1: \documentclass[10pt]{article}
2: \usepackage{amsmath, amssymb, epsfig, multicol, indentfirst}
3: \setlength{\oddsidemargin}{0in}
4: \setlength{\textwidth}{6.5in}
5: \setlength{\textheight}{8.5in}
6: \setlength{\headheight}{0in}
7: \setlength{\headsep}{0in}
8: \begin{document}
9: \begin{center}
10: \Large
11: Exploring the Gauge Hierarchy Problem at $N = 2$ with TeV$^{-1}$-sized
12: Black Holes\\[0.25in]
13:
14: % Identifying information (Added on 21-jul-04)
15: \normalsize
16: Jason Minamora$^{1, 2}$\\[1em]
17: {\footnotesize $^{1}$\emph{California Institute of Technology, Department
18: of Physics, Mailcode 356-48, 1200 E. California Blvd., Pasadena, CA 91125}\\
19: $^{2}$ \emph{Institute of Geophysics and Planetary Physics, University of
20: California, Los Angeles, Box 951567, Los Angeles, CA 90095}}
21: \end{center}
22: \small
23: \begin{abstract}
24: \noindent A mass Hierarchy of magnitude
25: $O(10^{16})$ GeV separates the gravitational and electroweak interactions.
26: Traditional proposed resolutions of this anomaly have included
27: supersymmetric theories (most notably, the MSSM and ESSM) and string
28: theory. However, energiers at which supersymmetric particles are expected to
29: appear are not accessible to experiment, and string theory is only
30: testable at $\sim \! M_{\text{pl}}$. A novel idea involves introducing
31: extra dimensions into a Minkowski space-time to reduce the effective
32: separation between the gauge and gravitational couplings. Theoretical
33: developments of the so-called Arkani-Dinopolous-Dvali (ADD)
34: extra-dimensional scenario, and its experimental verification, are presented
35: here.
36: \end{abstract}
37: \normalsize
38: %
39: %%\begin{multicols}{2}
40: %
41: \section{Background}
42:
43: The gap between the strengths of the electromagnetic and gravitational interactions places constraints on the applicability of the Standard Model (SM).
44: The cutoff scale between the gauge ($M_{\text{EW}} = 10^{3}$ GeV) and gravitational forces ($M_{\text{g}} \approx M_{\text{Pl}} = 10^{19}$ GeV) is
45: large~\cite{one} ($\Lambda \sim 10^{16}$), suggesting a necessary
46: extension to the SM.
47:
48: Supersymmetric extensions to the SM have been proposed to resolve the gauge hierarchy. The Minimal Supersymmetric Standard Model (MSSM)---see,
49: for example,~\cite{two}
50: ---assigns fermion superpartners to gauge bosons (including the unseen Higgs), and the spin-$\frac{3}{2}$ gravitino to the graviton. These particles, however, are yet to be seen; see Table 1, and also ref.~\cite{three} for discussion.
51:
52: \begin{table}[h]
53: \begin{center}
54: \begin{tabular}{l l l}
55: \hline\hline
56: SM Particle(s)&MSSM Counterpart(s)&UL(MSSM) [GeV] (CL = 95\%)\\\hline
57: $\gamma, Z^{0}, H^{0}$&$\overset{\sim}{\chi}^{0}_{i}$ (neutralinos) &66.4\\
58: $W^{\pm}, H^{\pm}$&$\overset{\sim}{\chi}^{\pm}_{i}$ (charginos) &67.7\\
59: $e$&$\overset{\sim}{e}$ &95.0\\
60: $g$&$\overset{\sim}{\chi}_{g}$ (gravitino) &(no data)\\
61: \hline\hline\\
62: \end{tabular}
63: \caption{Upper limits on MSSM particle searches. Note that all values are
64: within experimental reach. Values taken from ref.~\cite{four}.}
65: \end{center}
66: \end{table}
67: An alternative approach to the problem (discussed here) is to introduce extra spatial dimensions, each of radius $R$. In this scheme, the Planck scale is no longer fundamental.
68: Instead, a test mass $m$ will experience a potential
69: \begin{equation}
70: V(r) = \frac{m}{M^{N+2}_{*}R^{N}}\frac{1}{r} \quad (r \gg R)\label{E:N} \,
71: \end{equation}
72: a distance $r$ from the source. $M_{\text{pl}}$ is now replaced with the
73: effective ($4+N$)-dimensional Planck mass~\cite{five}:
74: \[
75: M^{2}_{\text{Pl}} = M^{2+N}_{*}R^{N} .
