GAS IN MERGERS
me allici y gas o he cen e . I ha was he case, he e would be a dilu ion o he me al
con en and he me allici y would show a dec ease. Ins ead, he obse ed inc ease in
me allici y could be a ibu ed o he inc eased s a o ma ion a e igge ed by he
me ge e en , ha would lead o me al injec ion and, he e o e en iching he ISM in
he cen al egions o he galaxies. This pic u e would be consis en wi h he enhanced
s a o ma ion ac i i y shown in Sec ion 2.5.1 o Chap e 2, bu o ully unde s and
his inc ease in me als in he cen al egions, one should also analyse how hei gas
con en changes wi h sepa a ion.
4.2.2 The gas con en o me ge s
As men ioned, he analysis o he gas con en o he cen al egions o galaxies in
me ge s can shed in aluable ligh on he gas dynamics du ing he encoun e . By
ocusing on s a - o ming gas- ich galaxies and inspec ing how he gas con en beha es
du ing a me ge e en , we can s a o cons ain he ole o in low o gas o inne
egions and con as his wi h he igge ed s a o ma ion o pain a pic u e on he
me allici y con en o cen al pa s o me ging galaxies.
Using he same app oach, we compa e he galaxies in he me ge sample wi h he
se o isola ed galaxies o in es iga e whe he me ging galaxies a e mo e gas ich
in hei cen al egions han he galaxies wi hou any neighbou nea by. Mo eo e ,
o obus ness, only galaxies classi ied as s a o ming (acco ding o he p esc ip ion
o B inchmann e al. (2004)) a e used, in a o al o 1658 galaxies. This ela i e
measu emen , o he di e ence be ween gas o galaxies in me ge s and hei ela i e
isola ed coun e pa s, is shown in Figu e 4.5.
Fo sepa a ions abo e 10 kpc, he e is i ually no di e ence in he median gas –
galaxies in me ge s p esen he same a io as isola ed galaxies. On he o he hand,
a end is seen o p ojec ed sepa a ions below 10 kpc owa ds nega i e alues o
∆ log gas meaning ha , in he cen al egion, me ging galaxies a e gas deple ed ela i e
o isola ed galaxies. This dec ease appea s consis en wi h he inc ease in me als ound
o galaxies in me ge s (see Sec ion 4.2.1) and also wi h he inc ease in s a o ma ion
obse ed o me ging galaxies when compa ed wi h isola ed galaxies (see Sec ion 2.5.1),
conside ing a simple scena io whe e he me al en ichmen caused by he inc ease in s a
o ma ion is no signi ican ly a ec ed by he in low o low-me allici y gas. None heless,
4.2. INVESTIGATING THE METALS AND GAS IN MERGERS 133
Figu e 4.5: The median o he a ia ion o ∆ log gas as a unc ion o p ojec ed
dis ance. The lowe and uppe shaded a eas ep esen he 68% and 84% pe cen iles,
espec i ely, as ob ained by boo s ap e-sampling, wi h n=999.

134 CHAPTER 4. METALS AND GAS IN MERGERS
an inc eased s a o ma ion capable o using up he gas and ejec ing me als back in o
he galaxy’s ISM would ha e o be consis en wi h lowe gas deple ion ime, i.e., he
a ailable gas o o m new s a s would ha e o be consumed mo e apidly, inc easing
bo h he a e o s a o ma ion and he amoun o me als.
4.2.3 The gas consump ion ime-scale in me ge s
As men ioned ea lie , he Robe s ime can be in e p e ed as a gas consump ion ime-
scale, which gi es in o ma ion on he ime i would ake o con e all he a ailable
gas in o s a s, wi h cons an s a o ma ion a e a he p esen alue, assuming he e
is no ecycling o he gas and ha no gas is being added o he egion s udied. Gi en
he p e ious esul s o a dec ease in gas coupled wi h an inc ease in s a o ma ion
o galaxies wi h p ojec ed dis ances below 10 kpc, analysing he ime-scale o gas
o con e in o s a s should gi e a clea pic u e o he p ocess behind gas deple ion
in me ge s. Once again, we use a ela i e measu e by compa ing TRo galaxies in
me ge s wi h he TRo isola ed galaxies - do me ge s and isola ed galaxies show simila
deple ion imes o hei gas?
