1: %\documentclass[twocolumn,epsf,showpacs,prd]{revtex4}
2: \documentclass[twocolumn,showpacs,prd,floatfix]{revtex4}
3: %\documentclass[12pt]{article}
4:
5: % Comment this out unless used for xdvi
6: %\usepackage[active]{srcltx}
7:
8:
9: \usepackage{graphicx}% Include figure files
10: \usepackage{dcolumn}% Align table columns on decimal point
11: \usepackage{bm}% bold math
12: \usepackage{amsmath}
13: \usepackage{epsf}
14:
15: \begin{document}
16:
17: \newcommand{\beq}{\begin{equation}}
18: \newcommand{\eeq}{\end{equation}}
19: \newcommand{\bea}{\begin{eqnarray}}
20: \newcommand{\eea}{\end{eqnarray}}
21: \newcommand{\lsim}{\stackrel{<}{\scriptstyle \sim}}
22: \newcommand{\gsim}{\stackrel{>}{\scriptstyle \sim}}
23: %
24: \def\bone{$B^{(1)}$}
25: \def\bone{B^{(1)}}
26: %\def\erone{e^{(1)}_R}
27: \def\etal{{\it et al.~}}
28: \def\eg{{\it e.g.~}}
29: \def\ie{{\it i.e.~}}
30: \def\DM{dark matter~}
31: \def\DE{dark energy~}
32: \def\GC{Galactic center~}
33: \def\susy{SUSY~}
34:
35: \title{Effects of atmospheric electric fields on detection of
36: ultrahigh-energy cosmic rays}
37:
38: \author{Alexander Kusenko and Dmitry Semikoz
39: }
40: \affiliation{Department of Physics and Astronomy,
41: UCLA, Los Angeles, CA 90095-1547, USA}
42:
43: \vspace{0.5truecm}
44: \begin{abstract}
45: We show that atmospheric electric fields may affect the cosmic ray
46: observations in several ways and may lead to an overestimation of the
47: cosmic ray energies. The electric field in thunderclouds can be as high as
48: a few kV/cm. This field can accelerate the shower electrons and can feed
49: some additional energy into the shower. Therefore, ground array
50: observations in certain weather conditions may overestimate the energy of
51: ultrahigh-energy cosmic rays if they don't take this effect into account.
52: In addition, the electric field can bend the muon trajectories and affect
53: the direction and energy reconstruction of inclined showers. Finally,
54: there is a possibility of an avalanche multiplication of the shower
55: electrons due to a runaway breakdown, which may lead to a significant
56: miscalculation of the cosmic ray energy.
57:
58: \end{abstract}
59:
60: \pacs{96.40.-z, 95.55.Vj, 95.85.Ry, 98.70.Sa
61: \hfill
62: UCLA/04/TEP/44}
63: \maketitle
64:
65: \vspace{1truecm}
66:
67: \section{Introduction}
68:
69: Ultrahigh-energy cosmic rays (UHECR) produce extensive air showers in the
70: atmosphere, which are observed by two different techniques: (1) optical
71: detection of fluorescent light, and (2) surface array detection of charged
72: particles in the shower. The first technique, employed, for example, by
73: HiRes detector \cite{hires}, can be used exclusively in good weather, while
74: the second technique, used, for example, by the AGASA
75: experiment\cite{AGASA}, is usually believed to afford accurate observations
76: regardless of the atmospheric conditions. The new Pierre Auger experiment
77: is using a combination of the two techniques \cite{AUGER}.
78:
79:
80: On a clear day the atmosphere is permeated by the electric fields of order
81: a few volts per cm~\cite{intro}. However, in thunderclouds the electric
82: field reaches much higher magnitudes, up to a few kV/cm~\cite{el_book}.
83: Balloon measurements~\cite{Marshall} show the electric fields ${\cal E}>
84: 1$~kV/cm, at altitudes 0-12 km, which may switch polarity at several
85: altitudes. A shower developing in the atmosphere and going through a
86: thundercloud may, therefore, pass between layers with either polarity at
87: some angle. The electric field along the path of a random shower may have
88: a magnitude of a few kV/cm and an arbitrary direction, or alternating
89: direction. As we discuss below, the electron energy gain in the electric
90: field of a few kV/cm can be comparable to the energy losses due to
91: ionization. It, therefore, important to examine what effect this may have
92: on detection of UHECR.
93:
94: Several experimental studies confirm the influence of electric fields on
95: secondary particles in a shower. First, observations of low-energy cosmic
96: ray electrons show short-term variations during thunderstorms
97: \cite{thunder_low_energy}. A similar effect has been observed for muons
98: with a higher energy,
99: $E>100$~MeV~\cite{thunder_low_energy,thunder_low_energy2}, but it is not
100: nearly as strong as the effect on the low-energy electrons. Second, it has
101: been established that giant electron-gamma bursts are triggered in
102: thunderclouds by the passage of an extensive air shower (EAS) from a
103: $10^{16}$ cosmic ray \cite{electr_thunder}. EAS were measured at the Tien
104: Shan Mountain in coincidence with detection of a radio signal from the giant
105: electron-gamma bursts. Theory of these bursts, based on the phenomenon of
106: runaway breakdown, has been developed in recent years~\cite{uspekhi}. It
107: has was also suggested that
108: acceleration of secondary electrons from EAS in thunderclouds can be used
109: for radio detection of cosmic rays with energies
110: $E>10^{17-19}$~eV~\cite{radio_high_energy}.
111:
112:
113: We will show that the atmospheric electricity associated with thunderclouds
114: may affect the ground array observations of $E>10^{19-20}$~eV cosmic rays
115: in several ways. First, acceleration of the shower electrons can alter the
116: low-energy spectrum and the number of particles in EAS at ground level.
117: This could lead to errors in energy determination by a ground array.
118: Second, the muon trajectories may be deflected by the atmospheric electric
119: fields, which could lead to errors in reconstruction of energy and
120: direction of inclined showers. Finally, a runaway breakdown and
121: discharge triggered by an UHECR EAS may produce a dramatic increase in the
122: number of electrons detected by the surface array.
123:
124:
125: \section{Energy of the electromagnetic component}
126:
127:
128:
129: \begin{figure}[ht]
130: \centerline{\epsfxsize=3.5in\epsfbox{proton_R.ps}}
131: \centerline{\epsfxsize=3.5in\epsfbox{iron_R.ps}}
132: \caption{ Effect of a uniform atmospheric electric field on the shower
133: profile for a proton primary (upper plot) and an iron primary (lower plot).
134: The dotted lines show AIRES simulations of the $10^{20}$~eV
135: (low line) and $1.2\times10^{20}$~eV (upper line) showers, respectively, in
136: the absence of atmospheric electric field. The thick and thin solid lines
137: correspond to a $10^{20}$~eV shower developing in the presence of a uniform
138: electric field $\pm 1$~kV/cm. The electric field makes the $10^{20}$~eV
139: shower look more like a $(1.1 - 1.2)\times10^{20}$~eV shower. }
140: \label{figure:R}
141: \end{figure}
142:
143:
144:
145:
146: \begin{figure}[ht]
147: \centerline{\epsfxsize=3.5in\epsfbox{proton_E.ps}}
148: \centerline{\epsfxsize=3.5in\epsfbox{iron_E.ps}}
149: \caption{ Effect of atmospheric electric field on the photon and electron
150: spectra. The the dotted lines represent the spectra of $10^{20}$~eV and
151: $1.2\times10^{20}$~eV showers, respectively, in the absence of atmospheric
152: electric field. The thick and thin solid lines correspond to a
153: $10^{20}$~eV shower developing in the presence of a uniform electric field
154: $\pm 1$~kV/cm. In the presence of electric field, the apparent energy of
155: the shower is higher than it's actual energy.
156: }
157: \label{figure:E}
158: \end{figure}
159:
160:
161: An electron or positron with momentum $\vec{p}=\vec{e}_x p_x + \vec{e}_y
162: p_y$ moving in the atmospheric field ${\cal E}(x,y)$ is described by the
163: following system of equations:
164: %
165: \beq
166: \left \{
167: \begin{array}{l}
168: \dot{p_x}=-\kappa(y) \frac{p_x}{p} \\
169: \dot{p_y}=e{\cal E}(x,y) -\kappa(y) \frac{p_y}{p},
170: \end{array}
171: \right.
172: \label{eqn_motion}
173: \eeq
174: %
175: where $\kappa$ accounts for the ionization losses. Here we have assumed
176: $\kappa $ to be independent of energy (a good approximation for the shower
177: electrons). Of course, $\kappa$ depends on the air density, and, hence, on
178: the height $y$ in the atmosphere. The electric field is also a function of
179: coordinates.
180:
181: For a vertical shower in a vertical electric field ${\cal E}={\cal E}(y)$,
182: the horizontal
183: component of the momentum vanishes, $p_x=0$, and the effect of the electric
184: field amounts to reducing the energy losses of the shower in the
185: atmosphere. The equation of motion is
186:
187: \beq
188: \dot{p_y} \approx \frac{d}{dy} p_y=e{\cal E}(y)-\kappa(y),
189: \eeq
190: %
191: where we have taken, as usual $c=1$. The solution of this equation,
192: %
193: \beq
194: p_y(y)= p_{y,0} - \int_{y_0}^{y} (e{\cal E}(\xi)-\kappa(\xi)) d\xi,
195: \label{integral}
196: \eeq
197: %
198: depends on the initial momentum of an electron and on the form of $(e{\cal
199: E}(y)-\kappa(y))$, where $\kappa(y)$ is a function of density only. From
200: the form of this solution it is clear that, for a vertical shower, the
201: electric field can reduce the energy losses and to make the atmosphere
202: effectively thinner. Of course, if the electric field is larger than
203: $\kappa/e$, the electron gains energy faster than it loses energy. This
204: possibility will be discussed below.
205:
206: The energy losses of the shower electrons due to ionization are
207: %
208: \beq
209: \kappa = \frac{dE}{dy}\approx \left ( \frac{\rho}{0.0013 {\ \rm g\,
210: cm}^{-3}} \right
211: ) \left [\ln
212: \frac{E}{m_e}
213: +6.3 \right ] \ \frac{\rm keV}{\rm cm} .
214: \eeq
215: %
216: This should be compared with the atmospheric electric field, which reaches
217: the values
218: %
219: \beq
220: {\cal E }\gsim
221: 1 \, \frac{\rm kV}{\rm cm}
222: \eeq
223: %
224: in thunderclouds~\cite{el_book}. Obviously, the effect of the electric field
225: is non-negligible.
226:
227:
228:
229: The experimental determination of UHECR energy is based on the
230: reconstruction of the total energy of photons and electrons at the shower
231: maximum based on the observed total energy at ground level. The
232: extrapolation to shower maximum involves, effectively, solving the
233: evolution of shower energy due to the energy losses as well as the change
234: in the spectrum and the number of particles due to a number of elementary
235: processes. Most of the particles reaching the ground are photons; in
236: particular, there are more photons than electrons at ground level by an
237: order of magnitude. These photons originate from (i) bremsstrahlung of
238: electrons, (ii) annihilations of positrons. If the electrons are
239: accelerated by the field, the positrons are slowed down, and vise versa.
240: The electric field along the path of a shower can have either polarity.
241: The electric field may also have different directions at different
242: heights~\cite{el_book}. On average, there is more positive charge in the
243: higher layers of the atmosphere, and the ionosphere is almost always
244: positively charged. However, the thunderclouds is where the field reaches
245: a high magnitude. The electric field in a cloud is not necessarily
246: uniform. It can have arbitrary direction along the shower path.
247:
248: If the electrons are accelerated, the atmosphere is effectively thinner for
249: them. Therefore, these electrons can produce the bremsstrahlung photons
250: over a longer period of time, and the resulting photons have, on average, a
251: higher energy. If positrons are accelerated, they have a higher average
252: energy when they annihilate, and, in addition, they annihilate closer to
253: the ground, on average. These are independent and, perhaps, competing
254: effects, so {\em a priori} it is difficult to know which of them dominates.
