Gamma-Ray Luminosity Function of Blazars and tGamma-Ray Luminosity Function of Blazars and the Cosmic Gamma-Ray Background: Evidence fohe Cosmic Gamma-Ray Background: Evidence fo

r the Luminosity-Dependent Density Evolutionr the Luminosity-Dependent Density Evolution

Takuro Narumoto(Department of Astronomy, Kyoto Univ.)

Tomonori Totani(Department of Astronomy, Kyoto Univ.)

3rd Workshop on the Nature of Unidentified High Energy Sources@Barcelona4th July 2006

T. Narumoto & T. Totani, 2006, ApJ, 643, 81

1. INTRODUCTION1. INTRODUCTION

Extragalactic Gamma-Ray Extragalactic Gamma-Ray Background (EGRB)Background (EGRB)

However, the origin of the EGRB is still an open problem blazars? (e.g., Stecker & Salamon 1996; Chiang & Mukherjee 1998) galaxy clusters? (e.g., Loeb & Waxman 2000; Totani & Kitayama 2000) dark matter annihilation? (e.g., Oda, Totani, & Nagashima 2005)

Strong et al. (2004)

EGRET confirmed the presence of the extragalactic gamma-ray background (EGRB)

EGRET

Gamma-Ray Luminosity Function of BlazGamma-Ray Luminosity Function of Blazar and the EGRBar and the EGRB

Most of the identified extragalactic EGRET sources are blazars blazars are the most likely candidate for the origin of the EGRB

however, the gamma-ray luminosity function (GLF) of blazars and its cosmological evolution are poorly understood

the estimate of the blazar contribution is highly uncertain SS96 (Stecker & Salamon 1996) blazar contribution to the EGRB is ~ 100% problem : overpredict the number of low-redshift blazars

CM98 (Chiang & Mukherjee 1998) blazar contribution to the EGRB is only ~ 25%

Earlier studies treated the cosmological evolution of the blazar GLF as a Pure Luminosity Evolution (PLE)

Cosmological Evolution of the Cosmological Evolution of the AGN X-Ray Luminosity AGN X-Ray Luminosity

Function (XLF)Function (XLF)

Peak redshift of the density evolution increases with AGN luminosity (e.g., Ueda et al. 2003; Hasinger et al. 2005)

Evolutionary nature of the AGN XLF is best described by the Luminosity-Dependent

Density Evolution (LDDE)

Ueda et al. (2003)

Redshift

Nu

mb

er

den

sit

y low luminosity

high luminosity

On the other hand, the cosmological evolution of the luminosity function of AGNs has been investigated in various wavelengths

Radio Detectability for Blazar Identification

• Most of the EGRET blazars are identified by finding radio counterparts, and they would remain unidentified if their radio counterparts are under the flux limit of radio surveys

identification probability must be included in the analysis

this probability was included in CM98, but it was calculated by assuming no correlation between gamma-ray and radio luminosities of blazars

we assume the gamma-ray and radio luminosity correlation based on the observations

In this study…

For the first time, in addition to the PLE, we introduce the LDDE into the blazar GLF and perform the likelihood analysis for the redshift and luminosity distribution of the EGRET blazars

In the likelihood analysis, we introduce the gamma-ray and radio luminosity correlation with a modest dispersion which is consistent with observations to calculate the radio detectability

Then, we examine the blazar contribution to the EGRB and address the prospects for the GLAST mission

2. BLAZAR GLF 2. BLAZAR GLF MODELSMODELS

PLE ModelPLE Model We derive the blazar GLF from the flat-spectrum radio-loud

quasar (FSRQ) radio luminosity function (RLF) by assuming a linear relation between the gamma-ray and radio luminosities of blazars

Since the faint-end slope index is poorly constrained, we take it as a free parameter

and are constrained by likelihood analysis

normalization factor

1

p 1

LDDE ModelLDDE Model We construct the blazar GLF based on the AGN XLF by assu

ming a linear relation between the blazar gamma-ray luminosity and the AGN X-ray luminosity

Since the faint-end slope index is poorly constrained, we take it as a free parameter

and are constrained by likelihood analysis

normalization factor « 1

1

q 1

X-ray luminosity of normal AGNs (not blazars)

3. RESULTS3. RESULTS

Constraints from Likelihood Constraints from Likelihood AnalysisAnalysis

PLE modelBest-fit parameters

is quite similar to the value obtained directly from the EGRET blazars ( )

is smaller (i.e., flatter faint-end slope) than that of the FSRQ RLF ( )

28.3 p

23.3 p

69.0 ,28.3 1 p

69.01

83.01

Constraints from Likelihood Constraints from Likelihood AnalysisAnalysis

LDDE modelBest-fit parameters

is a little larger (i.e., steeper faint-end slope) than the value inferred from the AGN XLF ( ), but the AGN XLF value is within the 95% CL contour

19.1 ,80.3 1 q

19.11

10.087.01

faint-end slope of the AGN XLF

Redshift and Luminosity Distribution of tRedshift and Luminosity Distribution of the EGRET Blazarshe EGRET Blazars

KS probability (redshift)

LDDE model : 67.8%PLE model : 3.1%

KS probability (luminosity)

LDDE model : 99.3%PLE model : 27.0%

The LDDE model can explain the redshift and luminosity distribution of the EGRET blazars b

etter than the PLE model

Steep faint-end slope (within the 68% CL contour) can explain 100% of the EGRB

However, since the PLE model poorly fits the observed data, it is not appropriate to derive any conclusion

Blazar Contribution to the EGRB (PLE MBlazar Contribution to the EGRB (PLE Model)odel)

Best-fit PLE model can explain only 50~55% of the EGRB

85.01

Blazar Contribution to the EGRB (LDDE Blazar Contribution to the EGRB (LDDE Model)Model)

Best-fit LDDE model can explain only 25~50% of the EGRB

26.11

faint-end slope of the AGN XLF

Steep faint-end slope (within the 68% CL contour) can explain 100% of the EGRBHowever, such a steep faint-end slope is not favored from the AGN XLF

4. PREDICTIONS FOR 4. PREDICTIONS FOR THE GLAST MISSIONTHE GLAST MISSION

Expected Number of GLAST BlazarsExpected Number of GLAST BlazarsThe Number of blazars detectable by GLAST is~ 3000 : best-fit LDDE model~ 5250 : best-fit PLE model~ 10000 : SS96 model

strongly dependent on the blazar GLF models

The LDDE model predicts three times fewer blazars than the previous estimate

Blazar GLF and its evolution can be constrained from the number count of GLAST blazars

Contribution of GLAST BlazarsContribution of GLAST Blazarsto the EGRBto the EGRB

The contribution to the EGRB decreses with decreasing flux just below the EGRET sensitivity limit

The resolvable fraction of the EGRB by GLAST blazars is

~ 20% (best-fit LDDE)~ 26% (LDDE with steeper faint-end slope)

~ 33% (best-fit PLE)~ 42% (PLE with steeper faint-end slope)

There are two peaks of the contribution to the EGRB as a function of flux, and major contribution comes from bl

azars under the GLAST detection limitthick lines : best-fit modelsthin lines : models that can explain 100% of the EGRB (with steeper faint-end slope)

SUMMARYSUMMARY• The LDDE model can explain the redshift and luminosity distributi

on of the EGRET blazars better than the PLE model

• Only 25%~50% of the EGRB can be explained by the best-fit LDDE model

• 100% of the EGRB can be explained by the LDDE model with steeper faint-end slope, but such a model is not favored from the AGN XLF

• The LDDE model predicts considerably fewer (by a factor of more than 3) blazars down to the GLAST sensitivity limit, compared with the previous estimate

• Based on the LDDE model, the contribution to the EGRB will decrease with decreasing flux just below the EGRET sensitivity limit

Redshift and Luminosity Distribution of tRedshift and Luminosity Distribution of the GLAST Blazarshe GLAST Blazars

Distributions of the LDDE model are wider than those of the other models