The esul s in Figu e 4.6 e eal a p onounced dec ease in TR, which is he esul o
a sho e gas consump ion ime-scale o galaxies in me ge s when compa ed wi h
isola ed galaxies: galaxies in he me ge sample a e unning ou o gas as e han
hei isola ed simila galaxies. Because TRdepends on he mass in gas Mgas and on
he s a o ma ion a e (SFR), we could p edic a dec ease in TRgi en he p e ious
esul s. The ac ha such beha iou is obse ed in TRis hen consis en wi h a
scena io whe e he cen al egion o he galaxy du ing a me ge p ocess is somehow
blocked o exchanges wi h he ou e egions o he galaxy.
One no iceable di e ence be ween he di e en ends seen be o e, is ha his does
no each a alue o ze o a la ge sepa a ions. Al hough i is no clea why his is he
case, i migh indica e ha he e is s ill some enhancemen o s a o ma ion ac i i y
o sepa a ions abo e 30 kpc. E en a small e ec o his enhancemen would become
impo an when measu ing TRsince i depends on bo h he SFR and he gas mass.
A special sample o isola ed galaxies ha would make a good compa ison sample in
his case would be cons i u ed o galaxies wi h no neighbou s wi hin a adii o >100
kpc. As he end in Figu e 4.6 seems o indica e, o sepa a ions la ge han 30 kpc
4.2. INVESTIGATING THE METALS AND GAS IN MERGERS 135
Figu e 4.6: The a ia ion in Robe s ime, o gas consump ion ime-scale, as a unc ion
o p ojec ed sepa a ion. The lowe and uppe shaded a eas ep esen he 68% and 84%
pe cen iles, espec i ely, as ob ained by boo s ap e-sampling, wi h n=999.
136 CHAPTER 4. METALS AND GAS IN MERGERS
i would e en ually each ze o.
4.3 Discussion
Conce ning he me allici y and gas con en o he cen al egions o galaxies in me ge s,
he esul s can be summa ized as ollows:
•An inc ease in me allici y o ∼a ac o o 1.5 o p ojec ed dis ances below 10
kpc.
•A dec ease in gas o ∼a ac o o 1.5 o p ojec ed dis ances below 10 kpc.
•A dec ease in TR, mo e p onounced o p ojec ed sepa a ions below 10 kpc, o
a ac o o 4.
•Addi ionally, om Chap e 2, an enhancemen o s a o ma ion ac i i y o
p ojec ed dis ances below 10 kpc up o a ac o o 3.
As men ioned in Chap e 1, se e al mechanisms can be p omp ed by he me ge
e en which con ibu e signi ican ly o he e olu ion o a galaxy - e.g. eedback om
igge ed AGN ac i i y - howe e , hese esul s o he me al and gas con en o cen al
egions o galaxies a e no expec ed o be a ec ed by such p ocesses, since hey a e
p obing he pa icula ime on he me ge e en be o e hose p ocesses ake place.
We ha e a measu emen o he inc ease in s a o ma ion ac i i y as a unc ion o
sepa a ion o he galaxies and coupled o ha we ind a dec ease in he gas con en ,
an inc ease in he me allici y and a dec ease in TR. Conside ed oge he hese esul s
seem o sugges a scena io whe e no in- alling, o negligible in- alling, o low-me allici y
gas akes place. In o he wo ds, hese esul s seem o be poin ing owa ds a simple
scena io o gas deple ion du ing he la e s ages o a me ge e en - a closed box model
o he cen al egions o galaxies. The closed box model, a o malism i s in oduced
by Tinsley (1980) o he chemical e olu ion o galaxies, can be in e p e ed as a) a
closed egion whe e b) he e is some ”ini ial” gas o o m new s a s and c) no in low
o ou lows o gas mass, leading o a me al en ichmen o he egion as s a s keep being
o med om he a ailable gas assuming ha d) he ini ial mass unc ion (IMF) does
4.3. DISCUSSION 137
no change.