255: In addition, one has to keep in mind that the ground detectors are
256: sensitive to particles about 0.6-1.2~km away from the shower axis. It is
257: difficult to evaluate the overall effect of the electric fields
258: analytically. We have employed a numerical calculation using
259: AIRES~\cite{AIRES} software package to simulate the effects of the
260: atmospheric electric field. To this end we changed the electron and
261: positron energy loss rate in accordance with eq. (\ref{integral}), assuming
262: a constant electric field along the direction of the shower. It turns out
263: that both the acceleration of electrons and positrons give similar effect.
264:
265: We illustrate the effect in Figures~\ref{figure:R} and \ref{figure:E}. In
266: Fig.~\ref{figure:R} we show the number of photons and electrons reaching
267: the ground as a function of distance from the shower core. The upper plot
268: corresponds to a proton primary, while lower plot is for iron. Dotted
269: (blue) lines present shower profiles with a primary energy $10^{20}$~eV
270: (lower line) and $1.2\times10^{20}$~eV (upper line). Thick (thin) solid
271: line corresponds to the case of a $10^{20}$~eV shower affected by a
272: 1~keV/cm electric field in the direction for which the electrons
273: (positrons) are accelerated. One can see that, in the presence of the
274: electric field, the number of photons increases, and the $10^{20}$~eV
275: shower begins to resemble a $1.2\times10^{20}$~eV shower. Although it
276: makes a difference whether electrons or positrons are accelerated
277: (decelerated), the overall effect increases the apparent shower energy in
278: both cases. One can also see from Fig.~\ref{figure:R} that number of
279: electrons and positrons reaching the ground changes with the direction of
280: the field, but that their number in any case is negligible in comparison
281: with the number of photons.
282:
283: In Fig.~\ref{figure:E} we show the energy distribution of photons and
284: electrons in the shower at ground level. The curve markings are similar to
285: those in Fig.~\ref{figure:R}. In the presence of an electric field, the
286: apparent shower energy may seem higher than the actual energy by as much as
287: 20\%.
288:
289:
290: We emphasize that our simulations were very crude: we assumed a uniform
291: electric field along the shower path. In reality, the electric field
292: configuration can be much more complicated, and this can either increase or
293: decrease the significance of the effect. Also, in these numerical
294: simulations we neglected the effect of an electric field on the muon
295: trajectories. More detailed simulations are required for more reliable
296: results. It would be desirable to measure the electric fields above the
297: experimental site and use these measurements in the data analysis.
298: Measuring the electric field at ground level is relatively straightforward.
299:
300:
301: \section{Muon arrival directions}
302:
303: The Pierre Auger detector is sensitive to both electrons and muons in the
304: shower. For a highly inclined shower, the electromagnetic component dies
305: out before the shower reaches the ground, and, therefore, the data analysis
306: is based on muons alone. The arrival times of muons are used to
307: reconstruct the angle of the shower. Here the effects of the atmospheric
308: electric field must also be taken into account because the electric field
309: bends the muon trajectories.
310:
311: Let us approximate $\kappa $ by a constant (constant density atmosphere),
312: and let us consider a uniform electric field. For a small
313: field, the angle is approximately constant and is close to the initial
314: value $\theta_0$,
315: %
316: \beq
317: \tan \theta_0 =\frac{p_{x,0}} {p_{y,0}}.
318: \eeq
319: %
320: As long as this is the case, the equations (\ref{eqn_motion}) have an
321: approximate solution
322: %
323: \beq
324: \left \{
325: \begin{array}{l}
326: {p_x}=p_{x,0}-\kappa \sin \theta_0 \, t \\
327: {p_y}=p_{y,0} + e{\cal E} t -\kappa\cos \theta_0 \, t. \\
328: \end{array}
329: \right.
330: \label{solu_muon}
331: \eeq
332: %
333: The components of the particle's velocity are
334: %
335: \bea
336: \dot{x} =\frac{p_x}{\sqrt{m^2+p^2}} \\
337: \dot{y} =\frac{p_y}{\sqrt{m^2+p^2} } ,
338: \eea
339: where, as before, $p=\sqrt{p_x^2+p_y^2}$. For the zenith angle one can write
340: %
341: \beq
342: {\rm cotan}\, \theta =\frac{dy}{dx}=
343: \frac{\dot{y}}{\dot{x}}=\frac{p_y}{p_x}.
344: \eeq
345: %
346: Using the approximate solution (\ref{solu_muon}), we get
347: %
348: \bea
349: \dot \theta & = & \frac{d}{dt} \arctan \left ( \frac{p_x}{p_y} \right )
350: \nonumber \\
351: & =& \frac{p_{x,0}(e {\cal E} - \cos \theta_0 \kappa)+ p_{y,0}\sin \theta_0 }{
352: p_0^2+2 A t +B t^2},
353: \eea
354: %
355: where
356: $A=( e {\cal E} - \cos \theta_0 \kappa) p_{y,0}-\kappa \sin \theta_0
357: p_{x,0} $ and $B= e^2E^2-2e E \kappa \cos \theta_0+\kappa^2 $. Therefore,
358: the change in the angle is of the order of
359: %
360: \beq
361: \Delta \theta \sim \frac{e {\cal E}t}{p} \sim 10^{-1} \left (\frac{3 {\rm
362: GeV}}{p} \right )
363: \left (
364: \frac{e E}{1 \, \frac{\rm kV}{\rm cm}}
365: \right )
366: \left (
367: \frac{ct}{3 \, {\rm km}}
368: \right )
369: \eeq
370: %
371: for a uniform electric field $E$. This change is non-negligible. It
372: introduces an error in the angle reconstruction of the arrival direction.
373: In addition, if one uses muons for energy estimation, this effect can
374: introduce a systematic error because one could miscalculate the length of
375: the shower path in the atmosphere.
376:
377: \section{Runaway breakdown}
378:
379: When the electric field in a thundercloud reaches some critical value of the
380: order a few keV, a runaway breakdown may generate a large number of
381: avalanche electrons with energies in the 0.1-10 MeV~range~\cite{uspekhi}.
382: The runaway electron avalanches are triggered by the seed electrons
383: in the extensive air showers initiated by the cosmic rays of energies
384: $10^{16}$~eV and higher~\cite{uspekhi}. There is a mounting evidence that
385: this form of discharge is the reason why the electric field in thunderclouds
386: is limited to a few keV. In the absence of cosmic rays the field could
387: reach values that are an order of magnitude greater than those observed in
388: clouds.
389:
390: If an electron avalanche accompanies an air shower observed by the ground
391: array, the number of electrons and photons registered by the detector could
392: increase significantly, and the cosmic ray energy can be greatly
393: overestimated.
394:
395: The runaway breakdown requires the electric field to reach a critical
396: value~\cite{uspekhi}, which is
397: %
398: \beq
399: {\cal E}_c \approx \frac{44 \, \pi e^3 Z n}{m_e} \approx 2.2\, \frac{\rm
400: kV}{\rm cm}
401: \label{E_c}
402: \eeq
403: %
404: in the lower atmosphere. Here $n$ and $Z$ are the molecular density and the
405: electron number, respectively. Since $10^{16}$~eV showers are apparently
406: sufficient to discharge the cloud, the field is unlikely to be critical at
407: the time and place of a (much less frequent) $10^{18}-10^{20}$~eV shower.
408: However, this probability is non-negligible and it is possible that some of
409: the high-energy showers can appear much more energetic to the ground array.
410: This effect should be taken into account by cosmic ray experiments.
411:
412: \section{Conclusion}
413:
414: We have shown that atmospheric electric fields may affect the energy
415: measurements by the ground arrays in certain weather conditions.
416: Our simplified calculations show that, in the presence of thunderclouds,
417: one may overestimate the shower energy by as much as 20\%.
418: In addition, the deflection of muons may affect the directional
419: reconstruction of inclined showers, which relies on muon arrival
420: times. Finally, if an UHECR shower is accompanied by a runaway breakdown,
421: the number of shower electrons can increase dramatically, which
422: may lead to a significant miscalculation of the cosmic ray energy.
423:
424:
425: \section*{Acknowledgments}
426:
427: We thank K.~Arisaka, J.~Lee, T. Ohnuki, A.~Tripathi, and the rest of the
428: Pierre Auger group at UCLA for numerous discussions and comments. We also
429: thank V.~Rubakov for helpful comments. This work was supported in part by
430: the DOE grant DE-FG03-91ER40662 and the NASA ATP grant NAG~5-13399.
431:
432:
433: \begin{thebibliography}{99}
434:
435: \bibitem{hires}
436: R.~U.~Abbasi {\it et al.} [High Resolution Fly's Eye Collaboration],
437: %``Measurement of the flux of ultrahigh energy cosmic rays from monocular
438: %observations by the High Resolution Fly's Eye experiment,''
439: Phys.\ Rev.\ Lett.\ {\bf 92}, 151101 (2004)
440: [arXiv:astro-ph/0208243].
441: %%CITATION = ASTRO-PH 0208243;%%
442:
443:
444: \bibitem{AGASA}
445: S.~Yoshida {\it et al.},
446: %``The Cosmic ray energy spectrum above 3 x 10**18-eV measured by the Akeno
447: %Giant Air Shower Array,''
448: Astropart.\ Phys.\ {\bf 3}, 105 (1995).
449: %%CITATION = APHYE,3,105;%%
450:
451:
452: \bibitem{AUGER}
453: L.~A.~Anchordoqui [The AUGER Collaboration],
454: %``The Pierre Auger Observatory: Science prospects and performance at first
455: %light,''
456: arXiv:astro-ph/0409470.
457: %%CITATION = ASTRO-PH 0409470;%%
458:
459:
460: \bibitem{intro}
461: R.G.~Fleagle and J.A.Businger, {\em An introduction to atmospheric
462: physics}, Academic Press, San Francisco (1980).
463:
464: \bibitem{el_book}
465: D.R. MacGorman and W.D. Rust, {\em The electrical nature of storms}, Oxford
466: Univ. Press, New York (1998).
467:
468: \bibitem{Marshall}
469: T.~Marshall, W. Rust, and M. Stolzenburg,
470: %Electrical structure and updraft
471: %speeds in thunderstorms over the southern Great Plains,
472: J. Geophys. Res., {\bf 100}, 1001 (1995).
473:
474: \bibitem{thunder_low_energy}
475: V.V.~Alexeenko {\it et al}, Phys.Lett. {\bf 301A}, 299 (2002).
476: \bibitem{thunder_low_energy2}
477: Y.Muraki {\it et al}, Phys. Rev. {\bf D69}, 123010 (2004).
478:
479: \bibitem{electr_thunder}
480: A.V.~Gurevich {\it et al}, Phys. Lett. {\bf 325A}, 389 (2004).
481:
482:
483: \bibitem{uspekhi}
484: A.V. Gurevich and K.P. Zybin,
485: %"Runaway breakdown and electric
486: % discharges in thunderstorms",
487: Phys. Usp., {\bf 44} (11), 1119 (2001).
488:
489:
490: \bibitem{radio_high_energy}
491: A.V.~Gurevich and K.P.~Zybin, Phys.Lett. {\bf 329}, 341 (2004);
492: A.V.~Gurevich, Yu.V.~Medvedev and K.P.~Zybin, Phys.Lett. {\bf 329}, 348
493: (2004);
494:
495: \bibitem{AIRES}
496: S.J.~Sciutto, {\it Proc. 27th ICRC} (Hamburg) {\bf 1}, 237 (2001). See also
497: \newline {\bf http://www.fisica.unlp.edu.ar/auger/aires/}
498:
499:
500:
501:
502: \end{thebibliography}
503:
504: \end{document}
505: