weak emission line quasars - university of texas … · weak emission line quasars ... colors and...

Ohad ShemmerUniversity of North Texas

The Inner Regions of Quasars, UT Austin, September 13, 2014

Weak Emission Line Quasars:

Confronting New Challenges in Understanding the Broad Emission Line Region

S.F. Anderson (U. Washington)

W.N. Brandt (Penn State)

A.M. Diamond-Stanic (U. Wisconsin)

X. Fan (U. Arizona)

P. Hall (York U.)

R. Lane (UNT)

S. Lieber (UNT)

P. Lira (U. Chile)

B. Luo (Penn State)

H. Netzer (Tel Aviv U.)

R. Plotkin (U. Michigan)

G.T. Richards (Drexel)

D.P. Schneider (Penn State)

M. Stein (UNT)

M.A. Strauss (Princeton)

B. Trakhtenbrot (ETH)

J. Wu (CfA)


In Collaboration with:


Outline✴ Weak emission line quasars (WLQs):

history and mystery

✴ Insights from multiwavelength observations

✴ What are WLQs telling us about the BELR

and SED?

✴ Summary of key questions


Weak Emission Line Quasars (WLQs):

History and Mystery


~100 SDSS sources, mainly at high redshift, with quasar-like continua but extremely weak or undetectable emission lines in their UV spectra.

Fan+99

Prototype: SDSS J153259.96-003944.1

Selected as a high-redshift

quasar candidate based on its

colors and lack of proper motion.

WLQs: History and Mystery

Have we found the long-sought BL Lacs at high redshift?

Diamond-Stanic+09

WLQs constitute >3σ deviation

Distributions of broad emission line equivalent widths (EWs) in SDSS quasars at z>3

WLQs were originally defined as quasars having EW ≤ 15.4 Å for the Lyα+N V emission-line complex (EW ≤ 10 Å for C IV).



Q1: What determines the shape of the line EW distributions?


Why are the UV emission lines in WLQs so weak or absent?


1100 1200 1300 1400 1500 1600 1700 1800Rest-Frame Wavelength [Å]

0

5

10

15

F! [

arbi

trar

y un

its]

Ly"

N V + Si II

S IV + O IV]

C IV

SDSS Quasar Template

Mean WLQ Spectrum

HST Spectrum of PHL 1811

Shemmer+09


Following Bev’s footsteps...

✴ Orientation?

✴ Obscuration?

✴ Polarization?

✴ Radio loudness?

✴ X-ray weakness?


...


Insights

from

Multiwavelength Observations

Insights from Multiwavelength Observations


UV

- op

tica

l

✴ quasar luminosities (LBol ~1047-1048 erg s-1).

✴ typical (blue) quasar continua.

✴ no broad absorption lines.

✴ no significant variability.

✴ no significant polarization.

✴ no detection of multiple images (not lensed).

Clues from the UV - optical band

(e.g., Diamond-Stanic+09):


X-ra

y

✴ no sign of significant absorption.

✴ typical quasar power-law spectra.

X-ray clues:

Shemmer+09

Joint fitting the X-ray spectra of

11 WLQs

(However, so far, results are tentative: based mostly on shallow Chandra observations.)

ΓNH


103

0.01

Coun

ts s

1 keV

1

SDSS J123132.38+013814.0z=3.2

10.5 2 5

20

2

Observed Frame Energy (keV)

Shemmer+09

Γ≅1.8

rad

io -

opti

cal -

X-ra

y

If WLQs are high-redshift BL Lacs, then where is the ‘parent’ population of X-ray and radio bright weak-lined sources at such redshifts?


5

4

3

2

1

0

-1

log R

-2-1.5-1-0.5ox

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

ro

WLQ at z>2.2 (this work)WLQ at z>2.2 (S06)WLQ at z<1 (S06)BL Lac (DXRBS)BL Lac (EMSS)BL Lac (RGB)BL Lac (Slew)BL Lac (SDSS)

Radio weakness

X-ray weakness

Radio-Quiet AGNsPHL 1811

2QZ J215454.3-305654

PG 1407+265

Shemmer+09; Plotkin+10

(Optical - X-ray `color’)

(Opt

ical

- R

adio

`co

lor’)

typical quasars



UV

- op

tica

l - m

id-IR

0.2

1

0.2

1

F [

Arb

itrar

y U

nits]

0.1 0.2 0.3 1 2 3 4 5Rest-Frame Wavelength [µm]

0.2

1

0.2

1a)

a)

b)

c)

d)

Diamond-Stanic+09; Lane+11

LBLPLBB

Ordinary-quasar SED

Luminous-quasar SED

z~6 quasar SED

Stacking SEDs of 18 WLQs using data from SDSS+NIR+Spitzer (Lane+11). WLQ SED is consistent with an ordinary-quasar SED.


WLQs are unbeamed quasars. The UV emission lines are intrinsically weak. Can be selected only via spectroscopic surveys.


What are WLQs Telling Us

About the BELR and SED?


Wu+11

Orientation? (see Niel Brandt’s talk for more details)What are WLQs Telling Us About the BELR and SED?


Laor & Davis+11

`Cold’ accretion disk?

Disk SED peak frequency decreases as black-hole mass increases. Fewer energetic photons are available for ionizing the BELR. Currently investigating new HST UV spectra of WLQs - searching for a predicted continuum drop below 1000Å (Plotkin et al., in prep.).

What are WLQs Telling Us About the BELR and SED?


Leighly+07

1100 1200 1300 1400 1500 1600 1700 1800Rest-Frame Wavelength [Å]

0

5

10

15

F! [

arbi

trar

y un

its]

Ly"

N V + Si II

S IV + O IV]

C IV


Mean WLQ Spectrum

HST Spectrum of PHL 1811

Shemmer+09

Are there low-redshift analogs to high-redshift WLQs?

Soft SED responsible for weak high-ionization BELR lines in WLQs?

Extremely high accretion rates?Clues from low redshift

4000 4500 5000Rest-Frame Wavelength [Å]

5.5

6

6.5

F [a

rbitr

ary

units

]

H

H

Fe II

H

Fe II


PHL 1811WLQ ?

Rest-Frame Wavelength [Å]

Fe II

Fe II

PHL 1811


Determining low-ionization emission line EWs and Eddington ratios[L/LEdd ∝ νLν(5100Å)0.5 FWHM(Hβ)-2]. Requires NIR spectroscopy.


Gemini-North VLT

4400 4600 4800 5000 5200 5400Rest-Frame Wavelength [Å]

1

2

F ! [1

0-17 e

rg s

-1 c

m-2

Å-1

]

H" [O III]

H" [O III]

1

2

3 SDSS J1141+0219

SDSS J1237+6301H"

[O III]

Shemmer+10z ~ 3.5

z ~ 3.5

– 30 –

1000 2000 3000 4000 5000 6000 7000 Rest!frame Wavelength (Å)

F ! (A

rbitr

ary

Uni

ts)

SiIV CIV CIII] MgII H" H#

SDSS J0836

z=1.749

SDSS J0945

z=1.683

SDSS J1321

z=1.422

SDSS J1411

z=1.754

Fig. 1.— Full X-shoter spectra. The UVB, VIS, and NIR arms are shown in blue, orange, and red,

respectively.

Plotkin+14, in prep.PRELIMIN

ARY

Exceptionally weak Hβ lines

Are the BELRs in WLQs unusually gas deficient or subject to exotic ionization conditions?


Extremely high accretion rates or anemic BELRs?


`Modified Baldwin Relation’

Excluding BAL quasars and RLQs

0.01 0.1 1L / LEdd

1

10

100

EW(C

IV)

[Å]

Baskin & Laor (2004)



0.01 0.1 1L / LEdd

1

10

100

EW(C

IV)

[Å]

Baskin & Laor (2004)Shemmer & Lieber, in prep.






0.01 0.1 1L / LEdd

1

10

100

EW(C

IV)

[Å]

Baskin & Laor (2004)Shemmer & Lieber, in prep.WLQ


Q2: Can we determine the Eddington ratio for all quasars?





XMM-Newton observations of WLQs to cross-check L/LEdd determinations: utilizing the hard-X-ray power-law photon index as an accretion-rate indicator (Stein+14, in prep.).

XMM-Newton

-1.5

-1.0

-0.5

0.0

log

(L /

L Edd)

L (5100Å) < 1046 ergs s-1

L (5100Å) > 1046 ergs s-1

BCESFITEXYMLE

1.5 2 2.5-1.0-0.50.00.51.0

log

(L /

L Edd)

BCES

log(L/LEdd) = (1.0 ± 0.3)! ! (2.5 ± 0.5)

Shemmer+08

XMM-Newton

Γ≅1.9




0.01 0.1 1L / LEdd

1

10

100

EW(C

IV)

[Å]







-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8log [EW(C IVobs) / EW(C IVpred)]

0

10

20

N

typical quasarsWLQs

Do WLQs mark a brief evolutionary phase thatall quasars go through (e.g., Hryniewicz+10)?

WLQs constitute a ≳4σ deviation.

Q3: How long does it take the BELR to form?



✴ Obtain additional NIR and X-ray spectra of WLQs.

✴ Analyze available HST/STIS spectra of six WLQs.

✴ Measure the relative strengths of low- and high-ionization emission lines in WLQs and compare with ordinary quasars.

✴ Check the dependence of relative line strengths on C IV blueshift.

✴ Ultimate goal: in conjunction with photoionization modeling, understand the roles that L, MBH, L/LEdd, the SED, density, and covering factor, play in determining the relative strengths of the BELR lines.

On the To Do List:



Summary of Key Questions

Summary of Key Questions


Feedback Welcome!

Q1: What determines the shape of the line EW distributions?

0.01 0.1 1L / LEdd

1

10

100

EW(C

IV)

[Å]


Q2: Can we determine the Eddington ratio for all quasars?

Q3: How long does it take the BELR to form?

-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8log [EW(C IVobs) / EW(C IVpred)]

0

10

20

N

typical quasarsWLQs

weak emission line quasars - university of texas … · weak emission line quasars ... colors and...

Documents