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

2: 

3: %% manuscript produces a one-column, double-spaced document:

4: 

5: %% \documentclass[manuscript]{aastex}

6: 

7: %% preprint2 produces a double-column, single-spaced document:

8: 

9: %% \documentclass[preprint2]{aastex}

10: 

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

12: \newcommand{\myemail}{vahep@stanford.edu}

13: \newcommand{\eg}{{\it e.g.}\,}

14: \newcommand{\ie}{{\it i.e.}\,}

15: \newcommand{\ea}{et al.\,}

16: \newcommand{\beq}{\begin{equation}}

17: \newcommand{\eeq}{\end{equation}}

18: \def\eq{equation\,\,}

19: \def\eqs{equations\,\,}

20: \def\t{\theta}

21: \def\cc{{\rm cm}^{-3}}

22: \def\g{\gamma}

23: \def\loss{{\rm loss}}

24: \def\Coul{{\rm Coul}}

25: \def\sy{{\rm synch}}

26: \def\scat{{\rm scat}}

27: \def\esc{{\rm esc}}

28: \def\eff{{\rm eff}}

29: \def\tr{{\rm cross}}

30: \def\ic{{\rm IC}}

31: \def\br{{\rm brem}}

32: \def\ph{{\rm ph}}

33: \def\crit{{\rm cr}}

34: \def\ac{{\rm ac}}

35: \def\min{{\rm min}}

36: \def\max{{\rm max}}

37: \def\tot{{\rm total}}

38: \def\rt{{\rm rad tot}}

39: \def\mug{\mu{\rm G}}

40: \def\ln{{\rm ln}}

41: \def\Clog{\ln\Lambda}

42: \def\3he{$^3$He\,}

43: \def\he4{$^4$He\,}

44: 

45: 

46: \shorttitle{$^3$He and $^4$He Distributions in SEPs}

47: \shortauthors{Petrosian, Jiang \& Liu}

48: 

49: 

50: \begin{document}

51: 

52: 

53: \title{Relative Spectra and Distributions of Fluences of $^3$He and $^4$He in

54: Solar Energetic Particles}

55: 

56: 

57: \author{Vah\'{e} Petrosian\altaffilmark{1,2,3}, Yan Wei Jiang\altaffilmark{1,2},

58: Siming Liu\altaffilmark{4}, George C. Ho\altaffilmark{5} and Glenn, M.

59: Mason\altaffilmark{5}}

60: 

61: \altaffiltext{1}{Department of Physics, Stanford University, Stanford, CA, 94305

62: email; vahep@stanford.edu; arjiang@stanford.edu}

63: \altaffiltext{2}{Kavli Institute for Particle Astrophysics and Cosmology,

64: Stanford University, Stanford, CA 94305}

65: \altaffiltext{3}{Also Department of Applied Physics}

66: \altaffiltext{4}{Los Alamos National Laboratory, Los Alamos, NM, 87545}

67: \altaffiltext{5}{Applied Physics Laboratory, Johns Hopkins University, 11100

68: Johns Hopkins Road, Laurel, MD  20723}

69: 

70: \begin{abstract}

71: 

72: Solar Energetic Particles  (SEPs) show a rich variety of spectra and relative

73: abundances of many ionic species and their isotopes. A long standing puzzle has

74: been the extreme enrichments of \3he ions. The most extreme enrichments are

75: observed in low

76: fluence, the so-called impulsive, events which are believed to be produced at

77: the

78: flare site in the solar corona with little scattering and acceleration during

79: transport to the Earth. In such events  \3he ions show a

80: characteristic concave curved spectra in a log-log plot.  In two earlier

81: papers (Liu  et al. 2004 and 2006) we showed  how such extreme enrichments

82: and such spectra can result in the model developed by  Petrosian \& Liu (2004),

83: where ions

84: are accelerated stochastically by plasma waves or turbulence.  In this paper we

85: address the relative distributions of the fluences of  \3he and  \he4 ions

86: presented by Ho et al. (2005) which show that  while the distribution of  \he4

87: fluence (which we believe is a good measure of the flare strength)

88:  like many other extensive characteristics of solar flare,  is

89: fairly broad, the \3he fluence is

90: limited to a narrow range. Moreover, the ratio of the fluences  shows a strong

91: correlation with the \he4 fluence. One of the predictions of our model presented

92: in the 2006 paper

93: was presence of steep variation of the fluence ratio with the level of

94: turbulence

95: or the rate of acceleration. We show here that this feature of

96: the model can reproduce the observed distribution of the fluences with very few

97: free parameters. The primary reason for the success of the model in both fronts

98: is because fully ionized \3he ion,  with its unique charge to mass ratio,

99: can resonantly interact with more plasma modes and accelerate more readily  than

100: \he4. Essentially in most flares, all

101: background $^3$He ions are accelerated to few MeV/nucleon range, while this

102: happens for $^4$He ions only in very strong events. A much smaller fraction of

103: $^4$He ions reach such energies in weaker events.

104: \end{abstract}

105: 

106: 

107: \keywords{Sun: flares, particle emissions -- acceleration--plasmas --

108: turbulence--waves }

109: 

110: 

111: \section{INTRODUCTION}

112: \label{intro}

113: Solar flares are excellent particle accelerators. Some of these particles on

114: open field lines are observed as solar energetic particles (SEPs) at one AU or

115: produce type III and other radio radiation. Those on closed field lines can be

116: observed by the radiation they produce as they interact with solar plasma and

117: fields. Electrons produce nonthermal bremsstrahlung and synchrotron photons in

118: the hard X-ray and microwave range, while protons (and other ions) excite

119: nuclear lines in the 1 to 7 MeV range or may produce higher energy gamma-rays

120: via $\pi^0$ production and its decay. It appears that stochastic acceleration

121: (SA) of particles by plasma waves or turbulence  plays an important role in

122: production of high energy particles and 

123: consequent plasma heating  in solar flares (e.g., Ramaty 1979; M\"{o}bius et 

124: al. 1980, 1982;  Hamilton \&  Petrosian 1992; Miller et al. 1997; Petrosian \&

125: Liu 2004, hereafter PL04).