76: \]
77: If we choose $M_{*} \approx M_{\text{EW}}$,
78: then gravity is of the same order as the gauge forces; substituting in
79: figures gives
80: \[
81: R \approx 10^{\frac{30}{n} - 17} \times \left(\frac{1 \text{ TeV}}{m_{\text{EW}}}
82: \right)^{\frac{2}{n} + 1}.
83: \]
84: The distance scale for $N = 1$ is too large---$R$ = 1 AU. The $N = 2$ case,
85: which gives TeV$^{-1}$ (millimeter-) sized $R$, is empirically
86: accessible~\cite{six}.
87:
88: One of the more interesting implications from the $N = 2$ theory
89: is production (and decay) of millimeter-sized black holes (BHs) at
90: experiments~\cite{seven, eight}. The simplest scenario arises when one considers
91: Hawking-type BHs. From a classical, two-dimensional
92: viewpoint~\cite{nine}, Hawking black holes decay forever. In (4 + 2) dimensions,
93: the BH lifetime can be taken to
94: be instantaneous. This is convenient, since one can produce
95: and compare the
96: decay spectrum of
97: $N = 2$ BHs against the (higher-dimensional) Hawking spectrum in a short time.
98: If BH branching fractions agree with those predicted by the Hawking model,
99: a new upper limit could be placed on the quantum gravity scale.
100:
101: \section{Phenomenological Considerations}
102:
103: The simplest description of a Hawking BH is the semiclassical one, valid for
104: masses $M_{\text{BH}} \gg M_{*}$.\footnote{Here, $M_{*}$ is the $N = 2$
105: effective Planck mass.}
106: The semiclassical approximation fails, however, when $M_{\text{BH}} \sim M_{*}$, since stringy corrections are no longer negligible. For our purposes, though, we will restrict discussion to the former case, where the spectroscopy of BH decays is well-described~\cite{ten}.
107:
108: \subsubsection*{Production and Decay of TeV$^{-1}$ -sized Black Holes}
109: %%
110: %% Figure 1 goes in this section
111: %%
112:
113: If BHs are produced with masses much larger than $M_{*}$, the
114: Schwarzschild radius $R_{s}$ can be found using classical
115: considerations~\cite{eleven}. For the six-dimensional case,
116: \begin{equation}
117: R_{s} = \pi^{-1/3}\left(\frac{3 M_{\text{bh}}}{2 M_{*}}\right)^{\frac{1}{3}}
118: \frac{1}{M_{*}} \, .\label{E:sch}
119: \end{equation}
120: In particular, a reduced mass $M_{*}$ of 1 TeV places an upper limit on
121: the size of these objects in the millimeter range, for production
122: energies of up to $\sim 10$ TeV. These energies are presently above the thresholds of experiment, but the mechanism for production is still valid.
123:
124: %%%%%%%%%%%Fig. 1
125: \begin{figure}[t]
126: \begin{center}
127: \begin{tabular}{l r}
128: \epsfig{figure = ./figures/bhform_4d.eps,
129: width = 3.00in, height = 1.0in}
130: &
131: \epsfig{figure = ./figures/shock.eps,
132: width = 1.75in, height = 2.25in}
133: \end{tabular}
134: \small
135: \caption{Parton scattering. (\emph{1A---Left}) Close-range interactions
136: between two partons $i$ and $j$ produce a vortex of size $\sim r_{\text{S}}
137: (\hat{s} = E_{\text{cm}})$. (\emph{1B---Right}) The
138: topological profile of space changes abruptly near the scattering region.
139: Regions (I), (II) and (III) fall outside the area cartooned in Fig. 1A; space
140: there is flat ($g_{\mu \nu}$ = $\eta_{\mu \nu}$). In the production region
141: (IV), the metric may be of Schwarzschild form, with a singularity at the
142: center.}
143: % Be sure to CITE 0212342
144: \end{center}
145: \end{figure}
146:
147: Consider the parton-parton\footnote{The six quarks and gluons.}
148: scattering process shown in Fig. 1. Such reactions may take place at
149: the LHC, and in collisions between neutrinos and baryons. If the impact
150: parameter $R_{s}(\hat{s})$ is smaller than $R_{s}$,
151: we expect a vortex---here, a black hole---to form~\cite{eight}.
