Spatially-resolved galaxy angular momentum Sarah Sweet Swinburne - - PowerPoint PPT Presentation

Spatially-resolved galaxy angular momentum

Sarah Sweet

Swinburne With Deanne Fisher, Karl Glazebrook (Swin), Danail Obreschkow, Claudia Lagos, Liang Wang (UWA)

Outline

• Background:

– Angular momentum is fundamental – Measuring spatially-resolved angular momentum

• Results:

– The pseudobulge evolutionary track – Angular momentum at cosmic noon – The probability density function of angular momentum

Galaxy evolution

Galaxy evolution in context

NAOJ cosmic noon cosmic evening cosmic dawn

Galaxy evolution in context

GiggleZ

Galaxy evolution in context

medium.com

Angular momentum is fundamental

“Most astronomers would agree that the list of important parameters should be headed by the total mass M, energy E and angular momentum J. Next on the list should probably be the relative contributions to these quantities from the disk and bulge components…” (Fall 1983)

Angular momentum is fundamental

• size
• density evolution
• disk thickness
• colour
• morphology

Mo+1998, Hernandez+Cervantes-Sodi 2006, Fall 1983, Romanowsky+Fall 2012, Obreschkow+Glazebrook 2014, Cortese+2016, Posti+2018, Sweet+2018, Fall+Romanowsky 2018

Stellar mass is also fundamental

• Remove stellar mass scaling

j baryons ~ j halo

• same tidal forces during spinup
• "angular momentum catastrophe”

Fall 1983 Catelan+Theuns 1996a,b; van den Bosch + 2001; Barnes+Efstathiou 1987 Governato+2010, Agertz+2011

j baryons

• more observable than j halo
• tracers: j*, jHa, jHI, jH2
• each with their own kinematics and mass profiles.

Angular momentum, stellar mass and morphology

• q depends on Hubble

type

• ! ~ 2/3
• hierarchical Universe
• Larger and later galaxies

have higher specific angular momentum.

Fall 1983, Romanowsky+Fall 2012

Angular momentum, stellar mass and morphology

• 3D fit:

– prefactor depends on B/T

• ! ~ 1

– disk stability

Obreschkow+Glazebrook 2014

Angular momentum, stellar mass and morphology

• ! ~ 1
• motivates this projection
• Galaxies with bigger

bulges have lower specific angular momentum per unit mass.

• B/T < 0.32

B/T Obreschkow+Glazebrook 2014

Bulge type

• pseudobulges

– rotationally-supported – secular evolution

• classical bulges

– pressure-supported – minor mergers or disk instabilities

• Galaxies with classical

bulges are more likely to have high bulge fraction than galaxies with pseudobulges

Toomre1977, Schweizer1990, Toomre1964 Kormendy+Kennicut2004, Wyse+1997 Fisher+Drory2016 Buta2011

Cosmic noon galaxies

• clumpy
• high gas fractions
• high rates of star formation
• enhanced turbulence
• predicted lower j/M
• alpha~2/3?

Buta2011

Glazebrook+1995b; Driver+1995; Abraham+1996a,b; Conselice+2000; Elmegreen+2005; Daddi+2010; Tacconi+2013; Bell+2005; Juneau+2005; Swinbank+2009; Genzel+2011; F ̈orster Schreiber+2009; Wisnioski+2011; Wuyts+2012; Fisher+2014; Lagos+2017; Teklu+2015; Obreschkow+15

Local analogues

• gain surface

brightness and spatial resolution

• are they truly

representative of high-z galaxies?

• Fisher+2017b

Baryons vs. haloes

• j baryons ~ j halo but
• baryons subject to more

physics

– feedback – tidal stripping PDF(j) j/jmean

space.com APOD

Catelan+Theuns1996; van den Bosch+2001; Sharma+Steinmetz2005

Open questions

Open questions

• Does the M* – j* – B/T relation extend to high bulge

fraction? Does it hold for classical bulges as well as for pseudobulge galaxies?

• What is the AM of high-z disks? Does it match that of

normal local disks? Does it match that of "local analogues"? What does that imply about their evolution?