A quali a i e assessmen o he possibili y o using a closed box model, whe e he
enhanced s a o ma ion ac i i y ha esul om he idal o ces pe u bing he gas
in o s a s is he solely esponsible o he gas dynamics, explains simply all he obse ed
ea u es ega ding me allici y and gas con en o he cen al egions o galaxies. In a
closed sys em, whe e no in lows o ou lows o ma e ial ake place, he inc eased s a
o ma ion a e gi es ise o a dec ease in gas since he gas is being consumed in o
s a s and he a io o gas mass o s ella mass dec eases. The dec ease in gas ac ion
combined wi h a inc ease in SFR, is ansla ed in o lowe alues o TR. Because he e
is no eplacemen o he gas, he e is no dilu ion o he me al con en and he inal
me allici y is highe han he ini ial me allici y due o he me al ejec ion ha in he
mean ime is en iching he medium.
Bu e en hough his esul s a e consis en be ween hem and wi h a closed box model
o he cen al egions o me ging galaxies, ano he app oach is needed: o e i y i
his simple model could p edic hese beha iou s o he a ia ions o gas, Z and TR.
To no e ha changes in gas ac ion and me al con en will likely be mass dependen
and likewise he inc ease in s a o ma ion a e. Howe e e en wi h hese la ge samples
we a e limi ed in ou abili y o s udy he e ec o mass on hese ends. Thus a e age
ends will be used h oughou his analysis o explo ing whe he he obse ed ela ions
a e consis en wi h a closed box solu ion o no . Fo his we will he e go h ough his
calcula ion using an a e age mass galaxy, o 1010 sola masses as an illus a i e case.
We ha e also calcula ed all hese ends as a unc ion o mass and ha e e i ied ha
his illus a i e case is a good ep esen a ion o he a e age. The e is a weal h o
in o ma ion in hese da a, bu o ex ac his ou u he we need o do a conside ably
mo e sophis ica ed analysis, including a p ope chemical e olu ion model and a mo e
ca e ul ea men o he ime-e olu ion o hese quan i ies, which is ou side he scope
o his i s a emp he e.
144 CHAPTER 4. METALS AND GAS IN MERGERS
Unde he assump ion o a simple closed box model o he cen al egions o galaxies,
he p edic ions om he SFR ega ding wha he gas con en will be, wha end
would he me allici y and gas consump ion ime scales e eal, a e consis en wi h
ou esul s. The blind p ojec ions conce ning he change in gas, Z and TRusing
analy ical es ima es ha e been shown o be consis en wi h each o he . This lends
some c edibili y o he no ion ha he cen al egions can be modeled as closed boxes.
Rega ding p e ious s udies, he main objec o s udy o in e ac ing and/o me ging
galaxies has been he en i e galaxy whe e he p o iles o me allici y g adien s a e aken
in o accoun as well as he in- alls o me al-en iched gas ha p oduce di e se s ella
popula ions (e.g. Kewley e al. 2006a;Michel-Dansac e al. 2008;Rupke e al. 2010;
Kewley e al. 2010). Ou s udy, being ocused solely in he cen al egions o he
galaxy, ep esen s, o he bes o ou knowledge, a i s app oach o his subjec and
he e o e is no possible di ec a compa ison be ween he wo me hods - cen al egion
s. en i e galaxy.
Fu u e Wo k
Gi en he expe imen al na u e o his s udy on he gas and me al con en o me ging
galaxies, he e is ce ainly mo e wo k o be done in o de o es ablish his closed box
beha iou o he cen al egions o me ging galaxies. The i s app oach would be o
apply his analysis o o he samples o me ging galaxies and in es iga e whe he his
beha iou holds. Howe e , he lack o gas es ima es, and he di icul y o ge eliable
measu emen s specially o la ge su eys, is a majo obs acle.
A possible s a ing poin would be o selec a small sample o some nea by sys ems o
galaxies used in his p ojec and ob ain eliable measu emen s - h ough bo h adio
obse a ions o map he HIcon en and in eg a ed ield uni spec oscopy (IFU) o
assess hei SFR igo ously. This p ocedu e could hen obus ly con i m he pa icula
dynamics o he cen al egions o me ging galaxies ha , gi en he esul s p esen ed
he e, seem o be qui e obli ious o i s su oundings.

Chap e 5
P obing single me ge s
The ad an ages o in es iga ing a la ge sample o me ge s a e mani old bu , mos
impo an ly, such samples a e c ucial o build a heo y o galaxy o ma ion and
e olu ion in he now p e ailing ΛCDM cosmology, whe e s uc u e g ows hie a chically
h ough me ge s o e e g owing clumps o ma e .
None heless, he exac p ocesses behind a me ge e en be ween any wo, o mo e,
galaxies can only be unde s ood in de ail h ough he inspec ion o indi idual exam-
ples o me ging galaxies. Fo ins ance, as e e ed in Chap e 1, he i s clue o
he impo ance o d y me ging in building up massi e galaxies was e ealed by he
disco e y o a iple ed me ge a z=1.27 in he clus e RX J0848+4453 in a wo k by
( an Dokkum e al. 2001). The e is a lack o well known me ge s be ween gas-poo
galaxies, due o hei almos unpe u bed mo phologies du ing he me ge p ocess and
o he ecen in e es in his ype o me ge s, bu o he indi idual me ge sys ems ha e
been exhaus i ely s udied specially gas- ich in e ac ing galaxies – e.g. he An ennae,
S ephan’s Quin e , Whi lpool galaxy (e.g. Nikola e al. 2001;Read 2003;O’Sulli an
e al. 2009). Such ocused in es iga ion allowed a be e unde s anding o phenomena
like he in e nal p ocesses leading o igge ed s a o ma ion o he me ge dynamics
causing ea u es like idal ails and b idges. The case o d y me ge s is no di e en .
145
146 CHAPTER 5. PROBING SINGLE MERGERS
5.1 A subse o d y me ge s
Du ing he cou se o his wo k, in pa icula du ing he isual inspec ion, se e al
in e es ing sys ems we e ound. A a a he low edshi ( o no e ha he high edshi
limi o he me ge sample is z= 0.2), some o he sys ems in his subse o me ge s
consis o mo e han wo, and possibly as many as i e, ellip ical galaxies wi h small
sepa a ions. These sys ems we e de ec ed because a leas wo o he objec s had
spec oscopic in o ma ion om he SDSS and we e wi hin he condi ions o a me ge
men ioned in Chap e 2, wi h a p ojec ed dis ance p≤30 h−1kpc and eloci y o se
sep ≤800 km/s. Howe e , p obably due in pa o he limi a ion imposed by he
placemen o he ibe s and o he selec ion o pho ome ic a ge s o spec oscopic
obse a ions, mos o hese sys ems lack he spec oscopic edshi needed o assess
hei me ging s a us.
Some o hese sys ems a e illus a ed in Figu e 5.1. The a ailable spec a o hese
galaxies show no emission lines, indica i e o an absence o ac i e s a o ma ion, and
he emaining galaxies ha e colou s consis en wi h ellip icals a he same edshi .
Ou es ima ions poin owa ds each o hese me ge candida es enclosing a o al s ella
mass in excess o 2.5×1011M⊙. The edshi s o hese sys ems ange om z= 0.10
o z= 0.15 and hei p ojec ed sepa a ions a e .20 kpc and hei eloci y o se s ≤
400 km/s. They may cons i u e key examples o he build-up o he massi e end o
he mass unc ion, in pa icula a low edshi s.