152:
153: %%%%%%%% Fig. 1
154:
155: In our approximation, millimeter-sized black holes decay via the Hawking
156: spectrum~\cite{twelve},
157: \begin{equation}
158: \frac{dN_{i}}{dt} \sim \frac{\Gamma_{s}(E, s)}{e^{E/T} - (-1)^{2s}}
159: \frac{d^{5}k}{(2\pi)^{5}}.\label{E:hawking}
160: \end{equation}
161: The \emph{greybody factor} $\Gamma_{s}$ is dependent on both the spin $s$, and the energy $E$; $\Gamma_{s}$, then, is effectively an absorption cross-section for particle species $i$. Because the Hawking distribution law~\eqref{E:hawking} contains a term in the (normalized) five-momentum, $d^{5} k$, we see that energy is radiated faster from a
162: TeV$^{-1}$-sized black hole, than from a classical Schwarzschild black hole,
163: because of the increased phase-space volume available to the decaying
164: particle.
165: In fact, lifetimes for the six-dimensional case are approximately 1 fs.
166:
167: Expression~\eqref{E:hawking} bears resemblance to the Planck
168: distribution function for a classical blackbody radiator:
169: \begin{equation}
170: \frac{dN(\lambda, T)}{dE} \sim \frac{1}{(\lambda T)^{2}}\frac{1}{e^{1/\lambda kT}
171: \pm 1} \, .\label{E:class}
172: \end{equation}
173: The prefactor in $\lambda^{-2}$ indicates that~\eqref{E:class} is valid in the three-dimensional case (if we identify the de Broglie momentum $p = 1/\lambda$). Classical BHs are, therefore, longer-lived than those in six-dimensional space.
174: Consequently, the momentum factor in~\eqref{E:hawking}
175: suggests that millimeter-sized black holes decay into
176: dominantly soft components (such as jets) compared
177: to the macroscopic case~\eqref{E:class}.
178:
179: One major obstacle is controlling Standard Model (SM) background.
180: Because the interaction between partons is mediated by either a $\gamma, W^{\pm}$, or $Z^{\circ}$, there may be a contribution from the mode $Z^{\circ}
181: \rightarrow ee$ to the leptonic component of the spectrum.
182:
183: \subsubsection*{Simulation of Black Hole Events}
184:
185: The scattering process outlined in Fig. 1 is the
186: primary means for producing black holes at the LHC and
187: elsewhere~\cite{seven, eight, twelve}; software to simulate the interaction has already been
188: constructed~\cite{fourteen}. Current simulation techniques model parton evolution from $pp$ and $p\overline{p}$ processes---those most often studied at the LHC.
189:
190: The software used to predict the decay products of
191: black holes is not subject to the two major
192: constraints of the theory outlined above; modern BH event generators
193: include analyses of time-varying behavior---that is, temperature and spectroscopy---and attempts to describe the terminal state of BH decay. One shortcoming of the semiclassical approximation is a steady-state model that assumes equal branching fractions for all of the 60 particles in the SM~\cite{five}; that
194: assumption is relaxed in a stochastic model, since the greybody
195: factor $\Gamma_{s}$ is a free parameter in the simulation. Table 2 shows results from a Monte-Carlo run using the CHARYBDIS
196: analysis package~\cite{thirteen}.
197:
198: \begin{table}
199: \begin{center}
200: \begin{tabular}{|c | c c c|}
201: \hline
202: & &Energy [GeV] &\\
203: Species&Est. Mean ($\pm 50$ GeV)&&Est. 90\% UL ($\pm 50$ GeV)\\\hline\hline
204: $H^{0}$&500&&1250\\
205: $e^{\pm}$&500&&1250\\
206: $\gamma$&650&&1200\\\hline
207: \end{tabular}
208: \caption{Mean values and 90\% upper limits on energy
209: for simulated decay spectra.
210: Values taken from~\cite{thirteen}.}
211: \end{center}
212: \end{table}
213:
214: \section{Experimental Foundations (Proposed)}
215:
216: A host of ground-based and fixed-target experiments plan to begin
217: searches for black holes once TeV physics is empirically realized.
218: Among these are the
219: LHC~\cite{seven, eight} (which will include the ATLAS
220: detector~\cite{seventeen}), and
221: AMANDA/ICECUBE~\cite{fourteen, fifteen, sixteen}.