• Can the distribution of j be used as a tracer of evolutionary

processes? As a tracer of morphology? As a kinematic decomposition tool?

Spatially resolving galaxies

Spatially resolving galaxies

• Galaxies are not all smooth, exponential disks
• Single-fibre and long-slit observations miss key

structure

Integral field spectroscopy

• A spectrum at every pixel;

an image at every wavelength

ifs.wikidot.com

Integral field spectroscopy

• Advantages:

– mitigating kinematic misalignment – accounting for structure in the disk

• Disadvantages:

– observationally expensive

Sweet+2016; Cecil+2015; Obreschkow+2015

Integral field spectroscopic samples

• THINGS:

The HI Nearby Galaxy Survey

• Romanowsky+Fall 2012
• CALIFA: Calar Alto Legacy

Integral Field Area

• KGES: KMOS Galaxy

Evolution Survey at z~1.5 in COSMOS, CDFS and UDS

• DYNAMO: clumpy,

star-forming z~0.1 galaxies

Walter+08; Leroy+08; Romanowsky+Fall12; Falcon-Barroso+17; Mendez-Abreu+17; Tiley+ip; Green+10,14

Measuring angular momentum

Measuring angular momentum: objectives

• minimise beam-smearing

– need adaptive optics- assisted observations

• trace the bulk of j

– need natural seeing data

• extrapolate to r=∞

– need model estimate

Sweet+2018,2019

Measuring angular momentum: details

1. measure spaxel-wise angular momentum separately for the adaptive optics and seeing-limited kinematic data 2. calculate spaxel-wise model angular momentum

– consistent with observed to 5%

3. mean j=J/M is calculated by integrating over a combination of 1) and 2) based on S/N

• AO contributes in inner regions (69%)
• seeing-limited in outer regions (18%)
• model contributes elsewhere (13%)

• minimise beam-smearing
• trace bulk of j
• trace spatial variation

in disk

– order-of-magnitude improvement

Our best practice advantages

Obreschkow + Glazebrook 2014

Angular momentum & bulges

Does the M* – j* – B/T relation extend to high bulge fraction? Does it hold for classical bulges as well as for pseudobulge galaxies?

Angular momentum & bulges

• Sweet, Fisher, Glazebrook +: 2018ApJ...860...37S
• tracing M* – j* – B/T over wide range of B/T,

and accounting for bulge type

• sample: THINGS, RF12, CALIFA
• seeing-limited + model approach

Mass—angular momentum—bulge fraction

• 2D fit: alpha = 0.56 +- 0.06
• Agrees with previous even

though:

– different bulge-disk decomposition methods – long-slit observations in Fall83 – data to 1 effective radius in Cortese+16 – extended range of morphology wrt OG14

• Agrees with CDM predictions

for DM haloes.

– M* = f(Mh) as j* = f(jh) – The M* - Mh relation is complex; need to test j* - jh.

8.5 9.0 9.5 10.0 10.5 11.0 11.5 1.5 2.0 2.5 3.0 3.5 4.0 log(M∗ [Msol]) log(j∗ [kpc km s−1]) β = 0 β = 0.2 β = 0.4

• THINGS

RF12 CALIFA pseudobulge (nb<2) classical bulge (nb>2) 3D fit 2D fit

Mass—angular momentum—bulge fraction

• 3D M* – j* – B/T:

alpha = 1.03 +- 0.11

• fit using R package

hyper.fit

– Accounts for measurement uncertainty in all variables as well as intrinsic scatter.

8.5 9.0 9.5 10.0 10.5 11.0 11.5 1.5 2.0 2.5 3.0 3.5 4.0 log(M∗ [Msol]) log(j∗ [kpc km s−1]) β = 0 β = 0.2 β = 0.4

• THINGS

RF12 CALIFA pseudobulge (nb<2) classical bulge (nb>2) 3D fit 2D fit

Robotham + Obreschkow 2015

! = 1

• B/T ~ (j/M)-1 is physically-motivated via disk stability:

– as j/M increases, surface density decreases – surface density ~ inverse Toomre Q – Q ~ instability against (pseudo)bulge formation – Q ~ (B/T )-1

• empirically-supported

Obreschkow + Glazebrook 2014

The pseudobulge track

• Trend of increasing B/T

with decreasing j/M describes galaxies with B/T <~0.4 only.

