The Population and Luminosity Function of AGNs in the SDSS

1
The Population and Luminosity Function of AGNs in
the SDSS

• Lei Hao
• Cornell University

Hao et al. 2005, AJ, 129, 1783 Hao et al. 2005,
AJ, 129, 1795
2
Low-Luminosity AGN

• Color selected AGN sample biased against the
  low-luminosity AGNs.
• Select from their spectroscopic characters,
  especially emission-line properties.

3
Low-Luminosity AGN
4
Narrow-line AGN selection
5
Stellar Subtraction

• The emission lines of AGN are often contaminated
  by the host galaxy absorption lines.
• Apply Principal Component Analysis (PCA) to pure
  absorption-line galaxies and use the eigenspectra
  as stellar templates.

6
Data Sample

• 100,000 galaxy and quasar spectra complete in
  1151 sq degrees.
• Main galaxy sample rlt17.77
• Low redshift (zlt0.33)

7
AGN Selections

• Broad Line AGNs

8
AGN Selections

• Broad Line AGNs

FWHM(Ha)gt1,200 km/s 1,317 are
identified
9
AGN Selections

• Broad Line AGNs

FWHM(Ha)gt1,200 km/s 1,543 are
identified

• Narrow Line AGNs

10
AGN Selections
11
AGN Selections

• Broad Line AGNs

FWHM(Ha)gt1,200 km/s 1,317 are
identified

• Narrow Line AGNs

3,074 are identified based on kewleys
criteria
10,700 are identified based on kauffmanns
criteria
12
The Luminosity Function of Seyferts

• We measure the luminosity function as a function
  of emission line luminosities.
• Assumption
• The luminosity of the nuclei is independent of
  the luminosity of its host galaxy.
• We measure the detection probability function for
  each AGN spectra in the sample using the
  Monte-Carlo simulation.

13
Luminosity Function Result
14
Luminosity Function Result

• AGN Number density
• 0.018/Mpc3, 19 of all galaxies
• Luminosity density
• 1.5 ?104 L?/Mpc3 in H? luminosity
• 1.11?106 L?/Mpc3 in r-band luminosity
• Nuclear activity contributes about 6 ? 10-3 of
  the total light of galaxies in r-band.

15
Ha luminosity to B band magnitude
16
LF compare With Previous Work
17
Luminosity Function Seyferts and QSOs
Croom et al. 2004
0.4ltzlt0.68
18
Luminosity Function of Type I vs. Type II
19
Luminosity Function of Type I vs. Type II
20
Host galaxy luminosity vs. Nuclear luminosity

• Galaxy Magnitude vs. Ha Luminosity plot

21
LF Assumption Check

• Imagine Lnu? Lhost
• Detection probability function underestimate
  (low L)
• overestimate (high L)
• LF will be flatter than it should be
• For a random chosen subgroup, the expected
  luminosity distribution will be biased compare
  with the observed luminosity distribution
  (low-redshift group shift to high luminosity)
• Divide the AGN sample into subgroups by their
  redshifts or host galaxy magnitudes.

22
LF Assumption Check
23
LF Assumption Check
24
Host galaxy luminosity vs. Nuclear luminosity

• Correlation test (Kendalls ?-statistic) study to
  check the independency.

• L(host) L(nuclei) ?

• ? is close to 0 for all the LFs.

• The principal quantity that determines the AGN
  luminosity is the Eddington ratio, not the black
  hole mass.

25
Conclusion

• A complete optical spectroscopically selected AGN
  catalog
• The LF probes a much lower-luminosity range.
• The number density of AGNs 20 of all galaxies,
  contribute 6 ? 10-3 of the total light of
  galaxies in r-band.
• The ratio of Seyfert1s vs. Seyfert 2s is a
  function of luminosity about one at low
  luminosity while at high luminosity Seyfert 1s
  outnumber Seyfert 2s by 2-4.
• The nuclear luminosity is not strongly correlated
  with the host galaxy luminosity. The principal
  quantity that determines the AGN luminosity is
  the Eddington ratio, not the black hole mass.

26
The mid-IR spectra of AGNs

• Low resolution modules (SL,LL)
• 5-40?m wavelength coverage
• Resolution 64-128
• High resolution modules (SH, LH)
• 9-38?m wavelength coverage
• Resolution 600

27
Unification model
Type I Seyfert1s, QSO, etc.
Type II Seyfert2s, QSO2s etc.
28
Quasars silicate emission
29
AGN mid-IR spectra
30
AGN mid-IR spectra
Seyfert1
Seyfert2
31
AGN mid-IR spectra
32
Expected
Type I
Silicate emission
Type II
Silicate absorption
Observed
AGN dusty structure is very complicated
33
Ongoing Work

• Spitzer mid-IR spectra of the AGNs selected from
  the SDSS sample
• Detailed study of the torus structure of the
  optical selected AGNs.