5.2 The quin uple d y me ge
One pa icula ly in e es ing d y me ge example is he case o a possible me ge o
up o i e galaxies, shown in Figu e 5.2 ( op). The objec s in his sys em ha e been
classi ied by he SDSS as ou galaxies and one s a . Howe e , ha is mos p obably an
inco ec classi ica ion due o he compac ness o he objec . This g oup has spec a
o wo o he galaxies, as indica ed in Figu e 5.2 bo om, ha show no signs o s a
o ma ion and place his sy em a z∼0.18. All galaxies in he image ha e simila
-band luxes and colou s consis en wi h being ellip icals a he same edshi . The
galaxies wi h spec oscopic da a a e a a p ojec ed sepa a ion o <12 kpc and ha e
a eloci y o se <410 km/s. Radio da a om he Fain Images o he Radio Sky a
Twen y-cm (FIRST) su ey, has shown a clea de ec ion o one o he sou ces ( he
5.2. THE QUINTUPLE DRY MERGER 147
Figu e 5.1: Th ee po en ial me ging sys ems aken om he me ge sample. The
galaxies wi h exis ing SDSS spec a a e iden i ied by he ed a ows. The images a e
1’ x 1’ in size and a e o ien ed wi h no h up and wes o he igh .
148 CHAPTER 5. PROBING SINGLE MERGERS
highligh ed galaxy in he le in Figu e 5.2) and a possible ex ended egion owa ds
he di ec ion o he o he ou galaxies.
Assuming a me ge o ou galaxies and ha all sou ces a e in ac a he same edshi ,
he o al s ella mass o each o he ellip ical galaxies in sligh ly below 1011M⊙and
he o al mass o he me ge is ∼3×1011M⊙. The e is no known o e densi y a ound
his me ge , bu Bes e al. (2006) ind ha adio-loud galaxies wi h s ella masses
<1011M⊙a e almos exclusi ely b igh es cen al galaxies, hus he e may be an
unde ec ed g oup a ound his me ge , in which case we migh in ac be looking a a
o ming b igh es cen al galaxy. The e o e, his sys em may be an excellen example
o he build-up o he massi e end o he ellip ical mass unc ion and i should o e a
numbe o insigh s in o he mechanics o gas-poo me ge s and how hey p oceed.
5.2. THE QUINTUPLE DRY MERGER 149
Figu e 5.2: The me ging sys em image om he SDSS DR6. The i e objec s a e
clea ly seen, wi h he one in he igh being he possibly misclassi ied galaxy. The
g een objec s in op image a e classi ied as s a s by he SDSS. The ed squa es in he
bo om image indica e he galaxies wi h spec oscopic in o ma ion.

Chap e 6
Conclusions and Fu u e Wo k
The s udy o me ging galaxies s a ed mo e han a hal a cen u y ago and e en oday,
while some ques ions ha e been answe ed, ou knowledge o his phenomenon is s ill
a om comple e. This holds pa icula ly ue o d y me ge s, since hei ole in
galaxy o ma ion and e olu ion was ecognized only a decade ago. Building he la ges
sample o d y me ge s o da e was he co e o his wo k and analysing in de ail all
he in o ma ion enclosed in his sample wi h close o 4000 me ging sys ems, wi h
app oxima ely 20% being d y me ge s, ep esen s an endea o ha ex ends a beyond
his PhD hesis.