222:
223: \subsubsection*{Partonic scattering at the LHC}
224:
225: Proton-anti-proton scattering with center-of-mass energy $E_{\text{cm}}$
226: greater than
227: the $N = 2$ BH production threshold $M_{\text{BH}}$ can be done with
228: current experimental thresholds.
229: experiment. If BHs are produced, they will be produced copiously at the
230: LHC.
231:
232: Collider experiments have already established empirical lower limits
233: on center-of-mass production energies; the most recent summary reports
234: a 1.3-TeV threshold for the $N = 2$ case. This is below the phenomenological
235: limit derived from the higher-dimensional Newton's constant, however
236: ($E_{N = 2} \gtrsim 1.6$ TeV), and so searches in higher-energy regimes
237: still look promising~\cite{eighteen}.
238:
239: The LHC collider beams are tentatively scheduled to operate at
240: 7.0 TeV, with $\sim 1 \times 10^{14}$ collisions per second. The BH
241: production rate, $\Gamma_{\text{BH}}$, depends
242: on $\sqrt{\hat{s}_{\text{beam}}}$,
243: the collision rate, and a composite efficiency
244: $\epsilon = \epsilon_{\text{b}} \times \epsilon_{\text{l}}$, based on
245: beam injection parameters and predicted losses during operation.
246: Estimates using simulated protons at injection and during procession around
247: the LHC storage ring give
248: efficiencies between $10^{-5}$ and $10^{-4}$~\cite{nineteen};
249: we therefore expect
250: $\Gamma_{\text{BH}} (p \overline{p}
251: \rightarrow BH) \lesssim 10^{10}$ s$^{-1}$; simulated
252: BH production cross-sections give similar predictions; see Fig. 2A.
253:
254:
255: %%%%%%%%%%%Fig. 2
256: \begin{figure}[t]
257: \begin{center}
258: \begin{tabular}{l r}
259: \epsfig{figure =./figures/fig1.eps,
260: width = 2in, height = 2in}
261: &
262: \epsfig{figure = ./figures/fig2.eps,
263: width = 2in, height = 2in}
264: \end{tabular}
265: \small
266: \caption{(\emph{2A---Left}) Scale-dependent parton cross-sections
267: [subpanel (a)]
268: and simulated multiplicity data [subpanel (d)] LHC collisions
269: ($\sigma_{\text{incl.}} = 1$ mb) should indeed produce BHs at a rate
270: $\Gamma_{\text{BH}}$ if they are seen, and are found to decay
271: via the Hawking mechanism.
272: (\emph{2B---Right}) Simulated number of BH events per unit reduced Planck mass
273: , for four different values of $M_{\text{BH}}$. Shaded regions give ranges for
274: $2 \leq N \leq 7$. Dashed and dotted lines are expected contributions
275: from the background modes
276: $\Gamma (Z^{\circ} \rightarrow \ell^{\pm} \nu) + \Gamma (
277: Z^{\circ} \rightarrow \gamma + X)$, and $\Gamma (
278: Z^{\circ} \rightarrow \gamma + X)$ only, respectively~\cite{nine}.}
279: % Be sure to CITE 'will we see BHs at the LHC?'
280: \end{center}
281: \end{figure}
282: \normalsize
283: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
284: ATLAS, which uses the $p\overline{p}$ beam, plans to take its
285: first science run in 2007. The proposed detector will be
286: sensitive to leptonic and electromagnetic decays from hard scattering events;
287: efficient detection of $\ell$ and $\gamma$ tracks
288: is important for reconstructing BH decay products, since the
289: observable portion of the Hawking spectrum is dominated by the intermediate
290: vector bosons and charged leptons---see Table 4.
291: One must also be clever in separating SM background events due to the decay
292: of the $Z^{\circ}$, since products from leptonic decay channels may
293: appear as BH signal. This background can be significant, as shown in
294: Fig. 2B. Obtaining accurate counts from electromagnetic processes
295: also presents difficulty, because of large photon multiplicities
296: from pseudoscalar meson decays---some examples of more exotic decays with
297: significant electromagnetic contributions are shown in Table 3.
298: Placing kinematic cuts on mass
299: differences between particles and decay products is a straightforward
300: way to crudely correct for SM events.
301:
302: \begin{table}[h]
303: \begin{center}
304: \begin{tabular}{|l||l|l|}
305: \hline
306: \multicolumn{3}{|c|}{Selected $p \overline{p}$ Scattering Modes}\\
307: \hline\hline &Primary Decay &Final State\\
308: \hline\hline
309: $p \overline{p} \rightarrow 3\pi^{0}$ &$3 \pi^{0}$ &$6\gamma$\\\hline
310:
311: &$\pi^{0}\pi^{0}\eta^{\prime}$ &$10\gamma, 6\gamma$\\\cline{2-3}
312: $p \overline{p} \rightarrow 2\pi^{0} + X$
313: &$\pi^{0}\pi^{0}\omega$ &$7\gamma$\\\cline{2-3}
314: &$\pi^{+}\pi^{-}\eta^{\prime}$
315: &$\pi^{+}\pi^{-1} 6\gamma$ (Dominant)\\\hline
316: &$\pi^{0}\omega \eta$ &$7\gamma$\\\cline{2-3}n
317: $p \overline{p} \rightarrow \pi^{0} + X$
318: &$\pi^{0}\eta\eta$ &$6\gamma$\\\cline{2-3}
319: &$\pi^{0}\eta\eta^{\prime}$ &$6\gamma$\\\hline
320: \end{tabular}
321: \caption{Electromagnetic $p \overline{p}$ channels with potential,
322: non-trivial contributions to the $\gamma$ component of the BH
323: Hawking decay spectrum. Many such secondary decays are expected to be
324: seen at the LHC; see ref.~\cite{twenty} for a full account.}
325: \end{center}
326: \end{table}
327:
328: \subsubsection*{Cosmic-Ray Production of BHs}
329:
330: Ultra-high-energy photons from extragalactic sources
331: produce neutrinos in matter, which are readily available for
332: detection by ground-based detectors.
333: The parton scattering model discussed earlier can similarly be used
334: here to discuss BH production in extensive air showers. Photons that initiate
335: electromagnetic cascades are typically in the range $10^{15}$--$10^{21}$ eV.
336: For reasons to be discussed below, ground-based detectors have the
337: highest likelihood of seeing BH events, because of the low background.
338: Neutrino arrays are sensitive
339: to the highest-energy portions of the cascade spectrum; BH formation should
340: thus take place simultaneously with presently observed neutrino-matter
341: interactions. If BHs are to be produced via parton interactions, then
342: the reaction $\nu N \rightarrow BH + \mu^{\pm}\ell^{\mp}X$
343: warrants consideration. Expected BH event rates (based on the ADD
344: scenario~\cite{eleven}) for the South Pole-based ICECUBE neutrino array
345: have been calculated based on cosmic neutrino fluxes~\cite{seven, eleven}:
346: \begin{align}
347: \Gamma_{\text{BH}} = \int N(E) \frac{d \sigma}{d \Omega} d \Omega &=
348: n_{p} \times \int_{M_{*}}^{E^{\nu}_{\text{max}}}
349: \frac{d\sigma}{dE} \Phi(E) \, dE\notag\\
350: &= (3.5 \times 10^{-35}) N \quad \text{(events/yr)} \, .\label{E:tot}
351: \end{align}
352:
353: Here, $\sigma = \sigma(E)$ is the ADD-estimated cross-section for
354: particle fluxes $\Phi(E)$. For $N = 10^{39}$, the number of
355: protons in 1 km$^{3}$ of ice, $\sim$100 BHs per day
356: should be seen from neutrino-nucleon scattering at ICECUBE.\footnote{The rate
357: calculated from~\eqref{E:tot} is for a six-dimensional BH (the $N = 2$ case).}
358:
359: Cosmic-ray experiments
360: offer a distinct advantage over beam-beam collision facilities like the LHC:
361: the freedom to choose between different active sources at various
362: distances from Earth makes background correction a very simple task.
363: Figure 3 shows the signal-to-noise ratio plotted against kinematic and
364: spatial parameters. Depending on the intensity of the source, background
365: begins to dominate at some critical distance $d_{c}$ (Fig. 3A). For objects
366: at distances $d < d_{c}$, the effective area of air showers fully
367: covers the field of view of the detector; background is no longer
368: significant, allowing easy reconstruction of decay matter from BHs.
369: Fig. 3B takes the form of a ``rate-versus-threshold'' curve: signal is
370: ultimately detector-limited, and is seen to level off at
371: Lorentz factors $\gamma \sim 450$.
372: However, the shape of such a curve does not yield much scientific
373: information, since it is the choice of threshold energy that determines
374: photomultiplier tube response to cascades.
375: The spectral profile of gamma-ray bursts (GRBs)
376: can be directly extracted from Fig. 3C; low levels of signal from far away
377: sources in higher-energy regions ($\alpha \gtrsim 2.4$) can be attributed
378: to attenuation effects; ICECUBE is sensitive to the hard component
379: of the spectrum for GRBs that are relatively nearby (dotted line in Fig. 3C).
380: Using ICECUBE as a nominal example of large-scale ground-based detectors,
381: we find that such detectors' ability to observe sources at a continuum
382: of distance scales, and their freedom to discriminate at different
383: thresholds, results in background minimization superior to fixed-target and
384: beam-beam experiments.
385: So far, black holes have not yet been observed in the mode
386: $\nu N \rightarrow \text{BH} + X$; upper
387: limits are given elsewhere~\cite{twentyone}.
388: Should one be produced, scintillations from heavier decay products
389: ($W^{\pm}\text{s}, \ell$) will be seen in the ice, even though
390: semi-classical arguments include
391: all SM particles in the BH decay spectrum. Unlike the LHC,
392: ground-based detection enjoys a high signal-to-noise ratio, and
393: should therefore
394: be sensitive
395: to not only the quarks and leptons (see Table 4), but also
396: the more exotic SM particles. It is likely that the first BH event will
397: be seen in cosmic-ray apparatuses.
398: %%%%%%Table ...
399: \begin{table}[h]
400: \begin{center}
401: \begin{tabular}{l c c c c c c c}
402: &$q$ &$\ell$ &$g$ &$W^{\pm}, Z$ &$\gamma$
403: &$G^{0}, G^{\pm}$ &$H$\\
404:
405: \hline
406: Relative Contribution(\%) &52.94 &17.65 &17.65 &6.62 &2.21 &2.21
407: &0.74\\
408: \end{tabular}
409: \caption{Relative contributions of SM constituents to the Hawking
410: Decay Spectrum. Values derived from multiplicity data given in~\cite{twelve}.}
411: \end{center}
412: \end{table}
413:
414:
415:
416:
417: \section{Summary and Conclusion}
418:
419: Higher-dimensional theories can be used to bridge the gap
420: between the electroweak (gauge) interactions and gravity. We have considered
421: one such application that introduces extra TeV$^{-1}$ (mm) -sized dimensions
422: into space, which in turn lowers the fundamental Planck scale $M_{*}$ to
423: within experimental bounds. Observing Black holes in $N = 2$ extra dimensions
424: would provide tangible evidence of quantum gravity at the TeV scale.
425: Parton-parton scattering in $p \overline{p}$ collisions, and in
426: $\nu N$ scattering events, are two proposed production methods.
427: If $N = 2$ black holes exist, one should see their copious production and
428: decay at the LHC, and at a lower rate in neutrino arrays such as
429: AMANDA/ICECUBE. Because of differences in resolution and background levels,
430: the LHC and neutrino telescopes have their own advantages---high event rates
431: are usually sacrificed for clean signal. As of this writing, no black
432: holes have been seen in neutrino arrays; studies at the LHC have yet to be
433: done.
434:
435: %%%%%%%%%%%Fig. 3
436: \begin{figure}[t]
437: \begin{center}
438: \begin{tabular}{l c r}
439: \epsfig{figure =./figures/sn_vs_z.eps,
440: width = 2in, height = 2in}
441: &
442: \epsfig{figure = ./figures/sn_vs_lorentz.eps,
443: width = 2in, height = 2in}
444: &
445: \epsfig{figure = ./figures/sn_vs_alphap.eps,
446: width = 2in, height = 2in}
447: \end{tabular}
448: \small
449: \caption{Signal-to-noise ratio for ICECUBE-type detectors. Three regimes
450: are shown, for three different values of the luminosity---dependence
451: on distance (\emph{3A, left}), outflow velocity (\emph{3B, center}), and
452: spectral index (\emph{3C, right})~\cite{sixteen}. These plots correspond to a
453: 250 GeV threshold; normal incidence is assumed.}
454: \end{center}
455: \end{figure}
456: \bibliography{bh}
457: \bibliographystyle{unsrt}
458: \end{document}
459:
460: