EXPECTED GAMMA-RAY EMISION FROM X-RAY BINARIES W lodek Bednarek - PowerPoint PPT Presentation

EXPECTED GAMMA-RAY EMISION FROM X-RAY BINARIES W� lodek Bednarek Department of Astrophysics, � L´ od´ z, Poland

I have to admit ⇓ I don’t know ? ⇓ It seems to me very complicated ! ⇓ I will try to convince you why I have such opinion

• Historical notes • Observations • General scenarios • Conditions within the binary system • IC e ± pair cascade processes (linear, isotropized, magnetic field driven) • Effects of e ± pair energy losses in the magnetic field • Dependence on the shock localization (variable stellar wind) • Effects of anisotropic stellar/pulsar winds • Effects of clumpy stellar wind • Effects of relativistic boosting of radiation • Double shock structure - acceleration of two populations of electrons • Acceleration of electrons and/or hadrons ?

Historical notes: • Pulsars within binaries → high energy emission (Bignami et al. 1977, Vestrand & Eichler 1982 - Cyg X-3; Maraschi & Treves 1981 - LS I 61 303) • TeV-PeV γ -ray emission (???): Her X-1, Vela X-1, Cyg X-3, ... (lack of confirmation - Weekes 1992) • Absorption of γ -rays in stellar radiation (Protheroe & Stanev 1987, Moskalenko et al. 1993) • Some binaries → GeV emitters (?) - EGRET error boxes LS 5039 (Paredes et al. 2000), Cyg X-3 (Mori et al. 1997), LS I 61 303 (Thompson et al. 1995), Cen X-3 (Vestrand et al. 1997) • IC e ± pair anisotropic cascades in the stellar radiation (Bednarek 1997, 2000) • Discovery of γ -ray binaries at TeV energies (LS2883/PSR1259 - Aharonian et al. 2005; LS 5039 - Aharonian et al. 2005; LS I 61 303 - Albert et al. 2006) • Recent modelling of γ -ray emission since 2005:

γ -ray observations - main features • Modulation of γ -ray signal from LS 5039: GeV light curve TeV light curve Figure 1: LS 5039: GeV emission from Abdo et al. (2011); TeV emission from Aharonian et al. 2006.

• Long term γ -ray variability Figure 2: LS I 61 303 TeV γ -ray light curves (2008-2010) from Acciari et al. (2011).

• Spectral features LS I 61 303 Eta Carinae Figure 3: LS I 61 303 from Abdo et al. (2011); Eta Carinae from Farnier et al. (2011).

Three types of gamma-ray binaries • Massive star + energetic pulsar: LS 2883/PSR1259-63, LS 5039, LSI 303 +61 • Massive star + accreting black hole: Cyg X-3, Cyg X-1 (?) • Two massive stars: Eta Carinae (1) The geometry of acceleration may or may not differ significantly: (2) Physical conditions rather differ significantly ( V p , ξ , B): jet shock rad rad 1 rad 2 star star disk pul

Conditions within the binary: Massive star Magnetic field structure Wind structure rad polar wind dip tor equatorial star star wind • B ( R ) ∝ R − 3 (dip); ∝ R − 2 (rad); ∝ R − 1 (tor). • Polar wind: v ∼ 10 3 km/s; Equatorial wind: v ∼ 10 − 100 km/s;

Propagation of γ -rays within binary system Figure 4: Star: surface temperature T ⋆ = 3 × 10 4 K, radius R ⋆ = 8 . 6 × 10 11 cm, distance of the injection place D = 1 . 4 R ⋆ , E γ = 1 TeV (from Bednarek 2000). Simple scaling for stars with other parameters: E o S T , T ⋆ , R ⋆ , D, α ) = S 3 T S R τ ( E o τ ( γ , T o , R o , D, α ), where S T = T ⋆ /T o , S R = R ⋆ /R o and D in R ⋆ or R o . γ

Types of the IC e ± pair cascade scenarios Aharonian et al. (2006), Cerutti et al. (2009); Bednarek (1997,2000,2006); Sierpowska & Bednarek (2005) magnetically driven linear isotropized γ e γ e e γ γ γ B γ γ star star star source source γ source γ γ Note: E e = 1 TeV, B = 1 G → R L ∼ 3 × 10 9 cm << R ⋆ . • Linear: γ -rays strongest to stellar limb • Isotropized: Focusing of γ -rays by stellar radiation • Magnetically driven: Re-directed γ -rays around B direction

Main features of the γ -ray cascades Figure 5: LS 5039 time averaged cascade spectrum: from Aharonian et al. (2006). Spectra: injected (dashed); cascade (solid); simple abs (dotted) GeV bump; TeV emission

Magnetically driven cascades: distribution of cascade γ -rays e γ star e γ star e γ star Figure 6: Distribution of γ -rays on the sky for injection angles: 90 o , 120 o , and 150 o (from Sierpowska & Bednarek 2005).

Synchrotron energy losses of e ± pairs P syn < P T IC ⇒ B s < B T = 40 T 2 G (stellar surface) 4 Figure 7: From Bednarek (1997): T s = 9 × 10 4 K. U B ∝ R − 4 , U rad ∝ R − 2 Periastron - TeV electrons → synchrotron losses important ? ⇓ Reason for some TeV γ -ray modulation (peri/apo) ?

Synchrotron spectra from cascade e ± pairs Synchrotron emission: primary electrons: Bednarek & Giovannelli (2007) Synchrotron emission: secondary cascade e ± pairs (constant B): ⇓ Figure 8: From Khangulyan et al. (2008); Bosch-Ramon et al. (2008)

Variable stellar wind: TeV emission at periastron ? Change in stellar wind ⇒ change in shock localization 2 α 1 1 α 2 star NS shock 1 shock2 Angles α 1 and α 2 differ significantly

Cascade spectra for different obs. angles Figure 9: IC e ± pair cascade spectra for different obs. angles: 30 o − 120 o (from Bednarek 2000)

Anisotropic stellar/pulsar winds Figure 10: Shock structure very complicated: from Sierpowska-Bartosik & Bednarek (2008). Complicated geometrical situations can be expected: • At some phases shock structures may change drastically • The shock structures may change with binary periods • Shock might appear very close to the pulsar or massive star

Both winds aspherical Pulsar wind aspherical: e.g. Bogovalov (1999) Be stellar wind aspherical: e.g. Waters et al. (1988) pulsar polar stellar wind wind NS pulsar wind pulsar shock II wind shock I NS Be star equatorial stellar wind pulsar wind

Shock structures: PSR 1259-63/SS2883 Figure 11: Location of the shock in PSR1259-63/SS2883: from Sierpowska-Bartosik & Bednarek (2008). Post-Shock flow can accelerate to γ ∼ 100: see Bogovalov et al. (2008) See also the case of LS I 61 303: Sierpowska-Bartosik & Torres (2009)

Effects of relativistic boosting of radiation See previous talk ⇒ Dr G. Dubus ⇓ Relativistic jet: Dubus et al. (2010a) Relativistic flow along pulsar cometary tail : Dubus et al. (2010b) Figure 12: From Dubus et al. (2010).

Double shock structure - two populations of electrons? Tavani & Arons (1997): PSR 1259-63/SS 2883 e,p radiation radiation e,p Eta Carinae WR wind wind Figure 13: Shock structure in massive binary system Eta Carinae: from Bednarek & Pabich (2011). different conditions at the shocks (B, ξ ) ⇓ acceleration of electrons (hadrons?) to different maximum energies

Gamma-ray emission from electrons accelerated at the shocks -9.5 ) -1 s -2 dN/dE / erg cm -10 2 -10.5 log(E -11 -11.5 -12 -5 -4 -3 -2 -1 0 1 2 3 log(E / GeV) Figure 14: Shock structure in massive binary systems: from Bednarek & Pabich (2011). Electrons from the shock in Eta Carinae wind (solid) and WR star (dashed/

Two populations of electrons in pulsar/massive star binaries ? sh , pul ≈ 63( ξ/B ) 1 / 2 ≈ 10 P 100 ( ξ − 1 D 12 /B 12 ) 1 / 2 E max TeV . (1) 4 ) ≈ 130 ξ 1 / 2 E max sh , w ≈ 1 . 3( ξB sh ) 1 / 2 ( D sh /T 2 − 4 B 100 /T 2 GeV . (2) 4 Figure 15: Energies of accelerated electrons at the pulsar, stellar shock: from Bednarek (2011, in preparation).

Effects of clumpy stellar wind ? Figure 16: From Zdziarski et al. (2010). Clumps R ∼ 10 11 cm; Pulsar wind mix (confined) with the matter and mag. field of clumps see also the model for jet-clump interaction, e.g. Owocki et al. (2009), Araudo et al. (2009)

Gamma-ray emission from electrons in clumpy wind Figure 17: Models: dominated by IC losses (upper), synchrotron losses (bottom): From Zdziarski et al. (2010). Pulsar: electrons with γ e ∼ 10 8 ; Stellar wind: magnetic field B ∼ 2 G. Transition between models: TeV γ -ray variability ?

Acceleration of electrons and/or hadrons ? Too strong synchrotron losses → Hadronic γ -rays ? ( Aharonian et al. 2005) Hadronic models: e.g. Romero et al. (2003,2005); Kawachi et al. (2004); Chernyakova et al. (2006); Torres & Halzen 2007; Araudo et al. (2009); Owocki et al. (2009); Bednarek & Pabich (2011) Figure 18: Neutrino spectra from Eta Carinae (Bednarek & Pabich 2011).

CONCLUSION: Many effects can play essential role in formation of emission features of γ -ray binaries • Very important role of geometry (processes occur aspherically) • Non-steady medium (aspherical, variable, inhomogenous winds) • Different radiation processes • Different populations of particles Binary systems are one of the best defined but quite complicated astrophysical objects ⇓ Reliable predictions of γ -ray emission features difficult