6.1 Assembling galaxies ia me ge s
As galaxies come close oge he du ing a me ge e en , he changes in hei p ope ies
a e a ec ed, mo e o less se e ely, h oughou he ime-scale ha he me ge encom-
passes. Du ing he encoun e , as hey app oach and e en collide, a ious p ocesses
ake place om mild mo phological changes, usually s a ing in he i s s ages o he
me ge , o iolen igge ing o nuclea ac i i y in wha is possibly he las s age be o e
he galaxies ha e combined and no longe can be iden i ied as wo, o mo e, sepa a e
sys ems. Meanwhile, du ing his ime, no only hei shapes a e changed d ama ically
bu bo h he s ella and gas con en o he indi idual galaxies can be a ec ed, mo e
o less s ongly, a he di e en s ages o he in e ac ion. Ha ing a sample o galaxies
150
6.1. ASSEMBLING GALAXIES VIA MERGERS 151
a se e al s ages o he me ge e en is hen key o ha e an insigh in o how di e en
p ocesses change o e ime.
6.1.1 S a o ma ion and nuclea ac i i y
The enhanced s a o ma ion ac i i y is p obably he mos immedia e consequence o
galaxy in e ac ion. Due o idal o ces playing du ing he in e ac ion, he pe u bed
gas in he galaxies is quickly ans o med in o s a s, o en in iolen me ge induced
s a -bu s s. Bene i ing om he spec oscopic in o ma ion, he EW(Hα) was used as
a p oxy o s a o ma ion ac i i y and ha e y phenomenon was obse ed in he
galaxies o he me ge sample ga he ed o his wo k. A s ong ela ion be ween galaxy
sepa a ion wi hin a g oup and s a o ma ion ac i i y was ound and, o e y close
galaxies (in hei la e me ge s ages), i.e. o p ojec ed sepa a ions below 10 kpc, an
enhancemen up o a ac o o h ee was ound.
The nuclea ac i i y, on he o he hand, did no show he same beha iou as he s a
o ma ion ac i i y. Theo e ical wo k has equen ly p oposed ha a galaxy-galaxy
collision may igge he AGN ac i i y in he galaxies in a e y la e s age o he
me ge , which could be ansla ed in o an inc ease o he AGN ac ion owa ds small
p ojec ed dis ances. E en hough we ha e no such galaxies in ou sample, we could
explo e he p og ess owa ds his s age, usually whe e he nuclei o he galaxies ha e
coalesced in o an appa en ly single objec , we e no p esen in ou me ge sample,
a dec ease o he ac ion o AGN in he me ge sample owa ds small sepa a ions
was ound. As men ioned, we make use o he classi ica ion as AGN-hos galaxy,
based on spec oscopic da a, as desc ibed in Kau mann e al. (2003), whe e a se
o spec oscopic indexes classi y a galaxy in s a - o ming, AGN o composi e. The
dec ease o he AGN ac ion was conside ed o be he esul o he lack o de ec ion
o his AGN classi ying emission lines, since he igge ed s a o ma ion ac i i y seems
o eil he p esence o he AGN.
6.1.2 Majo and mino me ge s
The ole o majo me ge s in building up massi e galaxies is expec ed o be mo e
signi ican when compa ed wi h mino me ge s, since hey would be mo e e icien
o assemble galaxies wi h highe masses. Ou me ge sample consis ed o oughly he
152 CHAPTER 6. CONCLUSIONS AND FUTURE WORK
same numbe o majo and mino me ge s, whe e we de ined mino me ge s o be hose
wi h me ge s wi h s ella mass a ios ≥1
3o <1
3, espec i ely, and majo me ge s ha e
been ound o play a bigge ole especially abo e M∗∼1011M⊙.
6.1.3 D y and non-d y me ge s
App oxima ely 80% o he me ge s in he me ge sample used in his wo k a e indeed
me ge s whe e a leas one o he galaxies is gas- ich. Es ablishing a limi on he
amoun o gas ha would di ide he me ge s in o gas-poo o gas- ich is a he a bi-
a y since he o e all p ope ies o hese galaxies has been shown o be a con inuum –
consequen ly, he e is a g adual ansi ion om d y o non-d y me ge s. Acco ding o
ou c i e ia o de ine d y and non-d y me ge s, he la e we e ound o be impo an in
he in e media e and low mass pa s o he galaxy mass unc ion. The main ques ion
mo i a ed by his wo k was hen:
Wha is he ole o d y-me ge s in building up he mos massi e end o he mass
unc ion?
The ac ion o galaxies in po en ial d y me ge s in he spec oscopic SDSS DR6 sample
cons i u es mo e han 2% o he galaxies. In ac , 20% o he me ge sample is made
up by a la ge sample o me ge s be ween gas-poo galaxies – which ansla es o ∼
700 sys ems. D y me ge s ha e been shown o be pa icula ly impo an o masses
g ea e han 5 ×1011M⊙, e ealing hei s a us as building blocks o e y massi e
galaxies - he ole o d y me ge s in building up he massi e end o he galaxy mass
unc ion was ound o be no negligible: up o 7% o he s ella mass in galaxies wi h
masses 11.0<log(M∗/M⊙)<11.6 has been in ol ed in his ype o me ge s.
This esul seems o econcile he di e en iews on he impo ance o d y me ge s in
building up he high mass end o he galaxy mass unc ion. Conside ing bo h majo
and mino d y me ge s (whose impo ance was highligh ed by L´opez-Sanjuan e al.
(2010)) we ha e shown ha some ea ly- ype galaxies ha e indeed been assembled a
la e imes, a leas since z= 0.2, in acco dance wi h nume ous wo ks poin ing owa ds
an inc ease o s ella mass enclosed in ea ly- ypes o z < 1. Wi hou ques ioning ha
he massi e end o he mass unc ion was mos ly in place a z∼1, some e olu ion
6.1. ASSEMBLING GALAXIES VIA MERGERS 153
appea s necessa y.
E en hough he e a e a numbe o poin s o conside when compa ing in de ail he
esul s o se e al s udies – e.g. di e en su eys, edshi in e als, me ge imescales
– which can hinde a di ec compa ison and, he e o e, he possibili y o ge ing a
clea pic u e o he ole o d y me ge s, he gene al di ec ions can be pa ched oge he
shaping up a new scena io o galaxy assembly h ough his ype o me ge s. To no e:
•The bulk o he massi e end o he mass unc ion being al eady in place by z∼1
(e.g Ab aham e al. 2007;Le Fe e e al. 2003;Sca la a e al. 2007)
•The inc ease o s ella mass enclosed in ea ly- ypes since z= 1 (e.g. B inchmann
& Ellis 2000;Hogg e al. 2002;Renzini 2006;Sca la a e al. 2007;Bundy e al.
2009)
•The me ge s be ween gas- ich spi als no able o ep oduce he obse ed p op-
e ies o e y massi e ellip icals (e.g. Bell e al. 2005;Naab e al. 2006)
•D y me ge s a la e epochs a e obse ed (e.g. an Dokkum e al. 2001;Bell e al.
2006;Masjedi e al. 2006)
These indings can be b ough oge he by he small, ye signi ican , ac ion o galaxies
assembled ia d y me ge s ound in his wo k, in pa icula o galaxies wi h s ella
masses abo e log(M∗/M⊙)>11.0. The inc ease in s ella mass in ea ly- ype galaxies
is likely o come om wo sou ces, d y me ge s a e y high masses and a mix u e
o d y and non-d y me ge s a lowe masses. E en hough he e is much wo k o be
done in disen angling he ime-scales o he me ge s hemsel es – i seems o be he
key, and p obably he las majo , obs acle o de e mine he place o d y me ge s (e.g.
Bundy e al. 2009;De P op is e al. 2010;Chou e al. 2011;Tojei o & Pe ci al 2011)
– he impo ance o his ype o me ge s in building e y massi e galaxies is de ini ely
signi ican wi hin he sampled low edshi in e al o 0.005 < z < 0.2.
6.1.4 En i onmen
I could be nai ely expec ed ha me ge s in high densi y en i onmen s would be mo e
equen when compa ed wi h low densi y egions. The common unde s anding ha