−8.0 −7.5 −7.0 −6.5 0.0 0.2 0.4 0.6 0.8 log(j∗/M∗ [kpc km s−1 Msol

−1])

β

• THINGS

RF12 CALIFA all pseudobulge(nb<2) classical bulge (nb>2) OG14

The pseudobulge track

• Galaxies with B/T > 0.4

have higher j* than predicted

• Most host classical bulges
• cf EAGLE - absence of

gas-poor mergers.

−8.0 −7.5 −7.0 −6.5 0.0 0.2 0.4 0.6 0.8 log(j∗/M∗ [kpc km s−1 Msol

−1])

β

• THINGS

RF12 CALIFA all pseudobulge(nb<2) classical bulge (nb>2) OG14

Schaye+2015, Lagos+2017

The pseudobulge track

• pseudobulges follow a well-

defined track

• secular evolution builds the

pseudobulge as disk material is transported to the centre; small increase in B/T

• stellar outflows cause small

amount of j* to be lost and a smaller change in M*

• galaxy moves upwards and

to the left along the track

−8.0 −7.5 −7.0 −6.5 0.0 0.2 0.4 0.6 0.8 log(j∗/M∗ [kpc km s−1 Msol

−1])

β

• THINGS

RF12 CALIFA all pseudobulge(nb<2) classical bulge (nb>2) OG14

The pseudobulge track

• Galaxies that host

classical bulges do not follow a well-defined track

– often higher B/T for j*/M*

• Classical bulges built by

mergers, which can significantly increase both M* and j*

• large increase in B/T

moves galaxy above track

−8.0 −7.5 −7.0 −6.5 0.0 0.2 0.4 0.6 0.8 log(j∗/M∗ [kpc km s−1 Msol

−1])

β

• THINGS

RF12 CALIFA all pseudobulge(nb<2) classical bulge (nb>2) OG14

The pseudobulge track

• galaxies with pseudobulges trace a well-defined track

in B/T-j * /M * space, but classical bulges do not.

• opposite to the black hole mass -- galaxy velocity

dispersion relation

• This is consistent with earlier suggestions that

classical bulges are sensitive to black hole evolution, while pseudobulge evolution is linked to disk, which dominates the j * budget.

The pseudobulge track

• Take-home messages:
• growth of many bulges is linked to decrease in j * /M *
• secular evolution less efficient at building

pseuodobulges than mergers at building classical bulges.

Angular momentum at cosmic noon

What is the angular momentum of high-z disks? Does it match that of normal local disks? Does it match that of “local analogues”?

Angular momentum at cosmic noon

• Sweet, Fisher, Savorgnan+

2019MNRAS.tmp..719S

• 2 high-redshift galaxies

(KGES, z~1.5) and 12 local analogues (DYNAMO, z~0.1)

• combine natural seeing

KMOS & GMOS IFS + with adaptive optics IFS OSIRIS + model estimate

COSMOS 127977 UDS 78317 KMOS KMOS OSIRIS OSIRIS

Improved measures of j*

• ur spaxel-wise integration within

6.75% of traditional rotation curve model method

• accounts for non-exponential

disks e.g. bright clumps and irregular velocity fields typical of high-z galaxies

• at z~0, GMOS + OSIRIS improves

measurement uncertainty on j* from 13% or 16% to 10%.

• at z~1.5, using only natural

seeing, or only AO-assisted

• bservations gives more uncertain

j* than a combination of the two.

2.0 2.5 3.0 3.5 4.0 0.0 0.1 0.2 0.3 0.4 log(j∗ [kpc km s−1]) (j∗−˜j∗)/j∗ z ˜ 1. 5 DYNAMO

Angular momentum of local analogues

• Clumpy, turbulent

DYNAMO galaxies have low j* for stellar mass

• low j* cannot simply be

explained by B/T, since these systems do not have large bulges.

8.5 9.0 9.5 10.0 10.5 11.0 11.5 1.5 2.0 2.5 3.0 3.5 4.0 log(M∗ [Msol]) log(j∗ [kpc km s−1])

β = 0 β = 0.2 β = 0.4

• COSMOS_127977

UDS_78317

• z ~ 1. 5

DYNAMO THINGS RF12 CALIFA pseudobulge (nb<2) 3D fit 2D fit

Angular momentum of local analogues

• DYNAMO galaxies have

low B/T for j*/M*.

• still building bulges
• summing clump mass

brings in range of pseudobulge track

−8.5 −8.0 −7.5 −7.0 −6.5 0.0 0.2 0.4 0.6 0.8 log(j∗/M∗ [kpc km s−1 Msol

−1])

β

• COSMOS_127977

UDS_78317

• z ~ 1. 5

DYNAMO THINGS RF12 CALIFA pseudobulge z=0 pseudobulge

Angular momentum at cosmic noon

• COSMOS 127977

consistent with normal local disk galaxies: no evolution of j*/M* with redshift?

• UDS 78317 consistent

with local analogues, but is a merger

8.5 9.0 9.5 10.0 10.5 11.0 11.5 1.5 2.0 2.5 3.0 3.5 4.0 log(M∗ [Msol]) log(j∗ [kpc km s−1])

β = 0 β = 0.2 β = 0.4

• COSMOS_127977

UDS_78317

• z ~ 1. 5

DYNAMO THINGS RF12 CALIFA pseudobulge (nb<2) 3D fit 2D fit

Angular momentum at cosmic noon

• COSMOS 127977

consistent with normal local disk galaxies: no evolution of j*/M* with redshift?

• UDS 78317 consistent

with local analogues, but is a merger

−8.5 −8.0 −7.5 −7.0 −6.5 0.0 0.2 0.4 0.6 0.8 log(j∗/M∗ [kpc km s−1 Msol

−1])

β

• COSMOS_127977

UDS_78317

• z ~ 1. 5

DYNAMO THINGS RF12 CALIFA pseudobulge z=0 pseudobulge

Distance to the pseudobulge track

• galaxies that deviate more are

– more dispersion-dominated due to turbulence / clumpiness – still in the process of building their bulges

• trend brings COSMOS

127977 into alignment with local analogues

• more complete picture of

galaxy evolution

−0.5 −0.4 −0.3 −0.2 −0.1 0.0 0.1 0.2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 ∆(Beta − log(j∗/M∗)) σ V COSMOS_127977

• z ~ 1.5

DYNAMO THINGS pseudobulge

Fisher+2017b, Cava+2018

Angular momentum evolution

• Take-home messages:
• scatter in M*-j* explained by

– variations in B/T – evolution with redshift – increased turbulence – treating mergers as rotating disks

Spatially-resolved angular momentum

Can the distribution of j* be used

• as a tracer of morphology?
• as a tracer of evolutionary processes?
• as a kinematic decomposition tool?

Spatially-resolved angular momentum

• Sweet, Fisher, Glazebrook + 2018b:

IAU proceedings; 10.5281/zenodo.1481585

• 25 barless CALIFA galaxies observed to >3re

Probability density function of j*

• motivation:

investigate distribution of j* and its utility as a tracer of morphology or a kinematic decomposition tool.

– fewer assumptions to measure j* than determine B/T

• especially at at high-z

– PDF(j*) contains more information than photometry or kinematics alone

Probability density function of j*

• spaxel-wise j*
• includes dispersion
• weighted by mass
• normalised to mean j*
• histogram => PDF

Hubble type ~ B/T ratio

NGC 6063

• Sc
• B/T = 0.04
• PDF(j) is broad,

symmetric, peaks near 1

PDF(j) j/jmean

NGC 2592

• S0
• B/T = 0.54
• PDF(j) is strongly-skewed,

peaks near 0

PDF(j) j/jmean

NGC 7653

• Sb
• B/T = 0.33
• PDF(j) is intermediate

PDF(j) j/jmean

PDF(j*/j*mean) is different to predicted

Sharma+Steinmetz 2005

• predictions fail:

– DM halo ~ Gamma fn – disk ~ x exp (-kx)

• fits for normal+lognormal

components fail

– degenerate – insufficient spatial resolution

Quantifying shape of PDF(j)

• Statistics of the underlying distribution:

– mean == 1 by construction – second moment standard deviation describes width – third moment skewness describes one-sidedness – fourth moment kurtosis describes tailedness

• kurtosis best for tracing galaxy morphology

PDF(j*) shape & galaxy morphology

• Galaxies with earlier

Hubble type T have more strongly tailed PDFs

• PDF shape correlates

more with T than with B/T

• T and PDF encode more

physical information than simple B/T ratio

0.6 0.8 1.0 1.2 1.4 Fisher Hubble type log(kurtosis of PDF(j∗/j*mean)) NGC2592 NGC6063 NGC7653

• Sd

Scd Sc Sbc Sb Sab Sa S0a S0 E

PDF(j*) shape & galaxy components

22 21 20 19 18 0.6 0.8 1.0 1.2 1.4 CALIFA disk central surface brightness log(kurtosis of PDF(j∗/j∗mean)) NGC2592 NGC6063 NGC7653

• 24

23 22 21 20 19 18 17 0.6 0.8 1.0 1.2 1.4 CALIFA bulge surface brightness log(kurtosis of PDF(j∗/j∗mean)) NGC2592 NGC6063 NGC7653

• bulge surface brightness correlates with PDF shape
• disk central surface brightness doesn’t

PDF(j*) shape & galaxy components

22 21 20 19 18 0.6 0.8 1.0 1.2 1.4 CALIFA disk central surface brightness log(kurtosis of PDF(j∗/j∗mean)) NGC2592 NGC6063 NGC7653

• 24

23 22 21 20 19 18 17 0.6 0.8 1.0 1.2 1.4 CALIFA bulge surface brightness log(kurtosis of PDF(j∗/j∗mean)) NGC2592 NGC6063 NGC7653

• bulge drives distribution of j*: – SMBH contribution
• distribution of j* is the same for disks of all sizes

– scale-invariance

PDF(j*) & kinematic decomposition

• a first look at bulge-disk

decomposition

• radius-selected bulge:

inner-most spaxels (yellow)

• remaining disk spaxels

(red)

• now: a priori B/T
• future: iterative

PDF(j*) & kinematic decomposition

• dispersion-selected bulge
• sensitive to clumps, thick

disk

PDF(j*) at cosmic noon

• Circular-only velocity

– no stellar dispersion measurement

• COSMOS 127977

intermediate between pure disk and bulge- dominated examples

• typical high-z morphology:

dispersion-dominated clumps embedded in a rapidly-rotating disk.

PDF(j*) takehomes

• Parametrizing the shape of the PDF with kurtosis

seems the sensible first step as a tracer of fundamental kinematic morphology – e.g. Hector

• In future (with better resolution – e.g. MAVIS) we will

be able to distinguish

– thin disk + bulge/thick disk/bar – clumps and bulges – pseudobulges from classical bulges

Conclusions

Conclusions

• Does the M* - j*- B/T relation extend to high bulge fraction?

Does it hold for classical bulges as well as for pseudobulge galaxies?

• What is the AM of high-z disks? Does it match that of

normal local disks? Does it match that of "local analogues"?

• Can the distribution of j be used as a tracer of evolutionary

processes? As a tracer of morphology? As a kinematic decomposition tool?

Conclusions

• Bulge growth linked to decrease in j*/M*, up to moderate B/T ~ 0.4.
• Secular evolution is less efficient at building pseudobulges than

mergers are at building classical bulges.

• Galaxies at cosmic noon have a wide range of angular momentum

properties due to high turbulence due to star formation, or because they are mergers.

• The probability density function of specific angular momentum is

more strongly tailed for earlier types with bigger bulges.

• In future (with better resolution) we will be able to decompose the

PDF(j*) into components and bulge types.