SLACS Spectroscopy

31  Download (0)

Full text

(1)

SLACS Spectroscopy

Observations, Kinematics & Stellar Populations

Oliver Czoske

Kapteyn Institute, Groningen, NL

“Strong Gravitational Lensing in the Next Decade”

Cogne, 22 June 2009

(2)

Collaborators

L´eon Koopmans (Kapteyn) Matteo Barnab`e (Kapteyn)

Tommaso Treu (UCSB) Matt Auger (UCSB)

Adam Bolton (IfA, Hawai‘i)

Scott Trager (Kapteyn)

(3)

Spectra of early-type galaxies

Galaxy spectrum s

(

λ

)

: sum of (many) Doppler shifted stellar spectra t

(

λ

)

s(λ) ∼

Z

G(v) · t 

ln λv c

 dv

5100 5150 5200 5250 5300

3.0e−164.5e−166.0e−16

lambda

value

Galaxy (J2238), σσ ==200km/s

5100 5150 5200 5250 5300

0.20.61.0

lambda

value

Star (HD 195506), Indo−US

G

(

v

)

: line-of-sight velocity distribution (LOSVD)

Moments:

v

= R

G

(

v

)

v dv σ2

= R

G

(

v

)(

v

v

)

2 dv

=⇒

galaxy spectra contain information on stellar kinematics

(4)

A little history

Rudolf Minkowski (1895 – 1978)

Minkowski (1950s)

the telescope as “analogue computer”

Burbidge2 & Fish (1961) numerical convolution

Brault & White (1971), Simkin (1974) introduction of FFT into astronomy Sargent et al. (1977) (Paul Schechter)

Fourier quotient method: ˜G

˜s/˜t Tonry & Davis (1979)

cross-correlation method Franx et al. (1989)

fitting in Fourier space

Rix & White (1992), van der Marel & Franx (1993), Cappellari & Emsellem (2004)

fitting in pixel space

4

(5)

Spatially resolved kinematics

E

N

a1 a3 PA

Longslit spectroscopy:

major and minor axis profiles of v and σ

=⇒

at least some ellipticals rotate

Tensor virial theorem: 2K

+

W

=

2

(

T

+

12Π

) +

W

=

0 v

σ

=

2

(

1

δ

)

 Wxx Wzz



(

ε

) −

2

Wxx/Wzz: ratio of components of potential energy tensor δ

=

Πzz/Πxx: anisotropy of random kinematic energy tensor

Roberts (1962), Binney (1978)

(6)

Spatially resolved kinematics

Illingworth & Schechter 1981

δ = 0

δ > 0

Davies et al. (1983)

MB < −20.5

MB < −20.5

× bulges

• massive galaxies are supported by anisotropic pressure

• less massive galaxies and bulges are supported by rotation

• occasional minor axis rotation: kinematical hint at triaxiality

6

(7)

Integral-field spectroscopy

λ

λ

0

λ

x y

• lenslets + fibres

• image slicer

. . .

=⇒

Three-dimensional data cubes

(

x, y, f

(

λ

)

)

=⇒

Two-dimensional maps of

v

(

x, y

)

, σ

(

x, y

)

, . . .

• line strength, metallicities, star formation rate, . . .

(8)

State of the art: SAURON

48 local (!) elliptical and lenticular galaxies

• SAURON IFS on WHT

• cz

<

3000 km s1

• Detailed dynamical modelling

• But: require fairly strong

assumptions, e.g. constant M/L

Emsellem et al. (2004)

see also ATLAS3D

Goal of our project:

extend structural studies of

early-type galaxies beyond the local universe

8

(9)

SLACS IFS Gravitational Lensing

Advantages of gravitational lensing:

• robust and model-independent estimate of total mass contained within Einstein ring

• sensitive to all kinds of matter (DM + stars + gas + . . . )

• insensitive to dynamic state of matter But: model degeneracies!

These can be broken by combination with, e.g., kinematic information.

Lenses Structures and Dynamics (L. Koopmans, T. Treu)

• detection of DM halos at high significance

• inner total mass profiles close to isothermal

etc...

(MG2016, L´eon Koopmans)

(10)

SLACS IFS Sloan Lens ACS Survey

SLACS is a lens selected gravitational lens survey:

• candidates are chosen from the SDSS luminous red galaxy sample +

“quiescent” MAIN sample

• selection criterion: presence of additional emission lines at higher redshift

Bolton et al. 2006

• follow-up observations with ACS or WFPC2: confirmation of lensing hypothesis

lenses are guaranteed to be bright and not outshone by the lensed background objects

SLACS lenses are normal early-type galaxies Treu et al. 2006

SLACS is the ideal parent sample for a combined lensing/dynamical analysis

⇐⇒

(11)

SLACS IFS Sloan Lens ACS Survey

SLACS is a lens selected gravitational lens survey:

• candidates are chosen from the SDSS luminous red galaxy sample +

“quiescent” MAIN sample

• selection criterion: presence of additional emission lines at higher redshift

Bolton et al. 2006

• follow-up observations with ACS or WFPC2: confirmation of lensing hypothesis

lenses are guaranteed to be bright and not outshone by the lensed background objects

SLACS lenses are normal early-type galaxies Treu et al. 2006

SLACS is the ideal parent sample for a combined lensing/dynamical analysis

⇐⇒

(12)

SLACS IFS Project

Goals of this project:

• obtain two-dimensional maps of galaxy kinematics, i.e. v

(

R

)

and σlos

(

R

)

using

1. VIMOS/IFU on VLT: 17 systems (Czoske)

2. “pseudo integral field” spectra from Keck: 13 systems (Treu, Gavazzi, Auger)

• Combine lens modelling with detailed modelling of the kinematical information in a fully self-consistent way (Barnab`e)

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35

0100200300400

Redshift

Velocity Dispersion (km/s)

SAURON E/S0 (Cappellari et al. 2007)

11

(13)

The Integral Field Unit of VIMOS on VLT

Photos from http://www.oamp.fr/virmos/

20

21 40

41 60

61 80

1

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 80

A

B C D Mask 3

(a)

Quadrant 4 (to Mask 4) Quadrant 1 (to Mask 1)

(c) (b) Quadrant 2 (to Mask 2)

Quadrant 3 (to Mask 3)

A

D B

2 1

3 4 5

1 3 2

4 5

C

5 4 3 2 1

5 4 3 2 1

D

C

B

A

Zanichelli et al. (2005)

(14)

SLACS IFS Project

Observations:

ESO/Large programme with VIMOS/IFU

Spectral resolution: R

2500

−→

∆v

110 . . . 85 km s1 Fibre size: 0.67 arcsec

1 . . . 2 kpc

Data reduction: VIPGI Kinematic analysis:

Template fitting in pixel space

−→

v

(

x, y

)

, σ

(

x, y

)

J162746.44005357.5 zlens = 0.2076 zsource = 0.5241

σv = 275±12

J021652.54081345.3 zlens = 0.3317 zsource = 0.5235

σv = 332±23

J230053.14+002237.9 zlens = 0.2285 zsource = 0.4635

σv = 283±18

J230321.72+142217.9 zlens = 0.1553 zsource = 0.5170

σv = 260±15

13

(15)
(16)
(17)
(18)

Integrated spectra (SDSS aperture, 3 arcsec)

4000 4200 4400 4600 4800 5000

1e−163e−165e−16

SDSS−J0037 'HD 249 ' sigma = 245.3 gof = 7.49

Wavelength

Rel. Flux

4000 4200 4400 4600 4800 5000

−1e−160e+00

Residuals

Wavelength

Residuals

●●

●●

●●

●●

●●

●●

4500 5000 5500 6000

1.6e−162.2e−16

SDSS−J1451 'HD 114092' sigma = 203.9 gof = 8.51

Wavelength

Rel. Flux

4500 5000 5500 6000

−4e−170e+003e−17

Residuals

Wavelength

Residuals

●●

●●

●●

●●

●●

●●

●●

4500 5000 5500 6000

8.0e−171.4e−16

SDSS−J1251 'HD 145328' sigma = 201.5 gof = 4.38

Wavelength

Rel. Flux

4500 5000 5500 6000

−2e−172e−17

Residuals

Wavelength

Residuals

●●

●●

●●

●●

●●

●●

4500 5000 5500 6000

6.0e−171.2e−16

SDSS−J1627 'HD 195506' sigma = 272.6 gof = 5.04

Wavelength

Rel. Flux

4500 5000 5500 6000

−2e−171e−174e−17

Residuals

Wavelength

Residuals

●●

●●

●●

●●

●●

●●

direct pixel-fitting method

template chosen from IndoUS database ( 1000 spectra

covering wide range of stellar parameters)

single stellar template for each system, chosen from global spectrum, used for each spaxel

additive and multiplicative polynomials to correct for continuum and sensitivity variations

emission lines, Balmer lines, Mg b masked, as well as sky line residuals

16

(19)

Integrated spectra (SDSS aperture, 3 arcsec)

Even when the fit doesn’t look so good: source features

(20)

Integrated spectra: Stellar population analysis

Oops – nothing there yet. . .

18

(21)

Integrated spectra: Stellar population analysis

stellar population analysis to yield stellar M/L and stellar age

• combined lensing/dynamics model yields total mass distribution and stellar distribution function, but requires assumption stellar M/L

• combine with surface brightness and stellar M/L: distribution of dark matter

• How do the structural histories/properties of the galaxies compare to their star formation histories?