126: This theory was applied 

127: to the acceleration of nonthermal electrons (Miller \& Ramaty 1987;  Hamilton \&

128: Petrosian 1992). It appears that it can produce many of the observed radiative

129: signatures such as broad band spectral features (Park, Petrosian \& Schwartz

130: 1997; PL04) and the commonly observed hard X-ray emission from the tops of

131: flaring loops (Masuda et al. 1994;  Petrosian \& Donaghy 1999). 

132: It is also commonly believed that the observed relative abundances of ions in

133: SEPs

134: favor a SA model (e.g. Mason et al. 1986 and Mazur et al. 1992). More recent

135: observations have

136: confirmed this picture (see Mason et al. 2000, 2002, Reames et al. 1994 and

137: 1997, and Miller 2003).

138: One of the most vexing problem of SEPs has been the  enhancement of \3he

139: in the so-called {\it impulsive  or \3he-rich events}, which sometimes can be

140: $3-4$ orders of

141: magnitude above the photospheric value%

142: \footnote{In addition there is charge-to-mass ratio dependent enhancement

143: relative to the photospheric values  of heavy ions in SEPs , and  in few flares

144: gamma-ray line emissions also points to anomalous abundance pattern of the

145: accelerated ions (Share \& Murphy 1998; Hua,  Ramaty \& Lingenfelter 1989). We

146: will not be dealing with these anomalies in this paper.}.

147: There have been many attempts to  explain this enhancement.

148: Most of the proposed models, except the  Ramaty and Kozlovsky (1974) model based

149: on  spalation (which has many problems), rely on resonant wave-particle

150: interactions and the unique charge-to-mass ratio of \3he  (see e.g. Ibragimov \&

151: Kocharov 1977; Fisk

152: 1978; Temerin \& Roth 1992; Miller \& Vi\~nas 1993; Zhang 1995; Paesold,

153: Kallenbach \& Benz 2003). Most  of these model assume presence of some

154: particular kind of waves which preferentially \underline{heats} \3he ions to a

155: higher temperature than \he4 ions, which then become

156: seeds for

157: subsequent acceleration by some (usually) unspecified mechanism (for more

158: detailed discussion see Petrosian 2008). None

159: of these earlier works did a compare  model spectra with

160: observations.

161: 

162: \begin{figure}[htb]

163: \begin{center}

164: \includegraphics[height=4.7cm]{f1a.eps}

165: \includegraphics[height=5.5cm]{f1b.eps}

166: \hspace{-1cm}

167: \includegraphics[height=5.9cm]{f1c.eps}

168: \end{center}

169: \caption{\scriptsize

170: {\it Left:} Variation of the ratio of \3he to \he4 fluences with the fluence of

171: \he4 showing a continuum of enrichments and a strong anti correlation. {\it

172: Middle:}  \3he vs \he4 fluences showing a much larger range for the latter while

173: the former seems to be limited to a small range. Note that the \3he fluences do

174: not concentrate at the lower end which would be the case if observational

175: threshold  was affecting their distribution.  {\it Right:} The distribution of

176: fluences of \3he and \he4. Note that the high end of the \he4 distribution may

177: be truncated because of the threshold of the fluence ratio (missing point in the

178: lower left triangle of the middle panel) [From Ho05].  The fluences are in units

179: of particles/(cm$^{2}$ sr MeV/nucleon).

180: }

181: \label{obs}

182: \end{figure}

183: 

184: In two more recent papers Liu, Petrosian \& Mason 2004  and 2006 (LPM04, LPM06)

185: have demonstrated that a SA model by parallel propagating waves

186: can explain both the extreme enhancement of \3he

187: and can  reproduce the observed \3he and \he4 spectra. In LPM06 it was shown

188: that the relative fluences of these ions, and to a lesser extent their spectral

189: indexes, depend on several model parameters so that in a large sample of events

190: one would expect some dispersion in the distributions of fluences and spectra. 

191: Ho et al. (2005; Ho05) analyzed a large sample of events and provide

192: distributions of \3he and \he4 fluences and the correlations between them. Our

193: aim here is to explore the possibility of explaining these observations

194: by the above mentioned dependence of the fluences on the model parameters.  In

195: particular we would like to explain the observations reproduced in Figure

196: \ref{obs} which shows a strong anti correlation of \3he/\he4 ratio with

197: \he4 fluence (left panel), but shows essentially no correlation between the two

198: fluences (middle panel). More strikingly,  the \3he fluence distribution appears

199: to be relatively narrow and follows a log-normal distribution, while \he4

200: distribution is much broader and  may have a power law distribution in the

201: middle of the range, where the

202: observational selection effects are unimportant. Often

203: the SEPs are divided into two classes; impulsive-high enrichment and

204: gradual-normal abundance classes. However, as evident from the left panel of the

205: above figure there is a continuum of enrichment extending over many orders of

206: magnitude.%

207: \footnote{The \he4 distribution shows a weak sign of bi modality but this is not

208: statistically significant. In this paper we will ignore this feature.}

209: 

210: 

211: In the next section we describe some of the model characteristics that can

212: explain these observations and in \S3 we compare the model predictions with the

213: observations, specifically the distributions of the fluences. A brief summary

214: and conclusion is given in \S4.

215: 

216: \section{MODEL CHARACTERISTICS}

217: \label{model}

218: 

219: The model used in LPM04 and LPM06 which successfully described the enrichment

220: and

221: spectra in several flares has several free parameters.  As usual we have the

222: plasma parameters

223: density $n$, temperature $T$ 

224: and magnetic field $B_0$. It turns out that the final results are insensitive to

225: the temperature  as long as it is higher than $2\times 10^6$K (see Fig.

226: \ref{ratios} below), which is the case for

227: flaring coronal loops. It also turns out that only a combination of density  and

228:  magnetic field ($\sqrt n/B_0$)  comes into play. We  express this as the

229: ratio of plasma to gyro-frequency of electrons,

230: $\alpha=\omega_{pe}/\Omega_e$ which is related to the Alfv\'en velocity in unit

231: of speed of light;

232: $\beta_A=\delta^{1/2}/\alpha$, where $\delta=m_e/m_p$ is the ratio of the

233: electron to proton masses. So in reality  we have only one effective free plasma

234: parameter $\alpha$ or $\beta_A$.  On the other hand, several parameters are

235: required to describe the spectrum of the turbulence.

236: Following the above papers we assume broken power laws for the two

237: relevant modes, the proton cyclotron (PC) and He cyclotron (HeC),

238: with an inertial range $k_\min<k<k_\max$, and similar power law indexes $q$ and

239: $q_h$ in

240: and beyond the inertial range, respectively.%

241: \footnote{In LPM06 we also have an index $q_l$ describing the power law below

242: the inertial range which is of minor consequence. For all practical purposes we

243: can assume a sharp cutoff below  $k_\min$ which means

244: $q_l\rightarrow \infty$.}

245: The only difference between the two branches is that the 

246: wave numbers  $k_\max$ and $k_\min$ for the PC mode are two times higher 

247: than those for the HeC mode. Finally there is the most important parameter

248: related

249: to the total energy density of  turbulence,  ${\cal E}_{\rm tot}$, which

250: determines both the rate of acceleration and, when integrated over  the

251: volume of the source region, determines the intensity or the strength of the

252: event. This parameter is  the characteristic time

253: scale $\tau_p$ or its inverse the  rate defined as  (see, e.g. Pryadko \&

254: Petrosian 1997) 

255: \begin{equation}

256: \tau_p^{-1} = {\pi\over 2}\Omega_e\left[{4\overline{\cal E}_0\over

257: B_0^2/8\pi}\right]\,\, \ \ \ \ {\rm with} \ \ \ \ \overline{\cal E}_0 =

258: {(q-1){\cal E}_{\rm tot}\over 

259: (k_\min c/\Omega_e)^{1-q}},

260: \label{taup}

261: \end{equation}  

262: for each mode. The factor of 4  arises from having two branches (PC and HeC)

263: and  two propagation directions of the waves (see LPM06 for details).

264: 

265: As shown in LPM04 and LPM06 papers the main difference between the acceleration

266: process of \3he and \he4 is in the difference between their acceleration rate or

267: timescales ($\tau_a$). The other relevant timescales, namely the loss

268: ($\tau_{\rm loss}$) and escape ($T_{\rm esc}$) times are

269: essentially identical for the two ions (e.g. see left panel of  Fig. 7 of

270: LPM06). The

271: acceleration timescales are different mainly at low energies (typically below

272: one MeV/nucleon), where the acceleration time of \he4 is a longer (by one  to

273: two orders of magnitude). As

274: a result at these low energies the \he4 acceleration time may be comparable or

275: longer than the loss time which makes it difficult to accelerate \he4 ions. Most

276: of \he4 ions are piled up below some energy (roughly where $\tau_a=\tau_{\rm

277: loss}$) and

278: only a few of

279: them accelerate into the observable range (e.g. see right panel of  Fig. 7 of

280: LPM06).

281: However, because the acceleration times

282: scale as $\tau_p$  while the loss time does not, for higher level

283: of turbulence (larger $\overline{\cal E}_0$), the acceleration time may fall

284: below the

285: loss time so that \he4 ions can be then accelerated more readily (see Fig.

286: \ref{spectra} below). On the other

287: hand,

288: essentially independent of values of any of the above parameters, the \3he

289: acceleration time at all energies, in particular at low energies, is always far

290: below its loss time so that in all cases   (except for very high densities or

291: very low values of $\tau_p^{-1}$)

292: \3he ions are accelerated easily to high energies. The relative values of the

293: escape and acceleration times (for

294: both ions) determine their  high energy spectral cutoffs.  

295: 

296: Figure \ref{acctimes} shows variation with energy of acceleration times of \3he

297: (thick lines) and  \he4 (thin lines) and their dependence on  parameters

298: $k_\min$, $\alpha$ and  $q$. The remaining parameters $q_h$ and $k_\max$ only

299: affect the slope of the low energy

300: end of \he4 which does not affect the spectra noticeably.

301: It is evident that  the general behavior of the acceleration time scales  described above

302: (consisting of a low and a high energy monotonically

303: increasing  branches with a declining transition in between)

304: is  present in all  models. These features

305: change only quantitatively and often by  small amounts.  As expected lowering

306: $k_\min$ decreases the acceleration times at the  high energy branch (left

307: panel). This is because the lower $k_\min$ waves interact  resonantly with

308: higher energy ions.  On the other hand, a lower value of $\alpha$ (or larger

309: Alfv\'en velocity or magnetization) decreases

310: the times at the low energy branch  (middle panel). Steeper spectra in the

311: inertial range,

312: produce a higher rate of acceleration (larger ${\cal E}_0$; see eq.

313: [\ref{taup}]) and  decrease the overall acceleration time scales (right panel) 

314: 

315: \begin{figure}[htb]

316: \begin{center}

317: \includegraphics[height=5.0cm]{f2a.eps}

318: \hspace{-0.6cm}

319: \includegraphics[height=5.0cm]{f2b.eps}

320: \hspace{-0.6cm}

321: \includegraphics[height=5.0cm]{f2c.eps}

322: \end{center}

323: \caption{\scriptsize

324: Dependence of the acceleration time of \he4 (thin, blue) and \3he (thick, red)

325: on $k_\min$ (left),

326: $\alpha$ (middle) and $q$ (right).  The lines are labeled with the corresponding

327: numbers of each parameter.

328: In each case  the solid lines are for the  fiducial model with

329: $\alpha=0.5,

330: k_\max=2\alpha\delta^{-1/2}=10k_\min, q=2$ and $q_h=4$. 

331: } 

332: \label{acctimes}

333: \end{figure}

334: Note  that in this and subsequent figures, $k_\max$ is in units

335: of $\Omega_p/c$ so that  $k_\max=2\alpha\delta^{-1/2}$ in the labels means an

336: actual $k_\max=2\Omega_p/v_A= \sqrt{2\beta_p}/r_{g,p}$, where

337: ${\beta_p}=2(v_{th,p}/v_A)^2$ is the plasma beta, and $v_{th,p}=k_BT/m_p$ and

338: $r_{g,p}=v_{th,p}/\Omega_p$ are the proton thermal velocity and gyro radius. The

339: scale of $k_\max$  is clearly beyond the MHD regime (where the wave frequency 

340: $\omega=v_A k\ll \Omega_p$), but is below the proton gyro radius for the chosen

341: parameters ($\sqrt{2\beta_p}\sim0.03$). 

342: 

343: Using these acceleration rates we calculate spectra of  the two ions (as  in

344: LPM04 and LPM06) for a range of 

345: parameters. Figure \ref{spectra} shows three sets of spectra where we

346: vary $k_\min, \alpha$ and $\tau_p^{-1}$. In each panel the sold lines are for

347: the fiducial model ($\alpha=0.5, k_\max=2\alpha\delta^{-1/2}=10k_\min, q=2$ and

348: $q_h=4$) chosen to fit the spectra observed by {\it ACE/ULEIS} for 30 Sep. 1999

349: event.

350: The spectral variations here reflect

351: the above  described variations of the acceleration timescales.  Lower $k_\min$

352: (or larger inertial range)  yields a larger tail for both ions (left panel).

353: Variation of $\alpha$ has a similar and smaller effect on \3he spectra but it

354: affects the \he4 spectra dramatically; for $\alpha\sim 1$ essentially there is

355: no \he4 acceleration but the  $\alpha\sim 1/4$ model accelerates a large number

356: of \he4

357: ions beyond 0.1 MeV/nucleon and into the observable range (middle panel). This

358: effect is even more pronounced for increasing values of $\tau_p^{-1}$, where a

359: factor of few increase in the general rate of acceleration (or the level of

360: turbulence) causes a large increase of the fluence of \he4  (right panel),

361: because, as stated above, its acceleration time  becomes shorter than its loss

362: time even al low energies.  All

363: these spectra show the same general characteristic features. While most \3he

364: ions are accelerated to  high energies for essentially all  model parameters

365: appropriate for solar coronal conditions and reasonable level of turbulence,

366: \he4

367: ions show a characteristic lower energy bump with a nonthermal hard tail. In

368: general, the lower energy bump is below the observation range except for low

369: $\alpha$ and high values of $\tau_p^{-1}$. Since a high level of turbulence is

370: expected for brighter and stronger events, this means that we get smaller

371: \3he/\he4 flux or fluence ratios for brighter events. Note that the spectra in

372: such cases may not agree with observations but this is not troublesome, because

373: as is well established, the stronger events (the so-called gradual events) are

374: associated

375: with CMEs and shocks which most likely will modify the above spectra which are

376: those of ions escaping the corona. Thus, the higher energy bumps in the spectra

377: shown here should be considered as seeds for

378: such further acceleration during the transport from the lower corona to the

379: Earth, which becomes more likely, and is

380: expected to change the above spectra more significantly,  for more energetic

381: events.  Thus if we give up the idea that there are two distinct classes of SEPs

382: (impulsive and highly enriched and gradual and normal abundance) but that there

383: is a continuum of events, which observations in Figure \ref{obs} show, then the

384: above scenario implies that the main acceleration occurs in the solar corona.

385: Subsequent interactions in CME shocks mainly modify the seed population escaping

386: the turbulent coronal site.

387: 

388: \begin{figure}[htb]

389: \begin{center}

390: \includegraphics[height=5.0cm]{f3a.ps}

391: \hspace{-0.6cm}

392: \includegraphics[height=5.0cm]{f3b.ps}

393: \hspace{-0.6cm}

394: \includegraphics[height=5.0cm]{f3c.ps}

395: \end{center}

396: \caption{\scriptsize

397: Dependence of the accelerated spectra of \he4 and \3he on $k_\min$ (left),

398: $\alpha$ (middle) and $\tau_p^{-1}$ (right; $\tau_p^{-1}$ in units of $\tau_p^{-1,0}=0.0055$

399: s$^{-1}$ ).  

400: The lines are labeled with the corresponding

401: numbers of each parameter.

402: In each case the  solid lines are for the  fiducial model  with $\alpha=0.5,

403: k_\max=2\alpha\delta^{-1/2}=10k_\min, q=2$ and $q_h=4$ that is chosen to fit the

404: data point shown for the 30 Sep. 1999 event observed by {\it ACE}. Note that for

405: a better

406: indication what energy particles dominate the spectra in a log-log plot we plot

407: particle energy time fluence.

408: }

409: \label{spectra}

410: \end{figure}

411: 

412: From the spectra we can calculate the ratio of \3he to \he4 fluences for

413: different models which could be then compared with the observed ratios shown in

414: Figure \ref{obs}. Inspection of observed spectra indicate that a

415: representative ion energy would be 1 MeV/nucleon. In Figure \ref{ratios} we

416: show the variation of this ratio with temperature (left panel) and  

417: $\tau_p^{-1}$ (middle and right panels) for several values of other important

418: parameters. As evident this ratio is most sensitive to the value of

419: $\tau_p^{-1}$ which represents the general rate of acceleration or the level of

420: turbulence. The ratio can change from the highest observed value  ($\sim 30$)

421: to  near photospheric value  ($\sim 2\times 10^{-4}$) for only a factor of 30

422: change in $\tau_p^{-1}$. It is natural to expect higher level of turbulence 

423: generation (i.e.  a larger value of $\tau_p^{-1 }$) in stronger  events.

424: Therefore, this predicted correlation is in agreement with the general trend of

425: observation shown in Figure \ref{obs} (left panel), if the strength of an event

426: is measured by the observed fluence of \he4 ions and most other ions like

427: carbon, nitrogen and oxygen.%

428: \footnote{It should be noted that while the observations are for fluences

429: integrated from 0.2 to 2.0 MeV/nucleon our theoretical ratios are calculated at

430: 1 MeV/ nucleon  which is near the geometric or algebraic mean of the range.}  

431: This seems reasonable and calls for  more quantitative comparison with

432: observations and model prediction. In the next section we present one such

433: comparison.

434: 

435: \begin{figure}[htb]

436: \begin{center}

437: \includegraphics[height=5.0cm]{f4a.eps}

438: \hspace{-0.6cm}

439: \includegraphics[height=5.0cm]{f4b.eps}

440: \hspace{-0.6cm}

441: \includegraphics[height=5.0cm]{f4c.eps}

442: \end{center}

443: \caption{\scriptsize

444: Variation  of the accelerated \3he to  \he4 fluence ratio (at $E=1$ MeV/nucleon)

445: with background plasma temperature $T$ (left) and

446: $\tau_p^{-1}$ (middle and right) for several values of other specified 

447: model parameters.  

448: The lines are labeled with the corresponding numbers of each parameter.

449: In each case the open circle stands for the model that fits spectra of  the 30

450: Sep. 1999 event, and the solid lines are for the  fiducial

451: model  with $\alpha=0.5, k_\max=2\alpha\delta^{-1/2}=10k_\min, q=2$ and $q_h=4$.

452: Note the weak dependence on the temperature for $T>2\times

453: 10^6$ K and a strong dependence on $\tau_p^{-1}$ for all model parameters with

454: saturates at chromospheric values of the ratio. The horizontal dot-dash line

455: shows the highest ratio observed so far (see Fig. \ref{obs}, left). 

456: }

457: \label{ratios}

458: \end{figure}

459: 

460: \section{Distributions of Fluences}

461: 

462: We have seen that the general observed behavior of the the ratio of the fluences

463: defined as $R=F_3/F_4$  is similar to the model predictions. In this section we

464: try to put this result on a firmer quantitative footing by considering the

465: observed distributions of the fluences of both ions as shown in Figure \ref{obs}

466: (right panel).  Except for the minor truncation at high values of $F_4$,  the

467: fluence

468: of \he4, the observed distribution of $F_3$, the fluences of \3he, seem to

469: be almost bias free and not affected significantly by the observational

470: selection effects. For example, there are well defined and steep decline both at

471: the high and low fluences away from the peak \3he value of $F_0\sim 10^{3.7}

472: {{\rm 

473: particles} \over {{\rm  cm}^{2} {\rm sr} {\rm (MeV/nucleon)}}}$. This is not

474: what

475: one would expect if the

476: data suffered  truncation due to a low observation threshold. In such a case one

477: would observe a

478: distribution increasing up to the threshold followed by a rapid cutoff below it.

479: Our model results described above

480: also seem to predict the observed behavior. As stressed in previous

481: section, the \3he

482: spectra and fluxes appear to be fairly independent of model parameters because

483: essentially under all conditions most \3he ions are accelerated and form a

484: characteristic concave spectrum. Thus we believe that it is safe to assume that

485: the observed \3he distribution is a true representations of the intrinsic

486: distribution (as produced on the Sun).  This distribution can be fitted very

487: nicely with a log-normal expression.%

488: \footnote{The truncation shown by the shaded area in the middle panel of Figure

489: \ref{obs} introduces a slight bias against detection of low fluences. We

490: estimate

491: that, because there are fewer events at the high \he4 fluence end, this means a

492: 10 to 20\% underestimation of the distribution of the three lowest values of the

493: \3he histogram (right panel, Fig. \ref{obs}). We will ignore this small

494: correction, whose main effect is to increase the value of $\sigma_3$ by a small

495: amount.} 

496: If we define the logs of the  fluences and their ratio as 

497: \beq

498: LF_3\equiv\ln (F_3/F_0),\,\,\,\,\  LF_4\equiv\ln (F_4/F_0) \,\,\,\,\ LR\equiv\ln

499: R,

500: \label{defs}

501: \eeq

502: then from fitting the observed distribution of \3he by a log-normal form we

503: get: 

504: \beq

505: \psi_3(LF_3)=\phi_0\exp\left({LF_3 \over \sigma_3}\right)^2\,\,\,\,\, {\rm

506: with}\,\,\,\,\, \sigma_3\sim 0.22,

507: \label{dist3}

508: \eeq

509: which is shown on the right panel of Figure \ref{results}. 

510: 

511: Using this distribution we now derive the distribution of \he4 fluences,

512: $\psi_4(LF_4)$. For this we use the model predicted relationship between the two

513: fluences as shown in Figure \ref{ratios} above. We will use the  two panels of

514: this

515: figure showing the dependence of the log of the fluence ratio $LR$ on 

516: $\tau_p^{-1 }$.

517: It turns out that most of these curves can be fitted by a simple function:

518: \beq

519: \ln (R/R_0)=LR-\ln R_0={A\over \ln (\tau_p^{-1 }/\tau_{p0}^{-1 })}.

520: \label{relation}

521: \eeq

522: The left panel of Figure \ref{results} shows fits to the curves in the right

523: panel of Figure \ref{ratios} with the indicated values of the the fitting

524: parameters $A, R_0$ and $\tau_{p0}^{-1 }$ (which is not the same as the 

525: $\tau_{p,0}^{-1 }=0.0055 {\rm s}^{-1}$ in Fig. \ref{spectra}). We

526: shall use this relation to transfer the \3he fluences and distributions to those

527: of  \he4.

528:  

529: \begin{figure}[htb]

530: \begin{center}

531: \includegraphics[height=5.2cm]{f5a.ps}

532: \hspace{+0.6cm}

533: \includegraphics[height=5.2cm]{f5b.ps}

534: \end{center}

535: \caption{\scriptsize {\it Left:} A simple analytic fit (curves) to the model

536: relations (points) between the fluence ratio and the acceleration rate or event

537: strength represented  by $\tau_p^{-1 }$    for the  three different values of

538: $k_\min$ of

539: the right panel of Figure \ref{ratios} with the indicated fitting parameters.

540: {\it Right:} The fitted log-normal distribution to the \3he fluences and 

541: predicted \he4 distributions of three models  compared with observations. The

542: solid line which gives the best fit is for  $n=2, k_\min=0.1k_\max$, the dashed

543: line is for $n=2, k_\min=0.2k_\max$  and the dash-dot line is for  $n=1.5,

544: k_\min=0.2k_\max$.

545: }

546: \label{results}

547: \end{figure}

548: 

549: For a given value of  $\tau_{p}^{-1 }$ the number of events with \he4 

550: log-fluences between  $LF_4$ and $LF_4+d(LF_4)$ (i.e. $\psi_4(LF_4)d(LF_4)$) is

551: equal to $\psi_3(LF_3)d(LF_3)$, the number of events with  \3he log-fluence

552: $LF_3 = LF_4 +LR(\tau_p^{-1})$ and  $LF_3+d(LF_3)$, where$d(LF_3)=d(LF_4)$and 

553: $LR(\tau_p^{-1})=\ln R_0 + A/\ln (\tau_p^{-1}/\tau_{p,0}^{-1 })$. Thus  we have

554: \beq

555: \psi_4(LF_4)=\psi_3(LF_4+LR[\tau_p^{-1}])=\phi_0\exp\left({LF_4+LR(\tau_p^{-1})

556: \over \sigma_3}\right)^2

557: \label{dist41}

558: \eeq

559: However, we expect not a single value for  $\tau_p^{-1 }$, which as stated above

560: is a proxy for the strength of

561: the event,  but a broad distribution of events with different strengths, say

562: $f(\tau_p^{-1 })$. Since, as argued above, the \3he fluence distribution

563: $\psi_3(LF_3)$ is independent of $\tau_p^{-1 }$,  then for a population of

564: events we have

565: \beq

566: \psi_4(LF_4)=\int_0^\infty \phi_0\exp\left({LF_4+LR(\tau_p^{-1}) \over

567: \sigma_3}\right)^2f(\tau_p^{-1 })d\tau_p^{-1 }.

568: \label{dist42}

569: \eeq

570: Every term in the above equations is determined by observations and our models

571: except the distribution $f(\tau_p^{-1 }$, which is a reflection of the level of

572: the distribution of the level of turbulence and, when multiplied by the volume

573: of the turbulent acceleration region (which does not affect the \3he/\he4

574: ratio), is related the overall strength of the event. 

575: Observations of solar flares show that most extensive  characteristics

576: which are a good measure of the flare strength or magnitude, such as X-ray,

577: optical or radio fluxes, appear to obey a steep power law distribution, usually

578: expressed as a cumulative distribution $\Phi(>F_i)\propto F_i^{-n}$  (or 

579: differential distribution $\phi(F_i)\propto F_i^{-n-1}$) with typically $n\sim

580: 1.5$ (see, e.g.  Dennis 1985 and reference therein). Such a distribution seems

581: to roughly

582: agree with the prediction of the so-called avalanche  model proposed by Lu \&

583: Hamilton (1991).

584: Now assuming that  $\tau_p^{-1 }$ also obeys such a power law distribution

585: (\i.e. $f(\tau_p^{-1 })\propto (\tau_p^{-1 })^{-(n+1)}$) we can write the

586: distribution of \he4 as:

587: \beq

588: \psi_4(LF_4)=\int_0^\infty \phi_0\exp\left({LF_4+\ln R_0 +A/x \over

589: \sigma_3}\right)^2e^{-nx}dx, \,\,\,\,\, {\rm with }\,\,\,\,\,

590: x\equiv \ln (\tau_p^{-1 }/\tau_{p0}^{-1 }).

591: \label{dist43}

592: \eeq

593: Using the above relations we have calculated the \he4 fluence distribution. The

594: results for three models are compared with the observations on the right panel

595: of Figure \ref{results}. Given  the other model parameters ($k_\min, \alpha$

596: etc.) we have only one free parameter namely the index $n$ for this fit. The

597: solid line obtained for the top curve of the left panel ($k_\min=0.2k_\max,

598: \alpha=0.5$), and for $n=2$ provides a good fit to the observed distribution of

599: \he4 fluences.  In order to demonstrate the sensitivity of the results to the

600: parameters we also show two other model predictions based on slightly different

601: parameter values. These results provide additional quantitative evidence (beside

602: those given in LPM04 and LPM06) on the validity of the SA of SEPs by turbulence,

603: and  indicate that  with this kind of analysis one can begin to constrain model

604: parameters. 

605: 

606: 

607: \section{SUMMARY AND CONCLUSIONS}

608: 

609: In this paper we have carried out further comparison between the prediction of

610: models based on stochastic acceleration of SEP ions by turbulence. In our

611: earlier works (LPM04, LPM06) we demonstrated that the extreme enrichments of

612: \3he and spectra of \he4 and \3he  observed in several events can be

613: naturally described in such a model. Using the results based on this model, here

614: we consider the relative distributions of \he4 and \3he fluences derived from a

615: large sample of event by Ho05. We show that with some simple and reasonable

616: assumptions we can explain the general features of these observations as well.

617: 

618: These are clearly preliminary results and are intended to demonstrate

619: that in addition to modeling only few bright events it is also important to look

620: at population as  a whole and ascertain that a model which can explain the

621: detail

622: characteristics of individual events can also agree with the distributions of

623: observables for a large sample of events. Here we have shown how the dispersion

624: in one parameter, namely the acceleration rate or the strength of the flare, can

625: account for the observed distributions of fluences. The key assumption here is

626: that the amount  of produced turbulence  (represented by $\tau_p^{-1}$) has a

627: wide dispersion and obeys a power law distribution similar to that observed for 

628: other extensive parameters that give a measure of the strength of a flare. 

629: The dispersion in

630: other model parameters can also influence the final outcome. However, the

631: dispersion of most of

632: the other important parameters, like  intensive parameters temperature, density

633: and 

634: magnetic field, are  expected to be much

635: smaller than that of an

636: extensive parameter like the overall strength of the event, the amount of

637: turbulence produced, the flare volume etc. In addition, as shown in the previous

638: section,

639: the intensive parameters play a lesser role than the extensive parameter

640: $\tau_p^{-1 }$ in determining the relative characteristics of \3he and \he4.

641: Given the dispersion of any other parameter one may carry out similar

642: integration

643: over its range. However, for the reasons given above we expect smaller changes

644: in the shapes of predicted distribution due to dispersion of most of the

645: intensive parameters. Given a more extensive set of data such improvements may

646: be needed can be  carried out.

647: 

648: The existing data may be used to test some of our assumptions, in

649: particular the assumption of constancy of the \3he distribution. We intend to

650: address these in future works. We can also make the above results more robust by

651: using model

652: fluences integrated over the same spectral range as the observations instead of

653: fluences at 1 MeV/nucleon. One can also expand this approach and address the

654: distributions of other characteristics besides the fluence, suc as the

655: spectral indexes or break energies (if any). The available data

656: contain this

657: information but require more analysis. 

658: 

659: \begin{thebibliography}{}

660: 

661: %\begin{references}

662: 

663: \bibitem

664: [Dennis(1985)]{Dennis85} Dennis, B. R. 1985, Sol. Phys. 100, 465

665: 

666: \bibitem

667: [Fisk (1978)] 

668: {Fisk78} Fisk, L. A.\ 1978, \apj, 224, 1048

669: 

670: \bibitem

671: [Hamilton \& Petrosian (1992)]

672: {Hamilton92} Hamilton, R. J., \& Petrosian, V.\ 1992, 

673: \apj, 398, 350

674: 

675: \bibitem[{Ho} et~al. 2005]{Ho2005}Ho, G. C., Roelof, E. C., \& Mason G. M. 2005,

676: \apjl,  621, L141

677: \bibitem[{Liu} et~al. 2004]{Liu2004} Liu, S., Petrosian, V., \& Mason, G. M.

678: 2004, \apjl, 613, L81

679: 

680: \bibitem

681: [Hua et al. (1989)] 

682: {Hua89} Hua, X. M., Ramaty, R., \& Lingenfelter, R. E.\ 1989, 

683: \apj, 341, 516

684: 

685: 

686: \bibitem[{Liu} et~al. 2006]{Liu2006} Liu, S., Petrosian, V., \& Mason, G. M.

687: 2006,

688: \apj, 636, 462

689: 

690: \bibitem[{Lu} et~al. 1991]{Lu1991} Lu, E. T.,  \& Hamilton, R. J. 1991, \apjl,

691: 380, L89

692: 2006,

693: \apj, 636, 462

694: 

695: \bibitem[Mason et al. (2002)] {Mason02a} Mason, G. M., Mazur, J. E., \& Dwyer,

696: J. R.\ 2002, 

697: \apj, 565, L51

698: 

699: \bibitem[Mason et al. (2004)] {Mason04} Mason, G. M., Mazur, J. E., Dwyer, J.

700: R., Jokipii, 

701: J. R., Gold, R. E., \& Krimigis, S. M.\ 2004, \apj, 606, 555

702: 

703: \bibitem[Mason et al. (1989)]{Mason89}{Mason, G. M., Ng, C. K., Klecker, B., \&

704: Green, G.\ 

705: 1989, ApJ, 339, 529} 

706: 

707: \bibitem

708: [Mason et al. (1986)] 

709: {Mason86} Mason, G. M., Reames, D. V., Klecker, B., Hovestadt, 

710: D., \& von Rosenvinge, T. T.\ 1986, \apj, 303, 849

711: 

712: \bibitem

713: [Mason et al. (2002)] 

714: {Mason02b} Mason et al.\ 2002, \apj, 574, 1039

715: 

716: \bibitem

717: [Masuda et al. (1994)] 

718: {Masuda94} Masuda, S., Kosugi, T., Hara, H., Tsuneta, S., \& 

719: Ogawara, Y.\ 1994, Nature, 371, 495

720: 

721: \bibitem

722: [Mazur et al. (1995)] 

723: {Mazur95} Mazur, J. E., Mason, G. M., \& Klecker, B.\ 1995, 

724: \apj, 448, L53

725: 

726: %\bibitem[McDonald \& Van Hollebeke (1985)] {McDonald85} McDonald, F. B., \& Van

727: %Hollebeke, 

728: %M. A. I.\ 1985, \apj, 290, L67

729: 

730: \bibitem[Miller (1997)] {Miller 97} Miller, J. A. 1997, \apj, 491, 939

731: 

732: \bibitem

733: [Miller (2003)]

734: {Miller03} Miller, J. A.\ 2003, Multi-Wavelength Observations of 

735: Coronal Structure and Dynamics eds. Petrus C.H. et al. (COSPAR Colloquia Series

736: Vol. 13), 387

737: 

738: %\bibitem[Miller et al. (1996)] {Miller 96} Miller, J. A., LaRosa, T. N.,

739: %\&Moore, R. L., 

740: %1996, \apj, 461, 445

741: 

742: %\bibitem[Miller \& Roberts (1995)] {Miller 95} Miller, J. A., \& Roberts, D. A.

743: %1995, \apj, 452, 912

744: 

745: \bibitem[Miller and Ramaty (1987)] {Miller87} Miller, J. and Ramaty, R. 1987,

746: Solar Physics 113, 195

747: 

748: \bibitem

749: [Miller \& Vi\~{n}as (1993)] {Miller93} Miller, J. A., \& Vi\~{n}as, A. F.\

750: 1993, 

751: \apj, 412, 386

752: 

753: \bibitem

754: [M\"{o}bius et al. (1980)] {Mobius80} M\"{o}bius, E., Hovestadt, D., Klecker,

755: B., 

756: \& Gloeckler, G.\ 1980, \apj, 238, 768

757: 

758: \bibitem

759: [M\"{o}bius et al. (1982)] 

760: {Mobius82} M\"{o}bius, E., Scholer, M., Hovestadt, D., 

761: Klecker, B., \& Gloeckler, G.\ 1982, \apj, 259, 397

762: 

763: \bibitem

764: [Paesold et~al. (2003)]{Paesold03} Paesold, G., Kallenbach, R., \& Benz, A. O.

765: 2003, \apj 582, 495 

766: 

767: 

768: \bibitem

769: [Park et al. (1997)] 

770: {Park97} Park, B. T., Petrosian, V., \& Schwartz, R. A.\ 

771: 1997, \apj, 489, 358

772: 

773: 

774: \bibitem

775: [Petrosian \& Donaghy (1999)]{Petrosian99}

776: {Petrosian, V., \& Donaghy, T. Q.\ 1999, \apj, 527, 945}

777: 

778: \bibitem

779: [Petrosian \& Liu (2004)] 

780: {Petrosian04} Petrosian, V., \& Liu, S.\ 2004, \apj, 

781: 610, 550 (PL04)

782: 

783: \bibitem

784: [Petrosian  (2008)] 

785: {Petrosian0} Petrosian, V.\ 2008, To appear in ISSI workshop proceedings on

786: Impulsive SEPs  (2005,2006)

787: 

788: \bibitem[Pryadko \& Petrosian 1997]{Pry97}Pryadko, J. M., \& Petrosian, V.\

789: 1997, \apj, 482, 774

790: 

791: \bibitem

792: [Ramaty (1979)] 

793: {Ramaty79} Ramaty, R.\ 1979, in Particle Acceleration Mechanisms in 

794: Astrophysics, eds. J. Arons, C. Max, \& C. McKee (New York: AIP) 135

795: 

796: \bibitem

797: [Ramaty \& Kozlovsky (1974)]{Ramaty74}Ramaty, R. \& Kozlovsky, B. 1974, \apj, 193,

798: 729

799: 

800: \bibitem

801: [Reames et al. (1994)] 

802: {Reames94} Reames, D. V., Meyer, J. P., \& von Rosenvinge, T.  T.\ 1994, \apjs, 90, 649

803: 

804: \bibitem

805: [Reames et al. (1997)] {Reames97} Reames, D. V., Barbier, L. M., von Rosenvinge, T. 

806: T., Mason, G. M., Mazur, J. E., \& Dwyer, J. R.\ 1997, \apj, 483, 515

807: 

808: 

809: \bibitem[Reames \& Ng (2004)] {Reames04} Reames, D. V., \& Ng, C. K.\ 2004,

810: \apj, 610, 510

811: 

812: \bibitem

813: [Reames et al. (1985)] 

814: {Reames85} Reames, D. V., von Rosenvinge, T. T., \& Lin, R. P.\ 1985, \apj, 292,

815: 716

816: 

817: \bibitem[Share \& Murphy (1998)]{Share98}Share, G. H., \& Murphy, R. J. 1998,

818: \apj, 508, 876

819: 

820: \bibitem[Temerin \&  Roth (1992)]{Temerin92}Temerin, M., \&  Roth, I. 1992,

821: \apjl,  391, L105

822: 

823: \bibitem[Zhang(1995)]{Zhang95}Zhang, T. X. 1995, \apj, 449, 916

824: 

825: \end{thebibliography}{}

826: 

827: \end{document}

828: 

829: 

830: