1: \documentclass[12pt,preprint,fleqn]{aastex}

2: %\documentclass{emulateapj}

3: %\usepackage{mathtext}

4: %\usepackage{amsmath}

5: \usepackage{graphicx}

6: \sloppy

7: \nonfrenchspacing

8: 

9: \newcommand{\vdag}{(v)^\dagger}

10: \newcommand{\myemail}{mboett@helios.phy.ohiou.edu}

11: 

12: \slugcomment{Submitted to {\it The Astrophysical Journal Letters}}

13: 

14: 

15: \begin{document}

16: 

17: \title{Photon-photon Absorption of Very High Energy Gamma-Rays from

18: Microquasars: Application to LS 5039}

19: 

20: \author{Markus B\"ottcher\altaffilmark{1}

21: \&

22: Charles D. Dermer\altaffilmark{2}}

23: 

24: \altaffiltext{1}{Astrophysical Institute, Department of Physics 

25: and Astronomy, Ohio University, Athens, OH 45701, USA}

26: \altaffiltext{2}{E. O. Hulburt Center for Space Research, Code 7653

27: Naval Research Laboratory, Washington, D.C. 20375-5352}

28: 

29: 

30: \begin{abstract}

31: Very high energy (VHE) $\gamma$-rays have recently been detected from 

32: the Galactic black-hole candidate and microquasar LS~5039. A plausible

33: site for the production of these VHE $\gamma$-rays is the region close

34: to the base of the mildly
relativistic outflow. However, at distances 

35: comparable to the binary
separation, the intense photon field of the 

36: stellar companion leads to substantial $\gamma\gamma$ absorption of 

37: VHE $\gamma$-rays.
If the system is viewed at a substantial inclination 

38: ($i \ne 0$), this absorption feature will be modulated on the orbital 

39: period of
the binary as a result of a phase-dependent stellar-radiation 

40: intensity and pair-production threshold. We apply our results to LS~5039 

41: and find that 

42: (1) $\gamma\gamma$ absorption effects will be substantial if the

43: photon production site is located at a distance from the central

44: compact object of the order of the binary separation ($\approx 2.5\times 

45: 10^{12}$ cm) or less;

46: (2) the $\gamma\gamma$ absorption depth will be largest at a few

47: hundred~GeV, leading to a characteristic absorption trough; (3) the

48: $\gamma\gamma$ absorption feature will be strongly modulated on

49: the orbital period, characterized by a spectral

50: hardening accompanying periodic dips of the VHE $\gamma$-ray

51: flux; and (4) $\gamma$-rays can escape virtually 

52: unabsorbed, even from within $\approx 10^{12}$ cm, 

53: when the star is located behind the production site

54: as seen by the observer.

55: \end{abstract}

56: 

57: \keywords{gamma-rays: theory --- radiation mechanisms: non-thermal 

58: --- X-rays: binaries --- stars: winds, outflows}

59: 

60: 

61: \section{\label{intro}Introduction}

62: 

63: Recent observations \citep{aharonian05} of $\gtrsim 

64: 250$ GeV $\gamma$-rays with the High Energy Stereoscopic System (HESS)

65: from the X-ray binary jet source LS 5039 establish that microquasars

66: are a new class of $\gamma$-ray emitting sources. These results 

67: confirm the earlier tentative identification

68: of LS 5039 with the EGRET source 3EG~J1824-1514 \citep{paredes00}. 

69: In addition to LS~5039, 

70: the high-mass X-ray binary LSI~$61^o303$  (V615~Cas) 

71: also has a possible $\gamma$-ray counterpart in the MeV -- GeV

72: energy range \citep{gregory78,taylor92,kniffen97}.

73: 

74: Microquasars now join blazar AGNs as a firmly established 

75: class of very-high energy (VHE; with 

76: energies $\gtrsim$~a few hundred GeV -- TeV) $\gamma$-ray sources. 

77: The nonthermal continuum emission of blazars 

78: is believed to be produced in a relativistic plasma jet 

79: oriented at a small angle with respect to our line of sight. Their radio 

80: through UV/X-ray emission is most likely due to synchrotron emission by 

81: relativistic electrons in the jet, while the high energy emission can be 

82: produced by Compton upscattering of lower-energy photons off relativistic 

83: electrons \citep[for a recent review, see, e.g.][]{boettcher02}, or through 

84: hadronic processes \citep{mb92,ad01,muecke03}. 

85: 

86: Because of their apparent similarity with their supermassive AGN cousins, 

87: it has been suggested that Galactic microquasars may be promising sites 

88: of VHE $\gamma$-ray production \citep[e.g.,][]{bosch05a}. While earlier 

89: work on $\gamma$-ray emission from X-ray binaries focused on neutron star 

90: magnetospheres as $\gamma$-ray production sites 

91: \citep[e.g.][]{moskalenko93,moskalenko94,bednarek97,bednarek00},

92: the VHE $\gamma$-ray detection of LS~5039 suggests that the $\gamma$-ray

93: production is more likely to be associated with the jets. High-energy 

94: $\gamma$-rays of microquasars can be produced via hadronic 

95: \citep[e.g.][]{romero03} or leptonic processes. In the latter case,

96: the most plausible site would be close to
the base of the mildly 

97: relativistic jets, where ultrarelativistic electrons can Compton 

98: upscatter soft photons.

99: Possible sources of soft photons are the synchrotron radiation produced 

100: in the jet by the same ultrarelativistic electron population \citep[SSC =

101: synchrotron self-Compton;][]{aa99}, or external photon fields \citep{bp04,bosch05a}. 

102: Both LS~5039 and V615~Cas are high-mass X-ray binaries which are rather 

103: faint in X-rays, with characteristic 1 -- 10~keV luminosities of $\sim 

104: 10^{34}$~ergs~s$^{-1}$. This is much lower than the characteristic bolometric 

105: luminosity of the high-mass companions of these objects, at $L_{\ast} \gtrsim 

106: 10^{38}$~erg/s. Consequently, the dominant source of external photons in 

107: LS~5039 and V615~Cas is the companion's optical/UV photon field. 

108: 

109: The intense radiation field of the high-mass companion will, however, also

110: lead to $\gamma\gamma$ absorption of VHE $\gamma$-rays in the $\sim 100$~GeV

111: -- TeV photon energy range if VHE photons are produced close to the base

112: of the jet. For the case of LS~5039, \cite{aharonian05}

113: have estimated that VHE emission produced within $\sim 10^{12}$~cm of the

114: central engine of this microquasar would be heavily attenuated ($\tau_{\gamma\gamma}

115: \sim 20$ for $E_{\gamma} \sim 100$~GeV), and the observed spectrum of VHE 

116: photons would be hardened compared to its intrinsic shape. For comparison,

117: the $\gamma\gamma$ opacity due to stellar radiation back-scattered by the

118: stellar wind in the vicinity of the binary system (for which the angle of

119: incidence between the line of sight and the target soft photon field would

120: be more favorable for $\gamma\gamma$ absorption) may be estimated as

121: $\tau_{\gamma\gamma}^{\rm max} (r) \sim (L_{\ast} \, \dot M \, \sigma_T^2) / 

122: (2.7 \cdot 48 \, \pi^2 \, r^2 \, m_p \, v_{\infty} \, m_e c^3 \, \epsilon_{\ast})$,

123: where $L_{\ast}$ is the stellar luminosity, $\dot M$ is the mass outflow rate, 

124: $v_{\infty}$ is the terminal velocity of the stellar wind,

125: and $\epsilon_{\ast}$ is the dimensionless mean photon energy of the stellar

126: radiation field. For LS~5039, $L_{\ast} \sim 7 \times 10^{38}$~ergs~s$^{-1}$,

127: $m_e c^2 \, \epsilon_{\ast} \sim 3.5$~eV, $\dot M \approx 10^{-6.3} M_{\odot}$/yr,

128: and $v_{\infty} \approx 2500$~km~s$^{-1}$ \citep{mg02}, which yields

129: $\tau_{\gamma\gamma}^{\rm max} (r) \sim 0.1 / r_{12}^2$, where $r =

130: 10^{12} \, r_{12}$~cm. Consequently, the $\gamma\gamma$ opacity will be

131: strongly dominated by the direct stellar photon field.
In the same spirit,

132: one can also estimate the effect of $\gamma$-ray absorption in the field of

133: atomic nuclei in the stellar wind. Using a total absorption cross section 

134: per unit mass of $\sigma_H \sim 0.012$~cm$^2$~g$^{-1}$ for GeV -- TeV 

135: $\gamma$-rays, we find $\tau_{\gamma Z} \approx 
\sigma_H \, \dot{M} 

136: / (4 \pi \, v_{\infty} \, r_0) \approx 1.3 \times 10^{-4} 

137: / r_{12}$ for LS~5093, which is also always much smaller than the 

138: $\gamma\gamma$ absorption depth.

139: 

140: In this {\it Letter}, we provide a more detailed analysis of the expected 

141: $\gamma\gamma$ absorption trough caused by direct companion star light, 

142: including its temporal modulation due to the orbital motion. The 

143: model description and the derivation of the general

144: expression for the $\gamma\gamma$ opacity as a function of $\gamma$-ray

145: photon energy and orbital phase is given in \S \ref{model}.

146: Numerical results for LS~5039 are presented in \S~\ref{results}.

147: \S~\ref{summary} contains a brief summary and discussion of our results.

148: 

149: \section{\label{model}Model description and analysis}

150: 

151: \begin{figure}[t]

152: \includegraphics[width=14cm]{f1.eps}

153: %\plotone{geometry.eps}

154: \caption{Geometry of the model. The direction of the radio jets defines

155: the $x_3$ axis. The orbital plane of the binary system is the ($x_1$, $x_2$) plane, 

156: defined in such a way that line of sight towards the observer lies in the

157: ($x_2$, $x_3$) plane, where the azimuthal angle $\phi = 0$.}

158: \label{geometry}

159: \end{figure}

160: 

161: We choose a generic model set-up as illustrated

162: in Fig.~\ref{geometry} \citep[for a comparable model setup in the AGN

163: case see][]{bd95}. The orbital plane of the binary system defines the

164: ($x_1, x_2$) plane. The jet, assumed to be perpendicular to this plane, 

165: defines the
$x_3$ axis. The system is inclined with respect to our line 

166: of sight by
an inclination angle $i$. An azimuthal (phase) angle $\phi$ 

167: is defined such that $\phi = 0$ in the direction of the $x_2$ axis. The 

168: line of sight
lies in the $(x_2, x_3)$ plane. The $\gamma$-ray production 

169: site is
located at a height $z_0$ along the jet. The phase angle of the 

170: companion
star at the time of production of a $\gamma$-ray photon at $z_0$ 

171: is denoted $\phi_0$.
The distance the photon travels along the line of sight 

172: is denoted by $l$, whereas
a starlight photon travels a distance $x$ before 

173: interacting with the $\gamma$-ray
photon. The angle of incidence between 

174: the two photons is $\theta$, and $\mu \equiv \cos\theta$. The star was 

175: located at azimuthal angle $\phi_1$ at the
time when a photon, leaving 

176: the star at that time, interacts with a $\gamma$-ray
photon that has 

177: travelled a distance $l$ from $z_0$. This yields
$\phi_1 = \phi_0 + (2\pi / P)

178: \, (l - x) / c$, where $P$ is the orbital period.

179: 

180: With the definitions as shown in Fig.~\ref{geometry}, the distance x can

181: be calculated as $x = \vert {\bf x} \vert = \vert {\bf r} - {\bf s} \vert$,

182: where $s$ is the orbital separation of the binary system. We find

183: \begin{equation}

184: x^2 = s^2 + l^2 + z_0^2 + 2 \, l \, (z_0 \, \cos i - s \, \sin i \, \cos\phi_1).

185: \label{x2}

186: \end{equation}

187: From Fig.~\ref{geometry}, $\mu = {\bf l} \cdot {\bf x} / (l x)$, 

188: which yields

189: 

190: \begin{equation}

191: \mu = {l + z_0 \, \cos i - s \, \sin i \, \cos\phi_1 \over x}.

192: \label{mu}

193: \end{equation}

194: 

195: With these quantities, we can evaluate the $\gamma\gamma$ opacity of a

196: $\gamma$-ray photon with dimensionless energy $\epsilon_0 = E_0 / (m_e c^2)$

197: as

198: 

199: \begin{equation}

200: \tau_{\gamma\gamma} (\epsilon_0, z_0, \phi_0) = \int\limits_0^{\infty} dl \,

201: (1 - \mu) \int\limits_{2 \over \epsilon_0 \, (1 - \mu)}^{\infty} d\epsilon

202: \, \sigma_{\gamma\gamma} (\epsilon_0, \epsilon, \mu) \, n_{\rm ph}^{\ast}

203: (\epsilon, x)

204: \label{taugg}

205: \end{equation}

206: where
$\sigma_{\gamma\gamma}$ is the pair production cross section and

207: 

208: \begin{equation}

209: n_{\rm ph}^{\ast} (\epsilon, x) = {15 \over 4 \, \pi^5 \, m_e c^3} \,

210: {L_{\ast} \, \epsilon^2 \over \Theta_{\ast}^4 \, x^2 \, \left( e^{\epsilon

211: / \Theta_{\ast}} - 1 \right)}.

212: \label{nph}

213: \end{equation}

214: The stellar spectrum
is approximated by a blackbody with dimensionless 

215: temperature $\Theta_* = kT_*/m_ec^2$,
and the star is approximated as a 

216: point source.

217: 

218: \section{\label{results}Results for LS~5039}

219: 

220: \cite{casares05} have analyzed optical intermediate-dispersion

221: spectroscopic observations of LS~5039, which provided an estimate

222: of the mass of the compact object of $M_X = 3.7^{+1.3}_{-1.0} \, 

223: M_{\odot}$, indicating
that it is likely to be a black hole. Other

224: relevant parameters of the binary system are $P = 3.91$~d, the 

225: luminosity of the
O6.5V type stellar companion, $L_{\ast} = 10^{5.3} 

226: \, L_{\odot}$, with an
effective surface temperature of $T_{\rm eff} 

227: = 39,000 \, ^o$K, a viewing angle of $i = 25^o$, and an orbital 

228: separation of $s \approx 2.5 \times 10^{12}$~cm \citep{casares05}. 

229: Note that the center of mass is rather close to the massive 

230: ($M \approx 23 M_\odot$) stellar companion. However, because 

231: the light-travel and jet propagation times on the length scales investigated 

232: here ($\lesssim 10^{14}$~cm) are much shorter than the orbital period, the

233: detailed dynamics of the orbital motion are irrelevant for our purposes: 

234: at any given phase $\phi_0$, the geometry is basically stationary. An 

235: additional complication could be
introduced by the substantial eccentricity 

236: of the orbit, $e = 0.35$, with periastron at a phase angle of $\phi \sim 0.6 

237: \, \pi$
\citep{casares05}. However, in view of the still rather poor quality 

238: of the data currently available and expected in the near future, this may be 

239: regarded as a higher-order effect which may be introduced at a later stage.

240: 

241: The inset to Fig.~\ref{tauz} illustrates the shape of the absorption trough

242: caused by $\gamma\gamma$ absorption and its dependence on the orbital phase.

243: Here we assume that the intrinsic $\gamma$-ray spectrum is a 

244: power-law with photon index $\alpha_{\rm ph} = 2.5$, and $z_0 = 10^{12}$~cm.

245: The various curves illustrate the orbital modulation of the absorption trough,

246: with the lowest (most heavily absorbed) curve corresponding to $\phi_0 = 0$

247: and the highest (least absorbed) curve corresponding to $\phi_0 = \pi$. The

248: modulation is a combined consequence of two effects: for phase angles closer 

249: to $\pi$, (a) the average distance
of the star to any point on the line of 

250: sight is longer and (b) the angle of
incidence $\theta$ is smaller. In addition

251: to effect (a) causing the overall photon number density of the stellar photon 

252: field to decrease, effect (b) causes the threshold for $\gamma\gamma$ pair 

253: production
to increase as $\epsilon_{\rm thr} = 2 / (\epsilon_{\ast} \, [1 - \mu])$. 

254: This leads to a decreasing overall depth of the absorption trough, and a shift 

255: of the minimum of the absorption trough towards higher photon energies. 

256: 

257: \begin{figure}[t]

258: \includegraphics[height=14cm]{f2.eps}

259: %\plotone{tau_vs_z.eps}

260: \caption{$\gamma\gamma$ opacity at 250~GeV and 1~TeV as a function 

261: of the distance of the photon production region from the central 

262: compact object at phase $\phi_0 = 0$. The figure illustrates that 

263: (1) VHE photons produced
within a few $\times 10^{12}$~cm (i.e., 

264: of the order of the orbital
separation of the binary system) would 

265: be subject to substantial $\gamma\gamma$ absorption; (2) the minimum 

266: of the absorption trough
(maximum of $\tau_{\gamma\gamma}$ as a 

267: function of photon energy)
is shifting towards higher energies for 

268: larger distances from the
central source. 

269: Inset: Orbital modulation of the expected $\gamma\gamma$ absorption

270: trough, assuming a power-law spectrum with photon index

271: $\alpha_{\rm ph} = 2.5$ and a photon production site at $z_0 = 10^{12}$~cm.

272: The different curves represent the escaping photon spectrum at various

273: orbital phases, from $\phi_0 = 0$ (lowest curve) to $\phi_0 = \pi$ (highest

274: curve) in steps of $\pi/10$.}

275: \label{tauz}

276: \end{figure}

277: 

278: Fig.~\ref{tauz} shows the dependence of the absorption

279: feature on the location $z_0$ of the VHE $\gamma$-ray production site. The

280: $\gamma\gamma$ opacity is plotted for two photon energies, $E = 250$~GeV, and

281: $E = 1$~TeV at $\phi_0 = 0$. GeV -- TeV photons produced within $z_0 \sim s$ 

282: from the compact object
will be heavily attenuated for this phase angle. 

283: For photons produced at $z_0 \gg s$, $\gamma\gamma$ attenuation becomes

284: negligible. The crossing-over of the two curves in Fig.~\ref{tauz}

285: illustrates the shift of the minimum of the absorption trough to higher

286: photon energies with increasing distance from the central engine. This 

287: is a consequence of the decreasing incidence angle as discussed above. 

288: 

289: \begin{figure}[t]

290: \includegraphics[height=14cm]{f3.eps}

291: %\plotone{flux_tau_photon.eps}

292: \caption{Orbital modulation of the integrated photon number flux above 

293: energies $E_0 = 250$~GeV (solid) and $E_0 = 1$~TeV (short-dashed), 

294: the $\gamma\gamma$ opacity at $E = 250$~GeV (dotted) and $E = 1$~TeV

295: (dot-dashed), and the local photon spectral index $\alpha_{500}$ at 

296: 500~GeV (long-dashed). As in the inset to Fig. \ref{tauz}, an underlying 

297: power-law of photon index $\alpha_{\rm ph} = 2.5$ and a photon production 

298: site at $z_0 = 10^{12}$~cm has been assumed. A periodic flux modulation is 

299: expected to be accompanied by positive spectral-index/flux correlation 

300: (spectral softening as the flux increases) at $E_0 \gtrsim 300$~GeV; the

301: opposite behavior is found at $E_0 \lesssim 100$~GeV.}

302: \label{flux_tau}

303: \end{figure}

304: 

305: Fig.~\ref{flux_tau} quantifies the orbital modulation of the $\gamma\gamma$

306: absorption trough for the same parameters as used in the inset to 

307: Fig.~\ref{tauz}. The dotted and

308: dot-dashed curves show the orbital modulation of the optical depth, illustrating

309: that (a) it is maximized for $\phi_0 \approx 0, 2\pi$, and 

310: (b) that the mimimum of the absorption trough shifts towards higher photon 

311: energies for phases closer to $\pi$. The most feasible way to detect an orbital

312: modulation would be to measure a periodicity of the observed photon flux. The

313: solid and short-dashed curves in Fig.~\ref{flux_tau} show the integrated 

314: photon number flux above energies $E_0 = 250$~GeV and
$E_0 = 1$~TeV, respectively. 

315: This illustrates that the relative modulation is
most pronounced at several 

316: hundred GeV, noting that the $\gamma\gamma$ absorption is largest 

317: at photon energies
$ E_1 \cong 2 \cdot 2 (m_ec^2)^2 / (k T_* [1 - 

318: \cos (\arctan(s/z_0) + i)])$. In this estimate, we have made use of the fact 

319: that the photon energy $\epsilon_{pk}$ at the peak of the $\gamma\gamma$ 

320: attenuation cross section is $\epsilon_{pk}\approx 2 \epsilon_{thr}$. For 

321: the case of LS 5039 with $z_0 = 10^{12}$ cm and $\phi_0 = 0$, $E_1 \cong 300$~GeV,

322: in good agreement with the inset in Fig.~\ref{tauz}. The spectral softening 

323: with
increasing flux is

324: also illustrated by the local photon spectral index $\alpha_{500}$ at 500~GeV, 

325: plotted as the long-dashed curve in Fig.~\ref{flux_tau}. The flux dip around

326: $\phi_0 \approx 0$ is accompanied by a hardening of the VHE $\gamma$-ray 

327: spectrum.

328: 

329: Because the bulk speed of microquasar jets is 

330: typically only mildly relativistic ($\Gamma \sim 2$), in dramatic contrast to 

331: blazars,
one might also need to 

332: consider the possibility of VHE $\gamma$-rays emanating from the counter jet. 

333: Thus, we have done a study as described above, for an inclination angle of

334: $i = 155^o$. While the results for this case are qualitatively very similar

335: to the $i = 25^o$ case, the resulting $\gamma\gamma$ opacities are typically

336: larger by a factor of $\sim 2$ -- 3 for photon production sites at $z_0 \lesssim

337: 10^{12}$~cm, as expected because of the more favorable photon collision angle

338: and the line of sight passing by the star closer than for the approaching-jet

339: case. Because of the stronger absorption of $\gamma$ rays from the 

340: counter jet throughout the binary orbit,

341: and the reduction in the flux due to Doppler deboosting, the counter jet is

342: unlikely to enhance the $\gamma\gamma$ absorption signature

343: in the signal appreciably.

344: As for $i = 25^o$, $\gamma\gamma$ absorption effects become negligible for 

345: $z_0 \gg s$. 

346: 

347: \section{\label{summary}Summary and Discussion}

348: 

349: We have presented a detailed analysis of the effect of $\gamma\gamma$ absorption

350: of VHE $\gamma$-rays near the base of the jet of a microquasar by the intense

351: photon field of a high-mass stellar companion. We include the time-dependent,

352: periodic modulation of this effect due to the binary's orbital motion. We

353: applied our results to the specific case of LS~5039, which has recently been

354: identified as the counterpart of the VHE $\gamma$-ray source HESS~J1826-148.

355: Our results can be summarized as follows:

356: 

357: (1) VHE $\gamma$-rays produced closer to the central engine than

358: $z_0 \sim$~a~few~$\times 10^{12}$~cm, which is of the order of the

359: binary separation $s$, would be subject to very strong $\gamma\gamma$ 

360: absorption due to the stellar radiation field at orbital phases 

361: close to $\phi_0 = 0$.

362: 

363: (2) For VHE photon production sites at $z_0 \lesssim s$, the $\gamma\gamma$ 

364: opacity --- and, thus, the VHE $\gamma$-ray flux --- would be strongly

365: modulated on the orbital period of the binary system ($P = 3.91$~d in

366: the case of LS~5039). At orbital phases close to $\phi_0 = \pi$, the

367: intrinsic VHE $\gamma$-ray flux would still be virtually unabsorbed

368: even for $z_0 \sim 10^{12}$~cm. 

369: 

370: (3) The orbital modulation of the VHE $\gamma$-ray flux would be

371: characterized by a spectral hardening in the $\sim 300$~GeV -- 1~TeV

372: range during flux dips. At lower energies, the spectrum softens with

373: decreasing flux.

374: 

375: The HESS collaboration has reported no evidence for either flaring

376: or periodic variability from LS 5039 \citep{aharonian05}, 

377: although \citet{casares05} suggest that there is weak evidence 

378: for variability of the HESS emission with the orbital 

379: period. Periodic variability may also be indicated by X-ray 

380: observations \citep{bosch05b} of LS 5039. 

381: 

382: Besides $\gamma\gamma$ opacity effects, there are a few alternative

383: scenarios which might cause a periodic modulation of the $\gamma$-ray 

384: flux: 

385: 

386: (a) As a consequence of the substantial eccentricity of the orbit,

387: the rate of mass transfer from the stellar companion to the compact 

388: object, which is believed to be dominated by wind accretion, is likely 

389: to be periodically modulated. This modulation would also be expected to 

390: appear at radio and X-ray energies.

391: 

392: (b) Analogous to the phase-dependent modulation of the incidence angle 

393: for $\gamma\gamma$ absorption, this geometrical effect would also yield

394: a more favorable angle for Compton scattering of starlight photons into

395: the $\gamma$-ray regime at phases $\phi_0 \approx 0$.

396: 

397: (c) The orientation of the jet may also be mis-aligned with

398: respect to the normal of the orbital plane \citep{maccarone02,butt03}

399: and possibly precessing about the normal \citep{larwood98,torres05},

400: leading to additional modulations, including a changing Doppler boosting

401: factor.

402: 

403: Effect (a) would be expected to lead to an overall hardening
of the 

404: $\gamma$-ray spectrum at all energies with increasing $\gamma$-ray
flux, 

405: while (b) would lead to an overall softening throughout the GeV -- TeV

406: photon energy range because of the Klein-Nishina cutoff becoming noticeable

407: at lower observed photon energies with increasing flux. Both effects are

408: in contrast to the $\gamma\gamma$ absorption trough investigated in

409: this {\it Letter}. A stationary misalignment of the jet could lead to

410: a slight enhancement of the orbital modulation (if the jet makes a smaller

411: angle with the line of sight than the orbital-motion axis) or reduce it

412: (in the opposite case), but would not change our results qualitatively.

413: A $\gamma$-ray flux modulation due to jet precession can easily be 

414: disentangled from the orbital modulation since the precession period 

415: is generally different from the orbital period, so that its effect

416: would average out when folding observational data with the orbital period.

417: Consequently, a measurement of a non-thermal absorption trough at VHE 

418: $\gamma$-ray energies modulated with the orbital period would firmly 

419: establish the
importance of $\gamma\gamma$ absorption effects and thus 

420: place a robust
limit on the distance $z_0$ of the VHE $\gamma$-ray 

421: production site in LS 5039.

422: 

423: \acknowledgments

424: We thank Guillaume Dubus, Mathieu de Naurois, and Valenti Bosch-Ramon 

425: for helpful correspondence, and the referee for a helpful and constructive

426: report. This work was partially supported by NASA through XMM-Newton GO 

427: grant no. NNG~04GI50G, NASA INGEGRAL Theory grant no. NNG~05GK59G, and 

428: GLAST Science Investigation no. DPR-S-1563-Y. The work
of C.\ D.\ D.\ is 

429: supported by the Office of Naval Research.

430: 

431: \begin{thebibliography}

432: 

433: \bibitem[Aharonian et al.(2005)]{aharonian05}Aharonian, F., et al.,

434: 2005, Science, 309, 746

435: 

436: \bibitem[Aharonian \& Atoyan(1999)]{aa99}Aharonian, F., \& Aharonian, A., 1999,

437: MNRAS, 302, 253

438: 

439: \bibitem[Atoyan \& Dermer(2001)]{ad01}Atoyan, A., \& Dermer, C.~D., 2001, 

440: Physical Review Letters, 87, 221102 

441: 

442: \bibitem[Bednarek(1997)]{bednarek97}Bednarek, W., 1997, A\&A, 322, 523

443: 

444: \bibitem[Bednarek(2000)]{bednarek00}Bednarek, W., 2000, A\&A, 362, 646

445: 

446: \bibitem[B\"ottcher(2002)]{boettcher02}B\"ottcher, 2002, in ``The 

447: Gamma-Ray Universe", proc. of the XXII Moriond Astrophysics Meeting, 

448: Eds. A. Goldwurm, D. N. Neumann, \& J. T. T. V$\hat{\rm a}$n, p. 151 

449: 

450: \bibitem[B\"ottcher \& Dermer(1995)]{bd95} B\"ottcher, M., \& Dermer, C.~D.\ 1995, \aap, 302, 37 

451: 

452: \bibitem[Butt(2003)]{butt03}Butt, Y., M., Maccarone, T. J., \& Prantzos, N.,

453: 2003, ApJ, 587, 748

454: 

455: \bibitem[Bosch-Ramon \& Paredes(2004)]{bp04}Bosch-Ramon, V., \& Paredes,

456: J. M., 2004, A\&A, 417, 1075

457: 

458: \bibitem[Bosch-Ramon et al.(2005a)]{bosch05a}Bosch-Ramon, V., Romero, G. E., 

459: \& Paredes, J. M., 2005a, A\&A, 429, 267

460: 

461: \bibitem[Bosch-Ramon et al.(2005b)]{bosch05b} Bosch-Ramon, V., 

462: Paredes, J.~M., Rib{\' o}, M., Miller, J.~M., Reig, P., \& Mart{\'{\i}}, 

463: J.\ 2005b, \apj, 628, 388 

464: 

465: \bibitem[Casares et al.(2005)]{casares05}Casares, J., Rib\'o, M.,

466: Ribas, I., Paredes, J. M., Mart\'i, J., \& Herrero, A., 2005, MNRAS,

467: in press (astro-ph/0507549)

468: 

469: \bibitem[Gregory \& Taylor(1978)]{gregory78}Gregory, P. C., \& Taylor, A. R.,

470: 1978, Nature, 272, 704

471: 

472: \bibitem[Kniffen et al.(1997)]{kniffen97} Kniffen, D.~A., et al.\ 1997, ApJ, 486, 126 

473: 

474: \bibitem[Larwood(1998)]{larwood98}Laarwood, J., 1998, MNRAS, 299, L32

475: 

476: \bibitem[Maccarone(2002)]{maccarone02}Maccaroone, T. J., 2002, MNRAS, 336, 1371

477: 

478: \bibitem[Mannheim \& Biermann(1992)]{mb92}Mannheim, K., \& Biermann, P. L.,

479: 1992, A\&A, 253, L21

480: 

481: \bibitem[Mannheim(1993)]{mannheim93}Mannheim, K., 1993, A\&A, 221, 211

482: 

483: \bibitem[McSwain \& Gies(2002)]{mg02}McSwain, M. V., \& Gies, D. R., 2002, 

484: ApJ, 568, L27

485: 

486: \bibitem[Mirabel \& Rodr\'\i guez(1994)]{mirabel94}Mirabel, I. F., \&

487: Rodr\'\i guez, L. F., 1994, Nature, 371, 46

488: 

489: \bibitem[Moskalenko et al.(1993)]{moskalenko93}Moskalenko, I. V., Karakula, S.,

490: \& Tkaczyk, W., 1993, MNRAS, 260, 681

491: 

492: \bibitem[Moskalenko \& Karakula(1994)]{moskalenko94}Moskalenko, I. V., \&

493: Karakula, S., 1994, ApJS, 92, 567

494: 

495: \bibitem[M\"ucke et al.(2003)]{muecke03}M\"ucke, A., et al., 2003,

496: Astropart. Phys., 18, 593

497: 

498: \bibitem[Paredes et al.(2000)]{paredes00}Paredes, J. M., Mart\'\i, J., Rib\'o, M.,

499: \& Massi, M., 2000, Science, 288, 2340

500: 

501: \bibitem[Romero et al.(2003)]{romero03}Romero, G. E., Torres, D. F., Kaufman Bernad\'o,

502: M. M., \& Mirabel, I. F., 2003, A\&A, 410, L1

503: 

504: \bibitem[Taylor et al.(1992)]{taylor92}Taylor, A. R., Kenny, H. T. Spencer, R. E.,

505: \& Tzioumis, A., 1992, ApJ, 395, 268

506: 

507: \bibitem[Torres et al.(2005)]{torres05}Torres, D. F., Romero, G. E., Barcons, X.,

508: \& Lu, Y., 2005, ApJ, 626, 1015

509: 

510: %\bibitem[Weekes(2002)]{weekes02}Weekes, T. C., 2002, Astropart. Phys.,
17, 221

511: 

512: \end{thebibliography}

513: 

514: \end{document}

515: