An excess of extreme TeV blazars from the Fermi LAT distribution on - PowerPoint PPT Presentation

An excess of extreme TeV blazars from the Fermi LAT distribution on the voidiness parameter T.A. Dzhatdoev, E.V. Khalikov (Moscow State University) 2017.07.18

Some abbreviations and definitions E 0 ― primary energy of a γ-ray (source restframe) E p0 ― primary energy of a proton (source restframe) z ― redshift; τ ― γγ pair production optical depth; γ ― spectral power-law index (when γ is a number) HE ― high-energy (E>100 MeV); VHE ― very high energy (E>100 GeV) EBL ― extragalactic background light; EGMF ― extragalactic magnetic field CMB ― cosmic microwave background PP ― pair production γγ→e + e - IC ― inverse Compton e - γ→e - 'γ' or e + γ→e + 'γ' SED ― spectral energy distribution K12 ― Kachielriess et al., Comp. Phys. Comm., 183 , 1036 (2012) BK16 ― Berezinsky & Kalashev, Phys. Rev. D, 94 , 023007 (2016) BW15 ― Biteau & Williams, ApJ, 200 , 58 (2015) KD10 ― Kneiske & Dole, A&A, 515 , A19 (2010) G12 ― Gilmore et al., MNRAS, 422 , 3189 (2012) H16 ― Horns, astro-ph/1602.07499 (2016) HM12 ― Horns & Meyer, JCAP, 033 (2012) NV10 ― Neronov & Vovk, Science, 328 , 73 (2010)

Typical AGN geometry (left, 1 ) and SED of blazar (right, 2 ); SED of EBL for several models (bottom, 3 ) 1 C.M. Urry, P. Padovani, PASP, 107 , 803 (1995); more details in: R. Antonucci, ARA&A, 31 , 473 (1993) 2 M. Boettcher et al., ApJ, 768 , 54 (2013) 3 EBL model KD10 is shown by red dashed line Extreme TeV blazars: high-energy peak in the SED at E>1 TeV, usually slow variability

The technique of calculations: Dzhatdoev et al., A&A, 603 , A59 (2017) (astro-ph/1609.01013) (hereafter D17) We are grateful to A. Kircheva and A. Lyukshin who participated in D17. D17 includes 6 extreme TeV blazars, 10 spectra (see Table 1 for journal references), (24 pages, ~70 plots) Cascade calculations: ELMAG (K12) E-mail for question and/or comments: timur1606@gmail.com

Some Fermi LAT blazars tend to be located towards the voids in the large scale structure (Furniss. et al., MNRAS, 446 , 2267 (2015) (F15), significance ~2.5 σ)

In these cases, observed flux is usually much higher (Furniss. et al., MNRAS, 446 , 2267 (2015), significance ~2.5 σ); x: voidiness runs from 0 to 1 EGMF-dependent effects? (for current state of EGMF constraints see NV10 and many citing papers)

A model of 1ES 1218+304 spectrum. B= 0.1 fG, L= 5 Mpc Magnetic suppression factor: from Alves Batista, astro-ph/1704.05137 (z=0.13 instead of 0.182).

Prediction: CTA will discover a new population of blazars with very hard spectra for the case of lines-of-sight with small voidiness HESS 100 hours Fermi LAT 10 years CTA 100 hours CTA 1000 hours

Another model of 1ES 1218+304 spectrum. B= 1 fG, L= 1 Mpc. Why an order of magnitude difference with the previous result?

Why an order of magnitude difference with the previous result? The estimate below is from Neronov & Semikoz, Phys. Rev. D, 80 , 123012 (2009); Alves Batista et al., Phys. Rev. D, 94 , 083005 (2016)) The factor 5 comes from different assumed z, primary energy and D; other factor of ~2 – the result of different assumed EGMF structure and other issues

But maybe the model is far too speculative and is not supported by other studies?

Arsioli & Chang A&A 598 , A134 (2017) found a plenty of blazars with hard primary spectra in Fermi LAT data (of course, TS is usually lower than in the official Fermi LAT catalogues)

Other effects? Yes(!) The combined statistical significance of the following “anomalies” (deviations from the absorption-only model) is 6-7 σ γ Electromagnetic cascade model of extragalactic -ray propagation applied to blazars Aharonian et al., A&A, 349 , 11 (1999) Aharonian et al., A&A, 384 , 834 (2002) d’Avezac et al., A&A, 469 , 857 (2007) Murase et al., ApJ, 749 , 63 (2012) Takami et al., ApJ Lett., 771 , L32 (2013) D17

The high-energy anomaly (HM12, H16) ― colored symbols denote absorption-corrected data (significance: originally 4.2 σ). A similar effect: Rubtsov & Troitsky, JETP. Lett., 100 , 355 (2014)

The ratio of best-fit model spectra for electromagnetic cascade model and the absorption-only model. Electromagnetic cascade model predicts up to 3 times more flux at E=8 TeV

Indication for a magnetically broadened cascade (Chen et al., Phys. Rev. Lett., 115, 211103 (2015), p-value~0.01), EGMF: B= 0.01-1 fG A likelihood ratio map is shown

Hadronic cascade model of blazar emission Uryson, JETP, 86 , 213 (1998) Essey & Kusenko, APh, 33 , 81 (2010) Essey et al., Phys. Rev. Lett., 104 , 141102 (2010) Essey et al., ApJ, 731 , 51 (2011) Murase et al., ApJ, 749 , 63 (2012) Takami et al., ApJ Lett., 771 , L32 (2013) Essey & Kusenko, APh, 57 , 30 (2014) Zheng et al., A&A, 585 , A8 (2016) D17, our version: Blazar emits >1 EeV protons (luminosity and spectrum: Tavecchio 2014). The source is embedded in galaxy cluster (Meyer et al., 2013, central magnetic field B 0 ). The proton beam may encounter another cluster at z 0

Constraints on hadronic cascade models. B 0 = magnetic field strength in the center of the cluster, z c = the termination redshift of the proton beam, in color: significance of exclusion

The Cherenkov Telescope Array (CTA): low threshold (20 GeV), improved sensitivity and angular resolution Our work may be interesting for astrophysicists working with space-based telescopes such as Fermi LAT, GAMMA-400, and the emulsion gamma-ray telescope GRAINE

Conclusions I. There are several indications that the absorption- only model is incomplete II. Electromagnetic cascade model may, in principle, explain all these effects III. If the effect of F15 is not a statistical fluctuation, there should be a new, still not discovered population of blazars with very hard spectra (“flood-effect”). IV. Extragalactic magnetic field strength and structure is unknown; the values (in voids!) of 1 aG – 1 fG on the 1 Mpc scale are still viable

Additional slides

Angle distribution of the MBC pattern (computed with the code of Fitoussi et al., MNRAS, 466 , 3472 (2017)). E 0 = 10 TeV, observable energy 9-11 GeV, pencil- beam; B= 0.1 fG, L= 1 Mpc. Black – max. positional polar angle 0.01 rad, red – 0.03 rad, green – 0.1 rad, blue – 0.3 rad

Smeared and integral angle distributions of the MBC pattern. B= 0.1 fG, L= 1 Mpc

Integral angle distribution of the MBC pattern. B= 0.1 fG, L= 1 Mpc

“Delta-plot”, cascade spectra for primary monoenergetic emission (ELMAG (K12) and ECS)

Inoue et al., ApJ, 768 , 197 (2013)

Gamma-ray horizon (from HM12, EBL model: KD10)

The ratio of the optical depth (blue ― KD10 as implemented in ELMAG 2.02, black ― G12 model) to the one for the original KD10 model. Solid ― z= 0.186, dashed ― z= 0.287. aa 34

For further classification of hadronic cascade models see D16, Section 4

Impact of voidiness (K= 1.0, 0.6, 0.4, 0.3, 0.2).The source is 1ES 0229+200 (z= 0.188). The high-energy part is better fitted for K<0.6, the low-energy part ― for K from 0.3 to 0.6.

The ratio of best-fit model spectra for electromagnetic cascade model and the absorption-only model

Constraints on the EGMF (the first such paper Neronov & Vovk, Science, 328 , 73 (2010) obtained B>0.3 fG, later this bound was relaxed by several orders of magnitude (variability!), e.g. Dermer et al., ApJ Lett., 733 , L21 (2011)) No MBC, but it is still possible to observe pair-halo (PH) Magnetically broadened cascade (MBC) solution

Spectral signatures of the electromagnetic cascade model: 1) high-energy cutoff, 2) “ankle” 3) “magnetic cutoff” 4) second ankle. B= 0

Signatures of the intergalactic electromagnetic cascade model (summary) I. Spectral signatures: 1. High-energy cutoff 2. Ankle 3. Magnetic cutoff 4. Second ankle II. Angular signatures: 5. Magnetically broadened cascade (MBC) and/or 6. Pair halo (PH) III. Timing signatures. Example: energy-dependent delay IV. Signatures in a sample of objects. Example: voidiness-dependent effect of F15