Stellar population studies will be more robust with larger wavelength coverage than VIMOS.

(22)

Results: Angular momentum parameter

−18 −19 −20 −21 −22 −23

0.00.10.20.30.40.50.60.7

MB λλR

SAURON SLACS/IFU

λR

=

i FiRi

|

vi

|

i FiRiq

v2i

+

σi2

Emsellem et al. (2007)

• fast rotators vs slow rotators, λR

0.1

• fast rotators tend to have low luminosity, MB

& −

20.5

• SLACS/IFU sample mostly slow rotators, consequence of selection of high velocity

dispersion systems

20

(23)

Results: v/σ–ε diagram

Binney (2005):

Updated version of v/σ–ε diagram, appropriate for two- (or

three?)-dimensional data:

h

ve2k

i

h

eσk2

i

:

=

i Fi v

2i

i Fi σi2 From tensor virial theorem derive

h

ve2k

i

h

eσk2

i = (

1

δ

)

Wxx/Wzz

1 α

(

1

δ

)

Wxx/Wzz

+

1 with

α = 1 Mhve2ki

Z

d3x vk(x) − vek(x)2 ρ(x)

For Hernquist profile with flat rotation curve: α

0.15

(24)

Results: v/σ–ε diagram

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

0.00.20.40.60.8

εε

vσσ

αα ==0 αα ==0.2

δδ ==0

δδ ==0.5

50 100 150 200 250 300 350

−0.10.00.10.20.30.40.50.6

σσ

δδ

SAURON SLACS/IFU

v2

σ2

= (

1

δ

)

Wxx/Wzz

1

α

(

1

δ

)

Wxx/Wzz

+

1 δ

=

1

Πzz Πxx

22

(25)

Future directions

• extend redshift coverage to z

1 with samples of adequate size

• extend mass coverage to σ

<

200 km s1

• better resolution: resolve central peak in velocity dispersion

• broad wavelength coverage for stellar population studies

Future lens surveys will provide ample material to choose from for in-depth studies of galaxy properties.

0.000100200300400 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Redshift

Velocity Dispersion (km/s)

SAURON E/S0 (Cappellari et al. 2007)

(26)

Next generation instruments and telescopes: XShooter on VLT

• large wavelength coverage: 3000 – 25,000 ˚A (three arms)

• spectral resolution: R

5100 (twice that of VIMOS)

• small field of view: 4

×

1.8 arcsec2

=⇒

GTO project to observe three massive lens galaxies at z

1 (PI: Koopmans)

24

(27)

Next generation instruments and telescopes: Parameterss

Resolution Field of view

(28)

Next generation instruments and telescopes: ELT et al.

• Telescopes of 30 m and more will be a reality within 10 years or so

• Spectroscopy will continue to be one of the main jobs of ground-based telescopes

• Integral-field spectrographs are complex, but provide high information density in their data

• Do lensing studies (not just of galaxies, but also of lensed sources, also in cluster lenses) have specific requirements for instrument design that we should try to push?

ESO-ELT: 42 m primary, fully adaptive

26

(29)

Conclusions

• Current IFS instrumentation gives useful data to begin studying galaxy kinematics beyond the local universe.

• VIMOS/IFU data on a subsample of SLACS lenses show that these form a continuation of the local SAURON sample

=⇒

massive, mostly slowly rotating galaxies

• full power of these data unleashed when combined with lensing information

=⇒

Matteo’s talk

• future instrumentation should improve coverage of galaxy parameters (redshift, mass), as well as improve the accuracy of kinematic studies

(30)

Contents

1

2 Collaborators

3 Spectra of early-type galaxies 4 A little history

5 Spatially resolved kinematics 6 Spatially resolved kinematics 7 Integral-field spectroscopy 8 State of the art: SAURON

9 SLACS IFS Gravitational Lensing

10 SLACS IFS Sloan Lens ACS Survey

11 SLACS IFS Project

12 The Integral Field Unit of VIMOS on VLT

13 SLACS IFS Project

14 15

16 Integrated spectra (SDSS aperture, 3 arcsec) 17 Integrated spectra (SDSS aperture, 3 arcsec) 18 Integrated spectra: Stellar population analysis 19 Integrated spectra: Stellar population analysis 20 Results: Angular momentum parameter

21 Results: v/σ–ε diagram 22 Results: v/σ–ε diagram 23 Future directions

28

(31)

24 Next generation instruments and telescopes: XShooter on VLT 25 Next generation instruments and telescopes: Parameterss

26 Next generation instruments and telescopes: ELT et al.

27 Conclusions 28 Contents

Figure

Updating...

References

Related subjects :