1: %% Beginning of file 'sample.tex'

2: %%

3: %% Modified 03 Jan 01

4: %%

5: %% This is a sample manuscript marked up using the

6: %% AASTeX v5.x LaTeX 2e macros.

7: 

8: %% The first piece of markup in an AASTeX v5.x document

9: %% is the \documentclass command. LaTeX will ignore

10: %% any data that comes before this command.

11: 

12: %% The command below calls the preprint style

13: %% which will produce a one-column, single-spaced document.

14: %% Examples of commands for other substyles follow. Use

15: %% whichever is most appropriate for your purposes.

16: 

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

18: 

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

20: 

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

22: 

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

24: 

25: %\documentclass[preprint2]{aastex}

26: 

27: %% If you want to create your own macros, you can do so

28: %% using \newcommand. Your macros should appear before

29: %% the \begin{document} command.

30: %%

31: %% If you are submitting to a journal that translates manuscripts

32: %% into SGML, you need to follow certain guidelines when preparing

33: %% your macros. See the AASTeX v5.x Author Guide

34: %% for information.

35: 

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

37: \newcommand{\myemail}{dastgeer@ucr.edu}

38: 

39: %% You can insert a short comment on the title page using the command below.

40: 

41: \slugcomment{ Preprint: {\it Astrophysical Journal}}

42: 

43: %% If you wish, you may supply running head information, although

44: %% this information may be modified by the editorial offices.

45: %% The left head contains a list of authors,

46: %% usually a maximum of three (otherwise use et al.).  The right

47: %% head is a modified title of up to roughly 44 characters.  Running heads

48: %% will not print in the manuscript style.

49: 

50: \shorttitle{Energy Cascades in Astrophysical Plasma}

51: \shortauthors{Dastgeer \& Zank}

52: 

53: %% This is the end of the preamble.  Indicate the beginning of the

54: %% paper itself with \begin{document}.

55: \newcommand{\itemfig}[1]{\item{\bf Fig (#1) }} 

56: \newcommand{\be}{\begin{equation}}

57: \newcommand{\ee}{\end{equation}}

58: \newcommand{\et}{{\it{et. al. }}}

59: \newcommand{\eqs}[2]{Eqs. (\ref{#1}) \& (\ref{#2})} 

60: \newcommand{\Eqs}[2]{Equations  (\ref{#1}) \& (\ref{#2})} 

61: \newcommand{\Eq}[1]{Equation (\ref{#1})} 

62: \newcommand{\eq}[1]{Eq. (\ref{#1})} 

63: \newcommand{\fig}[2]{Figs  (\ref{#1}) \& (\ref{#2})} 

64: \newcommand{\Fig}[1]{Fig. (\ref{#1})} 

65:  \newcommand{\eqa}{\begin{eqnarray}}

66: \newcommand{\eeq}{\end{eqnarray}}  

67: \newcommand{\eqsto}[2]{Eqs. (\ref{#1}) to (\ref{#2})} 

68: \newcommand{\Eqsto}[2]{Equations (\ref{#1}) to (\ref{#2})}

69: \newcommand{\sig}{\sigma_{\rm ex}}

70: \begin{document}

71: 

72: %% LaTeX will automatically break titles if they run longer than

73: %% one line. However, you may use \\ to force a line break if

74: %% you desire.

75: 

76: \title{Energy Cascades in a Partially Ionized Astrophysical Plasma}

77: 

78: 

79: %% Use \author, \affil, and the \and command to format

80: %% author and affiliation information.

81: %% Note that \email has replaced the old \authoremail command

82: %% from AASTeX v4.0. You can use \email to mark an email address

83: %% anywhere in the paper, not just in the front matter.

84: %% As in the title, you can use \\ to force line breaks.

85: 

86: \author{Dastgeer Shaikh\footnote{Electronic mail : {\tt dastgeer@ucr.edu}} 

87:   \and 

88: Gary P. Zank\footnote{Electronic mail : {\tt zank@ucr.edu }}}

89: \affil{Institute of Geophysics and Planetary Physics (IGPP),\\

90: University of California, Riverside, CA 92521. USA.}

91: 

92: 

93: 

94: 

95: 

96: 

97: %% Notice that each of these authors has alternate affiliations, which

98: %% are identified by the \altaffilmark after each name.  Specify alternate

99: %% affiliation information with \altaffiltext, with one command per each

100: %% affiliation.

101: 

102: 

103: 

104: 

105: %% Mark off your abstract in the ``abstract'' environment. In the manuscript

106: %% style, abstract will output a Received/Accepted line after the

107: %% title and affiliation information. No date will appear since the author

108: %% does not have this information. The dates will be filled in by the

109: %% editorial office after submission.

110: 

111: \begin{abstract}

112: A local turbulence model is developed to study energy cascades in the

113: interstellar medium (ISM) based on self-consistent two-dimensional

114: fluid simulations. The model describes a partially ionized

115: magnetofluid interstellar medium (ISM) that couples a neutral hydrogen

116: fluid with a plasma primarily through charge exchange interactions.

117: Charge exchange interactions are ubiquitous in warm ISM plasma and the

118: strength of the interaction depends largely on the relative speed

119: between the plasma and the neutral fluid. Unlike small length-scale

120: linear collisional dissipation in a single fluid, charge exchange

121: processes introduce channels that can be effective on a variety of

122: length-scales that depend on the neutral and plasma densities,

123: temperature, relative velocities, charge exchange cross section and

124: the characteristic length scales. We find, from scaling arguments and

125: nonlinear coupled fluid simulations, that charge exchange interactions

126: modify spectral transfer associated with large scale energy containing

127: eddies. Consequently, the warm ISM turbulent cascade rate prolongs

128: spectral transfer amongst inertial range turbulent modes. Turbulent

129: spectra associated with the neutral and plasma ISM fluids are

130: therefore steeper than those predicted by Kolmogorov's phenomenology.

131: 

132: 

133: \end{abstract}

134: 

135: %% Keywords should appear after the \end{abstract} command. The uncommented

136: %% example has been keyed in ApJ style. See the instructions to authors

137: %% for the journal to which you are submitting your paper to determine

138: %% what keyword punctuation is appropriate.

139: 

140: \keywords{magnetohydrodynamics: MHD-turbulence-methods:

141: numerical-(Sun:) solar wind-ISM: kinematics and

142: dynamics-interplanetary medium}

143: 

144: 

145: %% From the front matter, we move on to the body of the paper.

146: %% In the first two sections, notice the use of the natbib \citep

147: %% and \citet commands to identify citations.  The citations are

148: %% tied to the reference list via symbolic KEYs. The KEY corresponds

149: %% to the KEY in the \bibitem in the reference list below. We have

150: %% chosen the first three characters of the first author's name plus

151: %% the last two numeral of the year of publication as our KEY for

152: %% each reference.

153: 

154: 

155: \section{Introduction}

156: 

157: The local interstellar medium (LISM) surrounding the heliosphere is

158: warm ($\sim 7000^o K$) and consequently partially ionized. Although

159: the ionization fraction of the LISM is not conclusively established

160: \citep{slavin2006}[see also \citep{muller2006,zank2006}], it is

161: possible that the plasma number density is as small as $0.05 cm^{-3}$

162: and the neutral atomic H number density as much as $0.16

163: cm^{-3}$. Indeed, much of the interstellar medium is warm and

164: partially ionized. In the LISM, the low density plasma and neutral H

165: gas are coupled primarily through the process of charge exchange. On

166: sufficiently large temporal and spatial scales, a partially ionized

167: plasma is typically regarded as equilibrated; this is the case for the

168: LISM. However, the region between the bow shock of our heliosphere and

169: the heliopause \citep{zank1999} is not equilibrated because the charge

170: exchange mean free path (mfp) and the size of the region are

171: comparable. The neutral H distribution in the region is complex

172: \citep{pauls1995,zank1996,zank1996b,jacob2006} possessing an

173: approximately Maxwellian core that is broadened by hot H atoms produced

174: in the inner heliosheath \citep{zank1996b} and fast neutral H produced

175: in the supersonic solar wind \citep{zank1996b,zank1999}. Similarly,

176: the atmospheres of many stars, especially cooler stars, that are

177: embedded in a partially ionized interstellar medium will have outer

178: astrosheaths that are partially ionized. Several examples of such

179: partially ionized astropsheres have been observed using Lyman-$\alpha$

180: absorption measurements \citep{gayley1996, wood2000}.

181: 

182: 

183: Turbulent fluctuations in a partially ionized LISM are not only

184: potentially important in the context of the global heliospheric ISM

185: interaction, but are also instrumental to our understanding of many

186: astrophysical phenomena including the energization and transport of

187: cosmic rays, gamma-ray bursts, ISM density spectra etc. Besides the

188: ISM, a partially ionized plasma environment represent an important

189: component of the Saturnion magnetosphere, especially that part

190: influenced by mass loss from the moon Encephalus \citep{ip1997}, in

191: Neptune's magnetosphere e.g., \citep{hill1990}, and also in the

192: astrosphere of stars embedded in a cloud of neutral H or even some

193: stellar winds. For magnetosphere, the partially ionized plasma

194: sometimes comprises a mixture of water group and O neutrals and pickup

195: ions and solar wind plasma.  The partially ionized plasma environment

196: is also relevant to the study of neutron stars \citep{Potekhin2005}

197: where the plasma and strong magnetic field interaction still remains

198: an unresolved issue.

199: 

200: 

201:  The interaction of a neutral gas and plasma can be dated back to the

202:  seminal work by Kulsrud \& Pearce (1969) in the context of cosmic ray

203:  propagation, where it was shown that neutral component damps Alfv\'en

204:  waves. Neutrals interacting with plasma via a relative drag process

205:  results in ambipolar diffusion. Ambipolar diffusion plays a crucial

206:  role in the dynamical evolution of the near solar atmosphere,

207:  interstellar medium, and molecular clouds and star formation. For

208:  instance, Oishi \& Mac Low (2006) investigated the inability of

209:  ambipolar diffusion to set a characteristic mass scale in molecular

210:  clouds and find that substantial structure persists below the

211:  ambipolar diffusion scale because of the propagation of compressive

212:  slow mode MHD waves at smaller scales. It was further shown that the

213:  spectral behaviour is not influenced by the ambipolar diffusive

214:  process unlike viscous damping. Leake et al (2005) showed that the

215:  lower chromosphere contains neutral atoms, the existence of which

216:  greatly increases the efficiency of wave damping due to collisional

217:  friction momentum transfer.  They find that Alfv\'en waves with

218:  frequencies above 0.6Hz are completely damped and frequencies below

219:  0.01 Hz are unaffected. Khodachenko et al (2006) undertook a

220:  quantitative comparative study of the efficiency of the role of

221:  (ion-neutral) collisional friction, viscous and thermal conductivity

222:  mechanisms in damping MHD waves in different parts of the solar

223:  atmosphere.  It was pointed out by the authors that a correct

224:  description of MHD wave damping requires the consideration of all

225:  energy dissipation mechanisms through the inclusion of the

226:  appropriate terms in the generalized Ohm’s law, the momentum, energy

227:  and induction equations.  Molecular clouds are known to be supported,

228:  at least in part, by magnetic fields. The removal of magnetic fields

229:  thus represents an important component of the star formation

230:  process. In the most studied scenario, field removal occurs through

231:  the action of ambipolar diffusion, wherein magnetic fields are tied

232:  to the ionized component, which drifts relative to the more dominant

233:  neutral component of the gas (Mestel \& Spitzer 1956, Mouschovias

234:  1976, Nakano 1979, Shu 1983, Nakano 1984, Lizano \& Shu 1989, Basu \&

235:  Mouschovias 1994).  The role of magnetic fields and ion-neutral

236:  friction in regulating gravitationally driven fragmentation of

237:  molecular clouds was studied by Kudoh et al (2007).  Mestel \&

238:  Spitzer (1956) pointed out that even if clouds are magnetically

239:  supported, ambipolar diffusion (resulting from ion-neutral drag) will

240:  cause the support to be lost and stars to form.  Li \& Nakamura

241:  (2004) and Nakamura \& Li (2005) have studied turbulent fragmentation

242:  processes for a magnetized sheet including the effect of ion-neutral

243:  friction. They find that a mildly subcritical cloud can undergo

244:  locally rapid ambipolar diffusion and form multiple fragments because

245:  of an initial large scale highly supersonic compression wave.  Padoan

246:  et al (2000) calculated frictional heating by ion-neutral (or

247:  ambipolar) drift in turbulent magnetized molecular clouds and showed

248:  that the ambipolar heating rate per unit volume depends on field

249:  strength for constant rms Mach number of the flow, and on the

250:  Alfv\'enic Mach number. Furthermore, ion-neutral friction, has long

251:  been thought to be an important energy dissipation mechanism in

252:  molecular clouds, and therefore a significant heating mechanism for

253:  molecular cloud gas (Scalo 1977, Goldsmith \& Langer 1978, Zweibel \&

254:  Josafatsson 1983, Elmegreen 1985). These are only a few, amongst

255:  numerous other, studies that illustrate the importance of a neutral

256:  gas component in the dynamics of partially ionized magnetized

257:  astrophysical plasmas.

258: 

259: 

260: On even smaller scales, the edge region of a tokamak (a donut shaped

261: toroidal experimental device designed to achieve thermonuclear fusion

262: reaction in an extremely hot plasma) is partially ionized. In a

263: tokmak, the neutral particles result from effects such as gas puffing,

264: impurity injection, recombination, charge exchange, and possibly

265: neutral beam injection processes. The presence of neutrals can

266: potentially alter the dynamics of zonal flows and cross-field

267: diffusivity. For instance, several experiments have demonstrated that

268: the transition from L-mode (low confinement) to H-mode (high

269: confinement) can be significantly affected by neutral atoms in the

270: edge of a tokamak plasma \citep{singh2004}. The edge region

271: predominantly consists of neutral species resulting from recycling

272: from the wall, and from the limiter and/or divertor plate.  These

273: neutrals are present in significant numbers (e.g. $10^{10}- 10^{11}~

274: cm^{-3}$, which is about 5-8\% of the plasma density near the edge)

275: and affects the poloidal momentum balance and the L-H transition

276: threshold.  Thus they a play crucial role in regulating transport

277: processes in tokamak fusion plasmas.  The neutral gas can interact

278: with the magnetized plasma via charge exchange and other effects.

279: Furthermore, Doublet III-D tokamak (DIII-D) data shows that there is a

280: significant correlation between the power threshold for the L-H

281: transition and the poloidally averaged neutral density at the 95\%

282: flux surface \citep{groebner2001}. To understand the role of neutral

283: gas on H-mode physics, asymmetric gas puffing experiments have also

284: been performed in COMPASS-D and MAST devices

285: \citep{valovie2001,valovie2002}.

286: 

287: 

288: 

289: 

290: In all these environments, turbulence is ubiquitous. Turbulence

291: involves the nonlinear coupling and transfer of energy across a

292: multitude of spatial and temporal scales. The coupling of plasma to a

293: neutral gas via charge exchange, especially in a non-equilibrated

294: environment, introduces both a length scale distinct from the usual

295: collisional mfp and alternate channels and mode couplings for the

296: transfer of energy. Certainly, in the local ISM, the charge exchange

297: cross-section of e.g. neutral hydrogen atoms is larger than the

298: collisional cross-section (e.g. Zank 1999), implying that the former

299: governs much of the small-scale physics. Fluctuations with length

300: scales that exceed the charge exchange mean free path $\ell_{ce}$ will

301: obviously be effected by the neutral gas, such as the damping of

302: linear modes (Shaikh \& Zank 2006), but it is less clear how turbulent

303: plasma motion and nonlinear couplings will be mediated. For plasma

304: fluctuations on scales smaller than $\ell_{ce}$, the role of neutral

305: gas is even less clear.

306: 

307: Further, it is worth mentioning that the charge exchange process in

308: its simpler (and leading order) form can be treated like a friction or

309: viscous drag term in the fluid momentum equation, describing the

310: relative difference in the ion and neutral fluid velocities. The drag

311: imparted in this manner by a collision between ion and neutral also

312: causes ambipolar diffusion, a mechanism used to describe the Alfv\'en

313: wave damping by cosmic rays (Kulsrud \& Pearce 1969) and also

314: discussed by Oishi \& Mac Low (2006) in the context of molecular

315: clouds. The thermodynamical properties of the latter are significantly

316: different from that found in the heliosheath region. Interestingly,

317: unlike linear damping, ambipolar diffusion does not terminate an

318: isothermal MHD turbulence cascade (Oishi \& Mac Low 2006). Instead, it

319: damps some linear MHD waves. In the context of MHD turbulence, this

320: issue had also been addressed by Cho \& Lazarian (2003) who discussed

321: the viscosity-damped effects on inertial subrange associated with

322: magnetic field fluctuations in compressible (isothermal) MHD

323: turbulence. Ion-neutral friction nonetheless describes a leading order

324: interaction between the ion and neutral fluid species where thermal

325: speeds associated with both the fluids are of little significance or

326: ignorable and plasma phenomena are describable predominantly by

327: isothermal processes. This description is not appropriate to describe

328: the outer heliosheath plasma where proton as well as neutral hydrogen

329: temperatures can be as high as $20,000~^oK$. The thermal speeds

330: associated with the heliosheath protons and neutrals corresponding to

331: such a high temperature cannot therefore be ignored for the following

332: reasons. Firstly, since the thermal speed describes a significant

333: fraction of the energy and is proportional to the relative speeds in

334: the momentum and energy charge exchange sources, it should not be

335: neglected for plasmas that are not cold. It can potentially influence

336: the momentum and energy sources during the course of evolution. This

337: means that the interaction between the heliosheath protons and neutral

338: atoms is not mediated only by the corresponding bulk fluid speeds, but

339: it also depends critically on their thermal speeds.  Secondly, for

340: elastic and charge exchange collisions, the initial particle

341: distribution can be well described by Maxwellians even with different

342: streaming velocities and temperatures. Higher order effects are

343: introduced in the dynamics because the distributions are not really

344: Maxwellians due to prior collisions as pointed out by McNutt et

345: al. (1998).  The latter introduces Navier-Stokes-like corrections

346: (viscosity and thermal conductivity like terms) to the neutral

347: component of the equation. A consistent treatment of such processes,

348: describing the higher order corrections, in the context of the

349: heliosheath requires that the Maxwellians be multiplied by transfer

350: integrals, as done in Pauls et al (1995).  Thirdly, plasma and neutral

351: fluids in the heliosheath region are compressible enough to possess a

352: finite component of thermal energy associated with a non-adiabatic

353: exchange of energy amongst the local fluctuations.  Thus in the

354: context of supersonic heliosheath turbulent fluctuations, the

355: characteristic speed of plasma-neutral coupled fluid needs to be

356: treated on equal footing with the thermal speed. Consequently, thermal

357: corrections appear as higher order terms along with the leading order

358: ambipolar-like diffusive forces in the charge exchange expressions as

359: described in section 3.  This will be evident from the full energy

360: equation and the subsequent charge exchange energy transfer functions

361: employed in our model.

362: 

363: 

364: Multiple processes can couple the interstellar medium plasma and

365: neutral gas. Besides charge exchange scattering due to an induced

366: dipole moment in the neutral atom, depending on the temperature

367: (Banks, 1966), is possible. Interestingly enough, variations in

368: cross-sections can also be expected for different ions in the same

369: neutral gas at higher temperatures (Banks, 1966).  Further processes,

370: especially in the context of heliospheric processes, are

371: photoionization, electron impact ionization, recombination, and

372: H-H$^+$ Coulomb collisions (see e.g. Zank, 1999).  However,

373: cross-section associated with most of them is small compared to that

374: of charge exchange between hydrogen atom and ions in the heliosheath

375: region. We therefore do not incorporate them in our current model.

376: 

377: 

378: In view of these comments, we have developed a self-consistent

379: plasma-neutral ISM turbulence simulation model based on fluid modeling

380: of neutral and plasma fluids \citep{zank1999,pauls1995,zank1996b}. Our

381: model treats the plasma and neutral species as distinguishable fluids

382: and couples them through charge exchange.  The model essentially

383: employs a set of magnetohydrodynamic equations that describe the

384: turbulent plasma motions, whereas the neutrals are described by

385: hydrodynamic equations. A fluid approach for the local interstellar

386: medium is justifiable since the charge exchange and collisional mean

387: free path ensures approximate thermodynamical equipartition between

388: plasma and neutral species with the result that both plasma and

389: neutrals can be described by a Maxwellian distribution. Besides the

390: local ISM, the outer heliosheath can also be described reasonably well

391: by a fluid model of this kind, as was done in the model of

392: \citet{pauls1995}. For a non-equilibrated description, a more

393: elaborate multi-fluid model can be developed, as was done by

394: \citet{zank1996b,zank1996}, and this could be applied to

395: straightforwardly to both the outer heliosheath and very local

396: ISM. This would be more important to the outer heliosheath since a

397: multi-fluid model would then include hot inner heliosheath neutral

398: atoms deposited through secondary charge exchange in the outer

399: heliosheath \citep{zank1996b,zank1996,zank1999}.  The schematic in

400: \Fig{fig0} depicts the simulation regions in which we are particularly

401: interested i.e. the outer heliosheath and the local interstellar

402: medium. The importance of ISM turbulence is clearly evident from this

403: figure in the context of cosmic ray modulation, large-scale

404: heliospheric structure and particle acceleration.

405: 

406: In this paper, we focus on the multi-scale evolution of a partially

407: ionized plasma, comprising multiple fluids, where each fluid evolves

408: under the influence of the other through a complex interaction

409: process.  We investigate the energy cascade between LISM

410: turbulent-fluctuations in a partially ionized plasma where the plasma

411: and neutral gas are coupled through charge exchange. Here, we, present

412: a numerical simulation of ISM turbulence that self-consistently

413: evolves both the plasma and neutral fluids when coupled by charge

414: exchange. We restrict our attention to a neutral hydrogen (H) gas, so

415: that plasma and neutral densities are conserved by charge exchange

416: interactions.  Hence the corresponding continuity equations do not

417: include charge exchange terms.  One of the points that emerges from

418: our coupled multi fluid LISM simulations is that neutrals enhance

419: turbulent cascade rates and lead to much steeper spectra compared to

420: that predicted by the usual Kolmogorov theory \citep{k41} for pure MHD

421: or hydrodynamics \citep{shaikh2007}.  In Section 2, we discuss the

422: equations of a coupled plasma-neutral ISM turbulence model, their

423: validity, the underlying assumptions and the normalizations. Section 3

424: deals with a quantitative derivation of charge exchange source terms,

425: associated with the plasma and neutral fluids, based on the

426: time-dependent Boltzmann equation. Section 4 describes the results of

427: our nonlinear, coupled, self-consistent ISM fluid simulations. It is

428: shown, in section 5, that the coupling of plasma and neutral fluids

429: leads to an enhanced spectral transfer in the inertial range, which

430: results in a steeper ISM turbulent spectrum than predicted by the

431: Kolmogorov theory.  Finally, conclusions are presented in Section 6.

432: 

433: 

434: \section{Model Equations for a partially ionized  LISM}

435: The assumptions that are intrinsic to our model of turbulence in a

436: partially ionized ISM are the following.  (i) Fluctuations in the

437: plasma and neutral fluids are isotropic, homogeneous, thermally

438: equilibrated and turbulent, and (ii) Neither a mean magnetic field nor

439: velocity flows are present initially. Local mean flows may

440: subsequently be generated by self-consistently excited nonlinear

441: instabilities. (iii) The characteristic turbulent correlation

442: length-scales ($\lambda_c \sim 1/k_c$) are typically bigger than

443: charge-exchange mean free path lengths ($\lambda_{ce}\sim 1/k_{ce}$)

444: in the ISM flows, i.e $\lambda_c\gg \lambda_{ce}$ or $k_{ce}/k_c \gg

445: 1$. The latter inequality is also consistent with

446: \citet{Florinski2003,Florinski2005}.  Nevertheless, they are large

447: enough to treat any localized shocks as smooth discontinuities. In

448: other words, the characteristic shock length-scales are small compared

449: to the ISM turbulent fluctuation length-scales, and finally (iv)

450: boundary conditions are periodic, essentially a box of interstellar

451: plasma, which is a natural and most appropriate choice for modeling

452: turbulence in the local ISM.

453: 

454: While most of the above assumptions are appropriate to realistic ISM

455: turbulent flows, so allowing us to use MHD and hydrodynamic

456: descriptions for the plasma and the neutral components respectively,

457: the latter has been criticized in the context of heliospheric flows

458: within the heliopause owing to the large charge-exchange mean free

459: path (about the size or even bigger than the extent of the

460: heliosphere).  The use of a fluid description for neutrals in the

461: inner heliosheath is therefore inappropriate.  In the LISM and outer

462: heliosheath, we are nonetheless not restricted by a large mfp because

463: the plasma and neutral fluid remain close to thermal equilibirium and

464: behave as Maxwellian fluids.  Our model simulates the local ISM and

465: outer heliosheath. The fluid model describing nonlinear turbulent

466: processes in the interstellar medium, in the presence of charge

467: exchange, can be cast into plasma density ($\rho_p$), velocity (${\bf

468: U}_p$), magnetic field (${\bf B}$), pressure ($P_p$) components

469: according to the conservative form

470: \be

471: \label{mhd}

472:  \frac{\partial {\bf F}_p}{\partial t} + \nabla \cdot {\bf Q}_p={\cal Q}_{p,n},

473: \ee

474: where,

475: \[{\bf F}_p=

476: \left[ 

477: \begin{array}{c}

478: \rho_p  \\

479: \rho_p {\bf U}_p  \\

480: {\bf B} \\

481: e_p

482:   \end{array}

483: \right], 

484: {\bf Q}_p=

485: \left[ 

486: \begin{array}{c}

487: \rho_p {\bf U}_p  \\

488: \rho_p {\bf U}_p {\bf U}_p+ \frac{P_p}{\gamma-1}+\frac{B^2}{8\pi}-{\bf B}{\bf B} \\

489: {\bf U}_p{\bf B} -{\bf B}{\bf U}_p\\

490: e_p{\bf U}_p

491: -{\bf B}({\bf U}_p \cdot {\bf B})

492:   \end{array}

493: \right],\\

494: {\cal Q}_{p,n}=

495: \left[ 

496: \begin{array}{c}

497: 0  \\

498: {\bf Q}_M({\bf U}_p,{\bf V}_n, \rho_p, \rho_n, T_n, T_p)   \\

499: 0 \\

500: Q_E({\bf U}_p,{\bf V}_n,\rho_p, \rho_n, T_n, T_p)

501:   \end{array}

502: \right]

503: \] 

504: and

505: \[ e_p=\frac{1}{2}\rho_p U_p^2 + \frac{P_p}{\gamma-1}+\frac{B^2}{8\pi}.\]

506: The above set of plasma equations is supplemented by $\nabla \cdot {\bf

507: B}=0$ and is coupled self-consistently to the ISM neutral density

508: ($\rho_n$), velocity (${\bf V}_n$) and pressure ($P_n$) through a set

509: of hydrodynamic fluid equations,

510: \be

511: \label{hd}

512:  \frac{\partial {\bf F}_n}{\partial t} + \nabla \cdot {\bf Q}_n={\cal Q}_{n,p},

513: \ee

514: where,

515: \[{\bf F}_n=

516: \left[ 

517: \begin{array}{c}

518: \rho_n  \\

519: \rho_n {\bf V}_n  \\

520: e_n

521:   \end{array}

522: \right], 

523: {\bf Q}_n=

524: \left[ 

525: \begin{array}{c}

526: \rho_n {\bf V}_n  \\

527: \rho_n {\bf V}_n {\bf V}_n+ \frac{P_n}{\gamma-1} \\

528: e_n{\bf V}_n

529:   \end{array}

530: \right],\\

531: {\cal Q}_{n,p}=

532: \left[ 

533: \begin{array}{c}

534: 0  \\

535: {\bf Q}_M({\bf V}_n,{\bf U}_p, \rho_p, \rho_n, T_n, T_p)   \\

536: Q_E({\bf V}_n,{\bf U}_p,\rho_p, \rho_n, T_n, T_p)

537:   \end{array}

538: \right],

539: \] 

540: \[e_n= \frac{1}{2}\rho_n V_n^2 + \frac{P_n}{\gamma-1}.\]

541: 

542: Equations (\ref{mhd}) to (\ref{hd}) form an entirely self-consistent

543: description of the coupled ISM plasma-neutral turbulent fluid.

544: Several points are worth noting.  The charge-exchange momentum sources

545: in the plasma and the neutral fluids, i.e. Eqs. (\ref{mhd}) and

546: (\ref{hd}), are described respectively by terms ${\bf Q}_M({\bf

547:   U}_p,{\bf V}_n,\rho_p, \rho_n, T_n, T_p)$ and ${\bf Q}_M({\bf

548:   V}_n,{\bf U}_p,\rho_p, \rho_n, T_n, T_p)$. A swapping of the plasma

549: and the neutral fluid velocities in this representation corresponds,

550: for instance, to momentum changes (i.e. gain or loss) in the plasma

551: fluid as a result of charge exchange with the ISM neutral atoms

552: (i.e. ${\bf Q}_M({\bf U}_p,{\bf V}_n,\rho_p, \rho_n, T_n, T_p)$ in

553: Eq. (\ref{mhd})). Similarly, momentum change in the neutral fluid by

554: virtue of charge exchange with the plasma ions is indicated by ${\bf

555:   Q}_M({\bf V}_n,{\bf U}_p,\rho_p, \rho_n, T_n, T_p)$ in

556: Eq. (\ref{hd}). In the absence of charge exchange interactions, the

557: plasma and the neutral fluid are de-coupled trivially and behave as

558: ideal fluids.  While the charge-exchange interactions modify the

559: momentum and the energy of plasma and the neutral fluids, they

560: conserve density in both the fluids (since we neglect photoionization

561: and recombination). Nonetheless, the volume integrated energy and the

562: density of the entire coupled system will remain conserved in a

563: statistical manner. The conservation processes can however be altered

564: dramatically in the presence of any external forces. These can include

565: large-scale random driving of turbulence due to any external forces or

566: instabilities, supernova explosions, stellar winds, etc. Finally, the

567: magnetic field evolution is governed by the usual induction equation,

568: i.e. Eq. (\ref{mhd}), that obeys the frozen-in-field theorem unless

569: some nonlinear dissipative mechanism introduces small-scale damping.

570: 

571: 

572: The underlying ISM turbulence model can be non-dimensionalized

573: straightforwardly using a typical ISM scale-length ($\ell_0$), density

574: ($\rho_0$) and velocity ($v_0$). The normalized plasma density,

575: velocity, energy and the magnetic field are respectively;

576: $\bar{\rho}_p = \rho_p/\rho_0, \bar{\bf U}_p={\bf U}_p/v_0,

577: \bar{P}_p=P_p/\rho_0v_0^2, \bar{\bf B}={\bf B}/v_0\sqrt{\rho_0}$. The

578: corresponding neutral fluid quantities are $\bar{\rho}_n =

579: \rho_n/\rho_0, \bar{\bf U}_n={\bf U}_n/v_0,

580: \bar{P}_n=P_n/\rho_0v_0^2$. The momentum and the energy

581: charge-exchange terms, in the normalized form, are respectively

582: $\bar{\bf Q}_m={\bf Q}_m \ell_0/\rho_0v_0^2, \bar{Q}_e=Q_e

583: \ell_0/\rho_0v_0^3$. The non-dimensional temporal and spatial

584: length-scales are $\bar{t}=tv_0/\ell_0, \bar{\bf x}={\bf

585:   x}/\ell_0$. Note that we have removed bars from the set of

586: normalized coupled ISM model equations (\ref{mhd}) \& (\ref{hd}).  The

587: charge-exchange cross-section parameter ($\sigma$), which does not

588: appear directly in the above set of equations (see the subsequent

589: section for more detail), is normalized as $\bar{\sigma}=n_0 \ell_0

590: \sigma$, where the factor $n_0\ell_0$ has dimension of (area)$^{-1}$.

591: By defining $n_0, \ell_0$ through

592: $\sigma_{ce}=1/n_0\ell_0=k_{ce}^{-2}$, we see that there exists a

593: critical charge exchange wavenumber ($k_{ce}$) associated with the

594: coupled ISM plasma-neutral turbulent system.  For a characteristic

595: density, this corresponds physically to an area defined by the charge

596: exchange mode being equal to (mfp)$^2$.  Thus the larger the area, the

597: higher is the probability of charge exchange between plasma ions and

598: neutral atoms, as illustrated in Fig. (2).  To be precise, `$k_{ce}$'

599: is a (lengthscale)$^{-1}$ that typically helps us determine whether or

600: not a particular turbulent fluctuation length scale undergoes charge

601: exchange in our model. It is strictly in that sense, we refer to it as

602: a critical wavenumber. Clearly, there exist three different regimes

603: depending on whether (i) $k<k_{ce}$, (ii) $k \simeq k_{ce}$ or (iii)

604: $k>k_{ce}$. In the heliosheath, the probability that charge exchange

605: can directly modify those modes satisfying $k<k_{ce}$ is high compared

606: to modes satisfying $k>k_{ce}$.  Since the charge exchange

607: length-scales are much smaller than the ISM turbulent correlation

608: scales, this further allows many charge exchange interactions amongst

609: the nonlinear turbulent ISM modes before they cascade energy in one

610: unit convective time. This is illustrated schematically in Fig. (2).

611: 

612: It is interesting to compare our model with other work. Of particular

613: relevance is the classic work by Kulsrud and Pearce (1969) who

614: investigated the interaction of galactic comic rays and Alfi\'en waves

615: in the interstellar medium and also considered ion-neutral collision

616: as a damping mechanism for Alfv\'en waves.  Ion-neutral collisions are

617: also reported by McIvor (1977) to be a dominant mechanism for damping

618: interstellar medium turbulence.  The damping (due to ion-neutral

619: collisions) rates for Alfv\'en waves estimated by Kulsrud and Pearce

620: (1969) are based on $\nu_0 \simeq n_0 v_{th} \sigma$, where $\nu_0,

621: n_0, v_{th}$ and $\sigma$ are respectively the ion-neutral collision

622: frequency, neutral density, thermal speed and collision cross-section.

623: The damping wavenumber associated with this frequency can be estimated

624: to be $k\simeq n_0 v_{th} \sigma/V_A$. The latter determines

625: essentially the dissipation wavenumber associated with the damping of

626: small scale Alfv\'en waves in ISM turbulence (Kulsrud and Pearce

627: 1969). The turbulent cascade is further expected to be terminated

628: beyond this wavenumber. It is clear from this expression that the

629: higher the thermal speed, the larger the dissipation

630: wavenumber. Correspondingly in real space, dissipation will be

631: concentrated on smaller length-scales for higher thermal

632: speeds. Furthermore, the thermal speed is also proportional to the

633: temperature.  Thus, given the heliosheath parameters, $v_{Thu}$

634: associated with {\it heliosheath} neutral atoms will be at least two

635: orders of magnitude larger compared with that results from Eq. (47) in

636: Kulsrud and Pearce (1969).  This will correspond to a larger

637: dissipative wavenumber in the heliosheath region and essentially means

638: that charge exchange interactions are restricted to only those eddies

639: whose length-scales are comparable to the dissipative length-scales of

640: turbulence. It is not clear if this scenario holds in inner/outer

641: heliosheath turbulence where charge exchange interactions, in

642: principle, can modify any length scale in the inertial range.  Part of

643: the reason resides with the fact that charge exchange modifications

644: are proportional to the relative speed between ion and neutral fluids

645: and corresponding densities which can obviously alter the nonlinear

646: cascades by influencing the combined momenta and energies of both the

647: fluids in a self-consistent and subtle manner. Additionally, Kulsrud

648: and Pearce (1969), as does McIvor (1977), assumed a fixed neutral

649: density while computing the damping rates associated with Alfv\'en

650: waves in ISM turbulence. While a fixed neutral density can modify the

651: linear growth rates of the underlying waves, it does not itself evolve

652: and hence no back reaction on Alfv\'en waves and vice versa can be

653: expected.  By contrast, our model self-consistently evolves neutral

654: and plasma and describes the mutual feedback of one species on the

655: other.  This not only alters the linear interaction process, but it

656: also influences the nonlinear coupling in a subtle manner not readily

657: describable by means of any linear analytic theory.  A self-consistent

658: coupling is absolutely essential to understand nonlinear turbulent

659: cascades in the inner/outer heliosheath. It is to be noted further

660: that we do not attempt to investigate damping of heliosheath

661: turbulence by Alfv\'en waves that experience ion-neutral collisions,

662: unlike McIvor (1977).  While Alfv\'en waves are intrinsically present

663: in our model, turbulent cascades are least affected by them along the

664: mean magnetic field (Shebalin \& Montgomery 1983). Alfv\'en waves can

665: be crucial in providing a local anisotropy in the spectral transfer

666: due to a mean or local field though.  However, our prime focus here is

667: to develop an understanding of heliosheath turbulence where plasma and

668: neutral fluids evolve through a mutual interaction mediated by charge

669: exchange. As evident from the complexity associated with these

670: interactions, nonlinear turbulent cascades are likely to be affected

671: at any length scale in the heliosheath inertial range, unlike a linear

672: dissipative processes.

673: 

674: 

675: In the subsequent section, we derive an exact quantitative form of the

676: sources due to charge exchange in ISM turbulence.

677: 

678: 

679: 

680: \section{Charge Exchange Sources}

681: The charge exchange terms can be obtained from the Boltzmann transport

682: equation that describes the evolution of a neutral distribution

683: function $f_n=f({\bf x}, v_x, v_y, v_z, t)$ in a six-dimensional phase

684: space defined respectively by position and velocity vectors $({\bf x},

685: v_x, v_y, v_z)$ at each time $t$. Here we follow \citet{pauls1995} in

686: computing the charge exchange terms from various moments of the

687: Boltzmann equation. The Boltzmann equation for the neutral

688: distribution contains a source term proportional to the proton

689: distribution function $f_p$ and a loss term proportional to the

690: neutral distribution function $f_n$, \eqa

691: \label{bol}

692: \frac{\partial f_n}{\partial t} + {\bf v}_n \cdot \nabla f_n +

693: \frac{\bf F}{m} \cdot \nabla_{{\bf v}_n} f_n = 

694: f_{p}({\bf x}, v_x, v_y, v_z, t) \int f_n({\bf x}, v_x, v_y, v_z, t) |{ v}_n- {u}_p|

695: \sig(v_{rel}) d^3{v}_n \nonumber \\ 

696: -f_n({\bf x}, v_x, v_y, v_z, t) \int

697: f_p({\bf x}, v_x, v_y, v_z, t) |{ u}_p- { v}_n| \sig(v_{rel})

698: d^3{ u}_p.  \eeq Here $f_p, ~{u}_p$ represent respectively the

699: proton distribution function and velocity.  $\sig$ is the charge

700: exchange cross-section (between neutrals and plasma protons), $m$ is

701: the mass of particle, and ${\bf F}$ represents forces acting on the

702: fluid.  The charge exchange parameter has a logarithmically weak

703: dependence on the relative speed ($v_{rel}=|{u}_p- {v}_n|$) of

704: the neutrals and the protons through $\sig = [(2.1-0.092 \ln

705:   (v_{rel})) 10^{-7} cm]^2$ \citep{fite}.  This cross-section is valid

706: as long as energy does not exceed $1eV$, which usually is the case in

707: the inner/outer heliosphere. Beyond $1eV$ energy, this cross-section

708: yields a higher neutral density. This issue is not applicable to our

709: model and hence we will not consider it here.  The density, momentum,

710: and energy of the thermally equilibrated Maxwellian proton and neutral

711: fluids can be computed from \eq{bol} by using the zeroth, first and

712: second moments $\int f_{\xi} d^3\xi, \int m{\bf \xi} f_{\xi} d^3\xi$

713: and $\int m\xi^2/2 f_{\xi} d^3\xi$ respectively, where $\xi={u}_p$

714: or ${v}_n$. Since charge exchange conserves the density of the

715: proton and neutral fluids, there are no sources in the corresponding

716: continuity equations. We, therefore, need not compute the zeroth

717: moment of the distribution function.  Computing directly the first

718: moment from \eq{bol}, we obtain the neutral fluid momentum equation as

719: given by \eq{hd}.  The entire rhs of \eq{bol} can now be replaced by a

720: momentum transfer function ${\bf Q}_M({v}_n,{u}_p) $ which

721: reads \be

722: \label{qm}

723:  {\bf Q}_M({v}_n,{u}_p) = \bar{\mu}({u}_p,{v}_n)-\bar{\mu}({v}_n,{u}_p),

724: \ee

725: where ${\bf Q}_M$ and $ \bar{\mu}$, the transfer integral, are 

726: vector quantities. The transfer integrals describe  the

727: charge exchange transfer of momentum from proton to neutral fluid and

728: vice versa.  The first term on the rhs of

729: Eq. (4) can be expressed by

730: \[\bar{\mu}({u}_p,{v}_n) = f_p({\bf x}, v_x,v_y,v_z,t)\beta({u}_p,{v}_n)\]

731: where,

732: \[\beta({u}_p,{v}_n)=\int f_n({\bf x}, v_x,v_y,v_z,t) |{v}_n- {u}_p|

733: \sig(v_{rel}) d^3{v}_n.\]

734: Considering a Maxwellian distribution for the neutral atoms, we simplify 

735: $\beta({u}_p,{v}_n)$ as

736: follows,

737: \[\beta({u}_p,{v}_n)= \sig n_n V_{T_n} 

738: \sqrt{\frac{4}{\pi}+\frac{({v}_n- {u}_p)^2}{V_{T_n}^2}}.\] Note that

739: the above expression emerges directly from a straightforward

740: integration of sources in the rhs of the Boltzmann Eq. (3). To obtain

741: the expression for the momentum transferred from proton to neutral (or

742: vice versa), we need to take a second moment of the

743: $\bar{\mu}({u}_p,{v}_n)$ expression. This is shown in the following,

744: \[\bar{\mu}({u}_p,{v}_n) = m {\bf v}_n I_0({u}_p,{v}_n) + m ({\bf u}_p-{\bf v}_n) I_1({u}_p,{v}_n).\]

745: where $I_0$ and $I_1$ are transfer integrals that can be written as follows,

746: \[I_0({u}_p,{v}_n)= \int  f_p({\bf x}, v_x,v_y,v_z,t)\beta({u}_p,{v}_n) ~d^3{u}_p;\]

747: \[I_1({u}_p,{v}_n)= \int {v}_n f_p({\bf x}, v_x,v_y,v_z,t)\beta({u}_p,{v}_n) ~d^3{u}_p.\]

748: Assuming a Maxwellian distribution for plasma protons and using

749: $\beta({u}_p,{v}_n)$ from the above expression, we can

750: straightforwardly evaluate the transfer integrals $I_0$ and $I_1$ (see

751: Pauls et al (1995) for details).  We further write the

752: complete form of the first term on the rhs of Eq. (4) as follows,

753: \[\bar{\mu}({u}_p,{v}_n) = m\sig n_p n_n \left[U_{{u}_p,{v}_n}^\ast{\bf v}_n - ({\bf u}_p-{\bf v}_n)

754: \frac{  V_{T_n}^2}{\delta V_{{u}_p,{v}_n}} \right].\] In a similar

755: manner, we can evaluate the second term on the rhs of Eq. (4), which

756: yields the following form,

757: \[\bar{\mu}({v}_n,{u}_p) = m\sig n_{p}n_n \left[U_{{v}_n,{u}_p}^\ast {\bf u}_p - ({\bf v}_n-{\bf u}_p) \frac{V_{T_p}^2}{\delta V_{{v}_n,{u}_p}} \right],\]

758: where $\delta V_{{u}_p,{v}_n} =

759:   \left[4\left(\frac{4}{\pi}V_{T_p}^2+\Delta U^2 \right)

760:   +\frac{9\pi}{4}V_{T_n}^2 \right]^{1/2}, \delta V_{{v}_n,{u}_p} = \left[4\left(\frac{4}{\pi}V_{T_n}^2+\Delta U^2 \right)

761:   +\frac{9\pi}{4}V_{T_{n}}^2 \right]^{1/2}$ and $U^\ast=U_{{u}_p,{v}_n}^\ast = U_{{v}_n,{u}_p}^\ast =

762:   \left[\frac{4}{\pi}V_{T_{p}}^2+\frac{4}{\pi}V_{T_{n}}^2 +\Delta

763:   U^2\right]^{1/2}, \Delta {u} = {u}_p-{v}_n$ \citep{pauls1995}.  On

764:   substituting these expressions in the momentum transfer function, we

765:   obtain 

766: \be

767: \label{qm2}

768:  {\bf Q}_M({v}_n,{u}_p) = m\sig n_{p}n_n ({\bf v}_n -{\bf

769:    u}_p) \left[ U^\ast + \frac{V_{T_n}^2}{\delta V_{{u}_p,{v}_n}} -\frac{V_{T_p}^2}{\delta V_{{v}_n,{u}_p}}

770:    \right].  

771: \ee 

772: \Eq{qm2} together with \eq{hd} yields the momentum equation for

773: the neutral gas. Swapping the plasma and neutral fluid velocities

774: yields the corresponding source term for the proton fluid momentum

775: equation.  The gain or the loss in neutral or proton fluid momentum

776: depends upon the charge exchange sources, which depend upon the

777: relative speeds between neutrals and the protons. The thermal speeds

778: of proton and neutral gas in \eq{qm2} are given respectively by $

779: V_{T_p}^2 = 2K_BT_{p}/m$ (the factor 2 arises because of thermal

780: equilibration in that it is assumed that the temperature of the plasma

781: electrons and protons are nearly identical so that $T_p = T_e + T_{\rm

782:   proton} \simeq 2T_p$) and $V_{T_n}^2 = K_BT_{n}/m$. The

783: corresponding temperatures are related to the pressures by $

784: P_{p}=2n_{p}K_BT_{p}$ and $P_{n}=n_{n}K_BT_{n}$, where $n_n, T_n,

785: n_{p}, T_{p}$ are respectively the neutral and plasma density and the

786: temperature, and $K_B$ is the Boltzmann constant.

787: 

788: 

789: The  moment, $\int m\xi^2/2 f_{\xi} d^3\xi$, of the Boltzmann

790: \eq{bol} yields an energy equation for the neutral fluid whose

791: rhs contains the charge exchange energy transfer function 

792: \[Q_E({v}_n,{u}_p) = \eta({u}_p,{v}_n)-\eta({v}_n,{u}_p),\]

793: where $\eta({u}_p,{v}_n), ~\eta({v}_n,{u}_p)$ are the

794: energy transfer (from neutral to proton and vice versa) rates. These

795: functions can be computed as follows: 

796: \[\eta({u}_p,{v}_n) = \frac{1}{2}

797: mV_n^2 \sig n_{p}n_n U^\ast + \frac{3}{4}m V_{T_n}^2 \sig n_{p}n_n \Delta

798: V_{{u}_p,{v}_n}- m \sig n_pn_n {\bf v}_n\cdot ({\bf u}_p-{\bf

799: v}_n) \frac{V_{T_n}^2}{\delta V_{{u}_p,{v}_n}}\]

800:  and 

801: \[\eta({v}_n,{u}_p) = \frac{1}{2}mV_{n}^2 \sigma n_{p}n_n U^\ast +  \frac{3}{4}m

802: V_{T_{p}}^2 \sig n_{p}n_n \Delta V_{{v}_n,{u}_p}- m \sig

803: n_{p}n_n {\bf u}_p \cdot ({\bf v}_n-{\bf u}_p) \frac{V_{T_{p}^2}}{\delta

804: V_{{v}_n,{u}_p}}.\]  

805: The total energy transfer from neutral to proton fluid, due to charge

806: exchange, can then be written as,

807: \eqa

808: \label{qe}

809: Q_E({v}_n,{u}_p) = \frac{1}{2}m \sig  n_{p}n_n U^\ast (V_n^2-U_p^2)+

810: \frac{3}{4}m  \sig  n_{p}n_n (V_{T_{n}}^2\Delta V_{{u}_p,{v}_n}-V_{T_{p}}^2

811: \Delta V_{{v}_n,{u}_p}) 

812:  \nonumber \\

813: -m \sig  n_{p}n_n

814: \left[ {\bf v}_n \cdot ({\bf u}_p-{\bf v}_n)\frac{V_{T_n}^2}{\delta V_{{u}_p,{v}_n}}-

815: {\bf u}_p \cdot ({\bf v}_n-{\bf u}_p)\frac{V_{T_{p}^2}}{\delta V_{{v}_n,{u}_p}}\right],

816: \eeq

817: with $\Delta V_{{u}_p,{v}_n} = 

818: [\frac{4}{\pi}V_{T_p}^2+\frac{64}{9\pi}V_{T_n}^2 +\Delta U^2]^{1/2} ~{\rm and} ~

819: \Delta V_{{v}_n,{u}_p} = [\frac{4}{\pi}V_{T_n}^2+

820: \frac{64}{9\pi}V_{T_{p}}^2 +\Delta U^2]^{1/2}$.

821: A similar expression for the energy transfer charge exchange source

822: term of plasma energy in \eq{mhd} can be obtained by exchanging the

823: plasma and neutral fluid velocities. In the normalized momentum and

824: energy charge exchange source terms, the factor $m\sig$ in

825: \eqs{qm2}{qe} is simply replaced by $\bar{\sigma}$.

826: 

827: 

828: 

829: \section{ Nonlinear Simulation of Energy Cascades}

830: It is clear from previous sections that the interaction of the

831: neutrals with a plasma introduces characteristic length and time scales

832: that can separate characteristic time and length scales in ISM

833: turbulence. For instance, the plasma-neutral fluid coupling in the ISM

834: fluctuations introduces charge exchange modes, $k_{ce}$, that are

835: distinctively different from the characteristic turbulent mode

836: $k$. Typically, characteristic turbulent length-scales ($\lambda_c\sim

837: 1/k$) are much bigger than the charge exchange length-scales

838: $\lambda_{ce}\sim1/k_{ce}$, since for example, $\lambda_c \sim

839: (100-1000)\lambda_{ce}$ in the ISM cloud

840: \citep{Florinski2003,Florinski2005}. This essentially corresponds to

841: $k_{ce}/k\gg1$ \citep{zank1999,Florinski2003,Florinski2005}. It should

842: be noted that the charge exchange sources in the neutral and the

843: plasma fluids do not merely damp the smallest scale fluctuations, such

844: as is the case with linear collisional dissipation in turbulent fluid

845: flows, but they can alter the nonlinear cascade rate dramatically

846: \citep{shaikh2006,shaikh2007}. Because of its complex nonlinear

847: character (see \eqs{qm2}{qe}), charge exchange can be {\it effective on

848:   a variety of turbulent length-scales} unlike linear dissipation which

849: is operative on small length-scales only.  The effect and influence of

850: the charge exchange source terms on the turbulent inertial range of

851: coupled neutral and plasma fluids is therefore a subtle issue and is

852: not restricted only to the damping of small scales in partially

853: ionized, magnetized ISM turbulence. With this in mind, we carry out

854: two different sets of simulations of \eqs{mhd}{hd}, to discern the

855: effects of charge exchange on the ISM plasma-neutral system. In the

856: first set of simulations, \eqs{mhd}{hd} are decoupled (i.e. no charge

857: exchange coupling is assumed) and each evolves as an ideal fluid,

858: whereas in the second set the coupling is included self-consistently.

859: The former provides a reference simulation against which to

860: investigate the coupled system.

861: 

862: A two-dimensional (2D) nonlinear fluid code was developed to

863: numerically integrate \eqsto{mhd}{hd}. The 2D simulations are not only

864: computationally less expensive (compared to a fully 3D calculation),

865: but they offer significantly higher resolution (to compute inertial

866: range turbulence spectra) even on moderately-sized clusters like our

867: local Beowulf.  The spatial discretization in our code uses a discrete

868: Fourier representation of turbulent fluctuations based on a

869: pseudospectral method, while we use a Runge Kutta 4 method for the

870: temporal integration. All the fluctuations are initialized

871: isotropically (no mean fields are assumed) with random phases and

872: amplitudes in Fourier space.  This algorithm ensures conservation of

873: total energy and mean fluid density per unit time in the absence of

874: charge exchange and external random forcing. Additionally, $\nabla

875: \cdot {\bf B}=0$ is satisfied at each time step.  Our code is

876: massively parallelized using Message Passing Interface (MPI) libraries

877: to facilitate higher resolution. The initial isotropic turbulent

878: spectrum of fluctuations is chosen to be close to $k^{-2}$ with random

879: phases in both $x$ and $y$ directions.  The choice of such (or even a

880: flatter than -2) spectrum does not influence the dynamical evolution

881: as the final state in our simulations progresses towards fully

882: developed turbulence.  While the ISM turbulence code is evolved with

883: time steps resolved self-consistently by the coupled fluid motions,

884: the nonlinear interaction time scales associated with the plasma

885: $1/{\bf k} \cdot {\bf U}_p({\bf k})$ and the neutral $1/{\bf k} \cdot

886: {\bf V}_n({\bf k})$ fluids can obviously be disparate. Accordingly,

887: turbulent transport of energy in the plasma and the neutral ISM fluids

888: takes place on distinctively separate time scales.

889: 

890: Because of the different nonlinear time scales associated with the

891: plasma and neutral fluids, mode structures can be different in the two

892: fluids when they are evolved together and in isolation

893: (i.e. decoupled). The former is shown in \fig{fig3}{fig4}. The figures

894: respectively show the evolution of various physical quantities in the

895: plasma-neutral coupled system for a typical set of ISM parameters as

896: described in the captions (for typical ISM parameters, see also

897: \cite{zank1999,zank1996b,jacob2006,Florinski2003,Florinski2005}).  In

898: the absence of charge exchange (see \Fig{fig3}, right column), the

899: plasma fluid evolves as an ideal MHD fluid and is known to typically

900: develop sheet-like structures in the magnetic field

901: \citep{biskamp}. By contrast, the evolution of the magnetic field is

902: modified substantially when the two fluids are coupled through charge

903: exchange. For instance, the sheet-like structures present in the

904: turbulent magnetic field of the uncoupled plasma system instead are

905: smeared in the coupled plasma-neutral system on the typical

906: sheet-formation time-scales of ideal MHD.  The neutral fluid, under

907: the action of charge exchange terms, tends to modify the cascade rates

908: by isotropizing the ISM turbulence on a slower time scale. This point

909: is further elucidated in the subsequent section based on

910: Kolmogorov-like phenomenological arguments.  However if the coupled

911: system is evolved further, sheets do form eventually in the magnetic

912: field (after a relatively long time compared to the ideal MHD

913: time-scales) and a turbulent equipartition is set up in the plasma and

914: the neutral fluid modes.  In any event, the small-scale sheet-like

915: structures in magnetic field compress (or pinch) the plasma

916: density. Hence the density fluctuations develop identical (thinner

917: than) sheet-like structures that co-exist with small-scale turbulent

918: fluctuations in their spectrum as shown in \Fig{fig3}. The neutral

919: fluid, on the other hand, evolves isotropically as stated above by

920: forming relatively large-scale structures (see \Fig{fig4}).

921: 

922: 

923: 

924: 

925: 

926: \section{Energy Spectra}

927: Spectral transfer in ISM turbulence progresses under the action of

928: nonlinear interactions as well as charge exchange sources. Energy

929: cascades amongst turbulent eddies of various scale sizes and between

930: the plasma and the neutral fluids. In a freely decaying case, plasma

931: and neutral fluids evolve under the influence of charge exchange

932: forces which dramatically affect the energy cascades in the inertial

933: range. This is evident from \fig{fig5}{fig6} where Kolmogorov-like

934: fully developed turbulent spectra in the inertial range are shown

935: respectively for the coupled neutral and plasma fluids.  The neutral

936: fluid energy spectrum in the inertial regime for the coupled system

937: exhibits a $k^{-\alpha}$ spectrum, where the spectral index $\alpha

938: \approx 3.2$. On the other hand, the spectral index for plasma

939: magnetic and kinetic energy spectra for the coupled system is $\alpha

940: \approx 2.4$.  The inertial range spectral indices for the coupled

941: plasma-neutral system in our simulations show a {\it significant}

942: deviation from their corresponding uncoupled analogues which are

943: respectively $-3$ and $-1.7$ for the neutral and plasma fluids, also

944: plotted in \fig{fig5}{fig6}. The inertial range turbulent spectra are

945: much steeper for the coupled ISM plasma-neutral system than the

946: corresponding uncoupled or isolated neutral and plasma fluids in the

947: decaying turbulence regime. Consequently, the steep turbulent spectra

948: in the coupled ISM plasma-neutral system gives rise to an enhanced

949: spectral transfer amongst the inertial range modes, and the energy

950: cascade rates associated with the plasma and neutral (coupled) fluids

951: can be understood by invoking a Kolmogorov-like phenomenology as

952: described in the following.

953: 

954: The typical nonlinear interaction time-scale in ordinary

955: (i.e. uncoupled) plasma and/or neutral turbulence is given by

956: \be

957: \tau_{nl} \sim \frac{\ell_0}{v_\ell} \sim (kv_k)^{-1},

958: \ee 

959: where $v_k$ or $v_\ell$ is the velocity of turbulent eddies. In the

960: presence of charge exchange interactions, the ordinary nonlinear

961: interaction time-scale of fluid turbulence is modified by a factor

962: $k_{ce}/k$ such that the new nonlinear interaction time-scale of ISM

963: turbulence is now 

964: \be

965: \tau_{NL} \sim \frac{k_{ce}}{k} \frac{1}{kv_k}.

966: \ee 

967: The main justification of using the above expression originates from

968: the fact that turbulent cascade is determined typically by the

969: nonlinear interaction time ($\tau_{nl} \sim 1/kv_k$, where $\tau_{nl}$

970: is the eddy interaction time, $k$ is characteristic turbulent

971: wavenumber, and $v_k$ is characteristic speed of eddy in wavenumber

972: space) associated with the convection of turbulent eddies in the fluid

973: momentum equation. Note that this time scale follows from Kolmogorov's

974: phenomenology (Kolmogorov 1941) and corresponds essentially to the

975: eddy interaction time in an ideal, non interacting, and non magnetized

976: hydrodynamic-like turbulent fluid. Such time scales are essential to

977: estimate spectral energy transfer (or cascades) rates ($\varepsilon

978: \simeq E(k)/\tau_{nl}$, where $E(k)$ is turbulent energy per unit

979: mode) in a hydrodynamic turbulent fluid. For a non-ideal, interacting

980: and magnetized fluid, incorporating complex interactions amongst the

981: turbulent eddies with a background or local mean magnetic field, the

982: nonlinear eddy interaction time needs to be modified. This was first

983: pointed out by Kraichnan (1965) in the context of magnetohydrodynamic

984: (MHD) fluids where the interaction of turbulent eddies with Alfv\'en

985: waves, excited as a result of a mean magnetic field, was

986: addressed. Owing to the presence of waves in a MHD fluid, turbulent

987: correlations between velocity and magnetic field and the corresponding

988: energy transfer time are determined primarily by $\tau \sim (b_0

989: k)^{-1}$, where $b_0$ is a typical amplitude of a local magnetic field

990: and is dimensionally identical to that of the velocity field by virtue

991: of Els$\ddot{\rm a}$sser's symmetry (Kraichnan 1965). Note carefully

992: the modification in the energy transfer (or cascade) time scale

993: because of the wave-turbulent eddy interaction process.  Since the

994: convective time is the only nonlinear time scale involved in the

995: spectral transfer of turbulent energy, it it this time that accounts

996: entirely for the nonlinear cascade in the coupled plasma-neutral

997: heliosheath turbulence.  Likewise, we introduce an analogous

998: modification in the coupled plasma-neutral ISM turbulence where the

999: energy cascade time scale due to charge exchange interactions amongst

1000: turbulent eddies, is determined explicitly by the factor $k_{ce}/k$

1001: (where the symbols have their usual meanings).  It is this factor that

1002: determines how plasma and neutral fluids are coupled in heliosheath

1003: turbulence. Accordingly, the nonlinear spectral transfer time or the

1004: energy cascade rate is modified and is proportional to the factor

1005: $k_{ce}/k$ such that the new nonlinear eddy interaction time in the

1006: coupled plasma-neutral heliosheath turbulence is $\tau_{NL} \sim

1007: (k_{ce}/k) (1/kv_k)$. The interaction time $\tau_{NL}$ obtained in

1008: this manner for our coupled plasma-neutral problem not only adequately

1009: reproduces all three different regimes of interactions, but it also

1010: consistently describes the underlying physics in the context of

1011: heliosheath turbulence.

1012: 

1013: 

1014: 

1015: 

1016: On using the fact that $k_{ce}$ is typically larger than $k$ (or

1017: $k_c$, the characteristic turbulent mode, as defined elsewhere in the

1018: paper), i.e. $k_{ce}/k > 1$ in the ISM

1019: \citep{Florinski2003,Florinski2005,zank1999}, the new nonlinear time

1020: is $k_{ce}/k$ times bigger than the old nonlinear time i.e.

1021: $\tau_{NL} \sim (k_{ce}/k) \tau_{nl}$. This enhanced nonlinear

1022: interaction time in ISM turbulence is likely to prolong turbulent

1023: energy cascade rates.  It is because of this enhanced or prolonged

1024: interaction time that a relatively large spectral transfer of

1025: turbulent modes tends to steepen the inertial range turbulent spectra

1026: in both plasma and neutral fluids. By extending the above

1027: phenomenological analysis, one can deduce exact (analytic) spectral

1028: indices of the inertial range decaying turbulent spectra, as

1029: follows. The new nonlinear interaction time-scale of ISM turbulence

1030: can be rearranged as

1031: \be

1032: \tau_{NL} \sim

1033: \frac{k_{ce}v_k}{kv_k}\frac{1}{kv_k}\sim \frac{\tau_{nl}^2}{\tau_{ce}},\ee 

1034: where $\tau_{ce}\sim (k_{ce} v_k)^{-1}$ represents the charge exchange

1035: time scale. The energy dissipation rate associated with the coupled

1036: ISM plasma-neutral system can be determined from $\varepsilon \sim

1037: E_k/\tau_{NL}$, which leads to 

1038: \be

1039: \varepsilon \sim \frac{v_k^2}{k_{ce}/k^2 v_k}\sim \frac{k^2 v_k^3}{k_{ce}}.\ee

1040: According to the Kolmogorov theory, the spectral cascades are local in

1041: $k$-space and the inertial range energy spectrum depends upon the

1042: energy dissipation rates and the characteristic turbulent modes, such

1043: that $E_k \sim \varepsilon^{\gamma} k^{\beta}$.  Upon substitution of

1044: the above quantities and equating the power of identical bases, one

1045: obtains

1046: \be

1047: E_k \sim \varepsilon^{2/3} k^{-7/3}\ee

1048: for the plasma spectrum (the forward cascade inertial range). The spectral

1049: index associated with this spectrum, i.e. $7/3 \approx 2.33$, is

1050: consistent with the plasma spectrum observed in the (coupled ISM

1051: plasma-neutral) simulations (see \fig{fig5}{fig6}). Similar arguments in the

1052: context of neutral fluids, when coupled with the plasma fluid in ISM,

1053: lead to the energy dissipation rates

1054:  \be

1055: \varepsilon \sim \frac{k^2 v_k^2}{k_{ce}/k^2 v_k}.\ee

1056:  This further yields

1057:  the forward cascade (neutral) energy spectrum 

1058: \be

1059: E_k \sim \varepsilon^{2/3} k^{-11/3}\ee

1060:  which is close to the simulation result  shown in \Fig{fig6}.

1061: 

1062: 

1063: A key issue finally is to understand what role charge exchange modes

1064: play in coupled plasma-neutral ISM turbulence as they result

1065: essentially from the nonlinear charge exchange interactions. To

1066: address this issue, we plot charge exchange sources associated with

1067: the momentum and energy equations, \eqs{mhd}{hd}, in \Fig{fig7}. It

1068: appears from \Fig{fig7} that spectral energy is transferred

1069: predominantly at the larger scales by means of charge exchange mode

1070: coupling processes. The latter couples the large-scales, or smaller

1071: than $k_{ce}$ modes, efficiently to low-$k$ ISM turbulent modes.  It

1072: is primarily because of this $k_{ce}-k$ mode-coupling in the smaller

1073: $k$ part of the spectrum of the coupled plasma-neutral ISM turbulence,

1074: that energy is pumped efficiently at the lower $k$ inertial range

1075: turbulent modes. The efficient coupling of the Fourier modes at low

1076: $k$'s further enhances the nonlinear eddy time-scales associated with

1077: the coupled plasma-neutral turbulence system which is consistent with

1078: the scaling $\tau_{NL} \sim (k_{ce}/k) \tau_{nl}$, where $k_{ce}/k

1079: >1$, as described above.  This consequently leads to the steepening of

1080: the inertial range spectra observed in \fig{fig5}{fig6}. By contrast,

1081: higher $k$ modes, far from the energy cascade inertial range, are

1082: notably inefficient in transferring energy and momentum via charge

1083: exchange mode coupling interactions and are damped by small-scale

1084: dissipative processes in the coupled plasma-neutral ISM turbulence.

1085: 

1086: 

1087: 

1088: 

1089: 

1090: \section{Conclusions}

1091: In conclusion, we have developed a self-consistent fluid model to

1092: describe nonlinear turbulent processes in a partially ionized and

1093: magnetized ISM gas. The charge exchange interactions couple the ISM

1094: plasma and the neutral fluids by exciting a characteristic charge

1095: exchange coupling mode $k_{ce}$, which is different from the

1096: characteristic turbulent mode $k$ of the coupled system. One of the

1097: most important points to emerge from our studies is that charge

1098: exchange modes modify the ISM turbulence cascades dramatically by

1099: enhancing nonlinear interaction time-scales on large scales.  Thus on

1100: scales $\ell\ge\ell_{ce}$, the coupled plasma system evolves

1101: differently than the uncoupled system where large-scale turbulent

1102: fluctuations are strongly correlated with charge-exchange modes and

1103: they efficiently behave as driven (by charge exchange) energy

1104: containing modes of ISM turbulence.  By contrast, small scale

1105: turbulent fluctuations are unaffected by charge exchange modes which

1106: evolve like the uncoupled system as the latter becomes less important

1107: near the larger $k$ part of the ISM turbulent spectrum (see also

1108: \Fig{fig7}).  The neutral fluid, under the action of charge exchange,

1109: tends to enhance the cascade rates by isotropizing the ISM turbulence

1110: on a relatively long time scale.  This tends to modify the

1111: characteristics of ISM turbulence which can be significantly different

1112: from the Kolmogorov phenomenology of fully developed turbulence.  We

1113: believe that, it is because of this enhanced nonlinear eddy

1114: interaction time, that a large spectral transfer of turbulent energy

1115: tends to smear the current sheets in the magnetic field fluctuations

1116: and further cascade energy to the lower Fourier modes in the inertial

1117: range turbulent spectra. Consequently it leads to a steeper power

1118: spectrum. It is to be noted that the present model does not consider

1119: an external driving mechanism, hence the turbulence is freely

1120: decaying.  Driven turbulence, such as due to large scale external

1121: forcing, e.g. supernova explosion, may force turbulence at larger

1122: scales. This can modify the cascade dynamics in a manner usually

1123: described by dual cascade process.

1124: 

1125: There are issues that need further exploration.  For example, neutrals

1126: are observed to evolve isotropically in our coupled plasma-neutral ISM

1127: turbulence model. A relevant question, whether their coupling to the

1128: plasma fluid at large-scales reduces spectral anisotropy in the plasma

1129: fluctuations, remains to be investigated. Furthermore, our model is

1130: currently 2D. It is then intriguing whether variations in the third

1131: dimension, i.e. 3D, introduce non trivial effects to our present 2D

1132: studies because turbulence in 2D differs considerably from that in 3D,

1133: particularly in the context of 3D neutral hydrodynamics fluid, where

1134: vortex stretching effect (which is absent in 2D) tears apart the

1135: large-scale structures and forbids the inverse cascade phenomena. By

1136: virtue of latter, 2D and 3D neutral fluids possess distinct spectral

1137: features characterized essentially by the number of the inviscid

1138: quadratic invariants \citep{biskamp}. How coupling of the plasma and

1139: neutral fluids in partially ionized ISM turbulence is modified by the

1140: 3D interactions remains a subject of our future investigations.

1141: 

1142: We add finally that our self-consistent model can be useful in

1143: studying turbulent dynamics of partially ionized plasma in the

1144: magnetosphere of Saturn and Jupiter where outgassing from moons and Io

1145: and Encephalus introduces a neutral gas into the plasma.

1146: 

1147: 

1148: 

1149: 

1150: 

1151: 

1152: 

1153: %\begin{figure}

1154: %\includegraphics{fig1.ps}% Here is how to import EPS art

1155: %\caption{\label{fig1} A figure caption. The figure captions are

1156: %automatically numbered.}

1157: %\end{figure}

1158: 

1159: \acknowledgments

1160: 

1161: %\begin{acknowledgments}

1162:  The support of NASA(NNG-05GH38) and NSF (ATM-0317509) grants is  acknowledged.

1163: %\end{acknowledgments}

1164: 

1165: 

1166: 

1167: 

1168: 

1169: 

1170: 

1171: 

1172: %% The reference list follows the main body and any appendices.

1173: %% Use LaTeX's thebibliography environment to mark up your reference list.

1174: %% Note \begin{thebibliography} is followed by an empty set of

1175: %% curly braces.  If you forget this, LaTeX will generate the error

1176: %% "Perhaps a missing \item?".

1177: %%

1178: %% thebibliography produces citations in the text using \bibitem-\cite

1179: %% cross-referencing. Each reference is preceded by a

1180: %% \bibitem command that defines in curly braces the KEY that corresponds

1181: %% to the KEY in the \cite commands (see the first section above).

1182: %% Make sure that you provide a unique KEY for every \bibitem or else the

1183: %% paper will not LaTeX. The square brackets should contain

1184: %% the citation text that LaTeX will insert in

1185: %% place of the \cite commands.

1186: 

1187: %% We have used macros to produce journal name abbreviations.

1188: %% AASTeX provides a number of these for the more frequently-cited journals.

1189: %% See the Author Guide for a list of them.

1190: 

1191: %% Note that the style of the \bibitem labels (in []) is slightly

1192: %% different from previous examples.  The natbib system solves a host

1193: %% of citation expression problems, but it is necessary to clearly

1194: %% delimit the year from the author name used in the citation.

1195: %% See the natbib documentation for more details and options.

1196: 

1197: 

1198: 

1199: 

1200: %% Use the figure environment and \plotone or \plottwo to include 

1201: %% figures and captions in your electronic submission.

1202: 

1203: %\clearpage

1204: 

1205: 

1206: \begin{thebibliography}{99}

1207: 

1208: \bibitem[Banks (1966)]{banks}

1209: Banks, P., 1966, 

1210: Planetary and Space Sci. 14, 1105.

1211: 

1212: \bibitem[Basu et al (1994)]{basu}

1213: Basu, S., and Mouschovias, T. Ch. 1994, ApJ, 432, 720.

1214: 

1215: \bibitem[Biskamp (2003)]{biskamp}

1216: Biskamp, D., 2003,  Nonlinear MHD Turbulence, Academic Publisher.

1217: 

1218: \bibitem[Cho \& Lazarian (2003)]{cho2003}

1219: Cho, J. and  Lazarian, A., 2003, MNRAS, 345, 325.

1220: 

1221: \bibitem[Elmegreen (1985)]{elmegreen}

1222: Elmegreen, B.G. 1985, ApJ, 299, 196.

1223: 

1224: \bibitem[Fite et al (1962)]{fite}

1225: Fite, W. L., Smith, A. C. H., and Stebbings, R. F., 1962,

1226: Proc. R. Soc. London, em A. 268, 527.

1227: 

1228: \bibitem[Florinski et al (2003)]{Florinski2003}

1229: Florinski, V., G. P. Zank, and N. V. Pogorelov (2003), Galactic cosmic

1230: ray transport in the global heliosphere, J. Geophys. Res., 108(A6),

1231: 1228, doi:10.1029/2002JA009695.

1232: 

1233: \bibitem[Florinski et al (2005)]{Florinski2005}

1234: Florinski V., G. P. Zank, N. V. Pogorelov (2005), Heliopause stability

1235: in the presence of neutral atoms: Rayleigh-Taylor dispersion analysis

1236: and axisymmetric MHD simulations, J. Geophys. Res., 110, A07104,

1237: doi:10.1029/2004JA010879.

1238: 

1239: \bibitem[Goldsmith \& Langer (1978)]{goldsmith}

1240: Goldsmith, P.F. and Langer, W.D. 1978, ApJ, 222, 881.

1241: 

1242: \bibitem[Gayley et al (1996)]{gayley1996}

1243: Gayley, K., Zank, G. P., Pauls, H. L., Frisch, P. C., and Welty,

1244: D. E.: 1997, Astrophys. J. 487, 259. 

1245: 

1246: 

1247: \bibitem[Groebner et al (2001)]{groebner2001} 

1248: Groebner, R., Madhavi, M. A., Leonard, A. W., Osborne, T. H., Porter,

1249: G. D., Colchin, R. J., and Owen, L. W., 2001, GA Report GA-A23856,

1250: November; Groebner, R., Baker, D. R., and Burrell, K. H., et al, 2001,

1251: 41 No 12, 1789-1802.

1252: 

1253: 

1254: \bibitem[Hill \& Dessler (1990)]{hill1990}

1255: Hill, T. W.; Dessler, A. J., 1990

1256: Geophys. Res. Lett. (ISSN 0094-8276), vol. 17, Sept. 1990, p. 1677-1680.

1257: 

1258: 

1259: 

1260: \bibitem[Heerikhuisen et al (2006)]{jacob2006}

1261: Heerikhuisen, J.; Florinski, V.; Zank, G. P. 2006,

1262: JGR., 111,  A06110.

1263: 

1264: 

1265: \bibitem[Ip (1997)]{ip1997}

1266: Ip, W.-H.,  1997,

1267: Icarus, Volume 126, Issue 1, pp. 42-57.

1268: DOI:	10.1006/icar.1996.5618

1269: 

1270: \bibitem[Kolmogorov(1941)]{k41} 

1271: Kolmogorov, A. N., 1941, Dokl. Akad. Nauk SSSR, 31, 538.

1272: 

1273: 

1274: \bibitem[Kraichnan (1965)]{kraichnan}

1275: Kraichnan, R. H.,  1965, Phys. Fluids, 8, 1385.

1276: 

1277: \bibitem[Kudoh et al (2007)]{kudoh}

1278: Kudoh, T.,  Basu, S.,  Ogata, Y.,  Yabe, T.,  2007, MNRAS, 380,  499.

1279: 

1280: \bibitem[Kulsrud \& Pearce (1969)]{kulsrud}

1281: Kulsrud, R. and Pearce, 1969, ApJ 156, 445.

1282: 

1283: 

1284: \bibitem[Leake et al (2005)]{leake}

1285: Leake, J. E., Arber, T. D., Khodachenko, M. L.,  2005, 

1286: arXiv:astro-ph/0510265v1.

1287: 

1288: \bibitem[Li \& Nakamura (2004)]{li}

1289: Li, Z.-Y., Nakamura, F., 2004, ApJ, 609, L83.

1290: 

1291: \bibitem[Muller et al (2006)]{muller2006}

1292: %Muller, H.-R. and Zank, G. P., 

1293: %2006, ASPC., 359, 276.

1294:  Mueller, H.-R., Frisch, P. C., Florinski, V., and

1295: Zank, G. P., 2006, ApJ, 647, 1491. 

1296: 

1297: \bibitem[Mestel \& Spitzer (1956)]{mestel}

1298: Mestel, L., Spitzer, L., Jr., 1956, MNRAS, 116, 503.

1299: 

1300: \bibitem[McNutt et al (1998)]{nutt}

1301: McNutt et al. 1998, JGR, 103, 1905.

1302: 

1303: \bibitem[McIvor (1977)]{mcivor}

1304: McIvor, 1977, MNRAS 178, 85.

1305: 

1306: 

1307: \bibitem[Nakamura \& Li (2005)]{nakamura}

1308: Nakamura, F., Li, Z.-Y., 2005, ApJ, 631, 411.

1309: 

1310: 	

1311: \bibitem[Nakano (1979)]{nakano}

1312: Nakano, T. 1979, PASJ, 31, 697.

1313: 

1314: \bibitem[Oishi \& Mac Low (2006)]{oishi}

1315: Oishi, J. S. and Mac Low, M., 2006, ApJ, 638, 281.

1316: 

1317: \bibitem[Padoan et al (2000)]{padoan}

1318: Padoan, P.,  Zweibel, E., and Nordlund, A., 2000, ApJ, 540, 332.

1319: 

1320: 

1321: \bibitem[Potekhin et al (2005)]{Potekhin2005}

1322: Potekhin, A. Y.; Lai, Dong; Chabrier, G.; Ho, W. C. G., 2005,

1323: Ad. Sp. Res., 35, 1158.

1324: DOI:10.1016/j.asr.2004.12.010.

1325: 

1326: \bibitem[Pauls et. al (1995)]{pauls1995}

1327: Pauls, H. L., Zank, G. P., and Williams, L. L., 1995,  

1328: J. Geophys. Res. A11, 21595.

1329: 

1330: 

1331: \bibitem[Scalo (1977)]{scalo}

1332: Scalo, J.M. 1977, ApJ, 213, 705.

1333: 

1334: \bibitem[Shaikh et al (2006)]{shaikh2006} 

1335: Shaikh, D. Pogorelov, N. V, and Zank, G. P., 2006, 

1336: Astronomical Society of the Pacific

1337: Conference Series (ASPC),  {\it Numerical

1338: Modeling of Space Plasma Flows: Astronum-2006}, Edited by N. V. Pogorelov and G. P. Zank,

1339: 359, 91.

1340: 

1341: \bibitem[Shaikh \& Zank (2007)]{shaikh2007}

1342: Shaikh, D. and Zank, G. P., 2007, AIP Conf. Procs,

1343:  932, 111, DOI:10.1063/1.2778952.

1344: 

1345: \bibitem[Shebalin \& Montgomery (1983)]{shebalin}

1346: Shebalin, J. V., Matthaeus, W. H., and Montgomery, D.

1347: 1983, J. Plasma Phys., 29, 525.

1348: 

1349: \bibitem[Shu (1983)]{shu}

1350: Shu, F. H. 1983, ApJ, 273, 202.

1351: 

1352: \bibitem[Singh et al (2004)]{singh2004}

1353: Singh, R., Rogister, A., and Kaw, P., 2004, 

1354: Phys. Plasma, 11, 129.

1355: 

1356: \bibitem[Slavin \& Frisch (2006)]{slavin2006}

1357: Slavin, J. D. and Frisch P. C., 2006, 

1358: ApJ  {\bf 651},  L37-L40.

1359: 

1360: \bibitem[Valovie et al (2001)]{valovie2001}

1361: Valovie, M., Fielding, S. J., Carolan, P. G., et al., 2001,

1362: Contr. Fusion and Plasma Phys. 25A, 1829.

1363: 

1364: \bibitem[Valovie et al (2002)]{valovie2002}

1365: Valovie, M.,  Carolan, P. G., Fielding, S. J., et al., 2002,

1366: Plasma Phys. Contr. Fusion 44, A175.

1367: 

1368: 

1369: \bibitem[Wood et. al (2000)]{wood2000}

1370: Wood, B. E.;  Mueller, H.-R.; Zank, G. P, 2000, 

1371: ApJ., 542, 493.; 

1372: Wood, B. E.; Linsky, J. L.; Zank, G. P., 2000, 

1373: ApJ., 537, 304.

1374: 

1375: \bibitem[Zank \& Pauls (1996)]{zank1996}

1376:  Zank, G. P. and  Pauls, H. L., 1996, Sp. Sci. Rev., 78, 95.

1377: 

1378: \bibitem[Zank et al. (1996)]{zank1996b}

1379: Zank, G. P.; Pauls, H. L.; Williams, L. L.; Hall, D. T., 2006, 

1380: JGR, 101, A10, p. 21639-21656.

1381: 

1382: 

1383: 

1384: \bibitem[Zank (1999)]{zank1999}

1385:  Zank, G. P., Sp. Sci. Rev., 89, 413-688, 1999.

1386: 

1387: 

1388: \bibitem[Zank et al (2000)]{zank2000}

1389: Zank, G. P., Lu, J. Y., Rice, W. K. M.,  and Webb, G. M., 2000, J. Plasma

1390: Phys. {\bf 64}, 507-541.

1391: 

1392: \bibitem[Zank et al (2006)]{zank2006}

1393: %Heliospheric Variation in Response to Changing Interstellar

1394: %Environments by 

1395: Zank, G. P.,  Mueller, H. -R., Florinski, V., and

1396: Frisch, P. C.,  2006, Chapter 2 in Solar Journey:  Significance of our

1397: Galactic Environment for the Heliosphere and Earth, editor

1398: P.C. Frisch, Springer, pp. 281-316. 

1399: 

1400: 

1401: \bibitem[Zweibel \& Josafatsson (1983)]{zweibel}

1402: Zweibel, E.G. and Josafatsson, K. 1983, ApJ, 270, 511.

1403: 

1404: 

1405: \end{thebibliography}

1406: 

1407: 

1408: 

1409: \begin{figure}

1410: \plotone{fig1.ps} \figcaption{\label{fig0} Schematic illustrating

1411:   the plasma neutral simulation regions both in the local interstellar

1412:   medium and in the outer heliosheath (see \cite{zank1999} for a

1413:   review of the solar wind interaction with the local ISM).}

1414: \end{figure}

1415: 

1416: \clearpage

1417: 

1418: \begin{figure}

1419: \plotone{fig2.ps} \figcaption{\label{fig1} Schematic

1420: illustrating the significance of the charge exchange mode, which

1421: corresponds to charge exchange interactions per unit area in the

1422: computational domain. This mode, being different from the

1423: characteristic turbulent mode $k$, is excited naturally when plasma

1424: and neutral fluids are coupled in the ISM by charge exchange. }

1425: \end{figure}

1426: 

1427: \begin{figure}

1428: \plotone{fig3.ps} \figcaption{\label{fig3} ({\bf Left column}) Mode

1429:   structures of plasma when coupled with the neutral ISM fluid.  The

1430:   local ISM case is considered where typical LISM parameters are;

1431:   $\gamma=5/3, n_{0_{plasma}} = 0.06 cm^{-3}, n_{0_{neutral}} = 0.18

1432:   cm^{-3}, T = 6500 K, V_{th} = 10 km/s, \sigma = 6.9\times 10^{-15},

1433:   l_0 = 200 AU, L({\rm box length}) = 5 l_0$. ({\bf Right column}) The

1434:   corresponding quantities for an uncoupled ISM plasma simulation

1435:   measured at a similar time step (i.e. $t=20$ in normalized units).

1436:   Notice the magnetic field structures in the uncoupled case showing

1437:   the formation of thin (or sheet-like) structures unlike the coupled

1438:   system. Also notice the differences in the density fields.  }

1439: \end{figure}

1440: 

1441: \clearpage

1442: 

1443: \begin{figure}

1444: \plotone{fig4.ps} \figcaption{\label{fig4} The corresponding mode

1445:   structures of neutral fluid in the presence of a plasma

1446:   fluid. Simulation parameters are described in the previous figure.

1447: }

1448: \end{figure}

1449: 

1450: \clearpage

1451: 

1452: \begin{figure}

1453: \plotone{fig5.ps} \figcaption{\label{fig5}The inertial range

1454:   turbulent spectrum for a neutral fluid shows a $k^{-3.2}$ spectrum when

1455:   coupled through charge exchange to a plasma fluid.  The spectral

1456:   index exhibited by this spectrum is larger than that of an ordinary

1457:   (hydrodynamic) fluid. The latter exhibits a Kolmogorov-like $k^{-3}$

1458:   spectrum in the forward cascade regime (drawn schematically). The

1459:   numerical resolution for our spectral studies is $1024^2$ modes in

1460:   2D.}

1461: \end{figure}

1462:  

1463: 

1464: \clearpage

1465: 

1466: \begin{figure}

1467: \plotone{fig6.ps} \figcaption{\label{fig6} A similar steepening of

1468:   turbulent spectra occurs in the plasma fluid when coupled to the

1469:   neutrals through charge exchange. An uncoupled magnetofluid plasma

1470:   exhibits a Kolmogorov-like $k^{-5/3}$ spectrum in the forward

1471:   cascade regime (drawn schematically).}

1472: \end{figure}

1473: 

1474: 

1475: \clearpage

1476: 

1477: \begin{figure}

1478: \plotone{fig7.ps} \figcaption{\label{fig7} 

1479: Charge exchange spectrum associated with momentum and energy in both plasma

1480: and neutral fluids. Clearly the energy cascades are dominated by large-scale

1481: interactions, whereas small scales are not efficient enough to influence the

1482: turbulent spectra in the local ISM.

1483: }

1484: \end{figure}

1485: 

1486: \clearpage

1487: 

1488: 

1489: 

1490: 

1491: 

1492: 

1493: \end{document}

1494: 

1495: \be

1496: \label{mhd:cont}

1497:  \frac{\partial \rho_p}{\partial t} + \nabla \cdot (\rho_p{\bf U}_p)=0;

1498: \ee

1499: \be

1500: \label{mhd:mom}

1501: \rho_p \left(  \frac{\partial}{\partial t} + {\bf U}_p \cdot \nabla \right) {\bf U}_p

1502: = -\nabla P_p + \frac{1}{c} {\bf J} \times {\bf B}+{\bf Q}_M({\bf U}_p,{\bf V}_n, \rho_p, \rho_n, T_n, T_p);

1503: \ee

1504: \be

1505: \label{mhd:mag}

1506:  \frac{\partial {\bf B}}{\partial t} = \nabla \times ({\bf U}_p \times {\bf B});

1507: \ee

1508: \eqa

1509: \label{mhd:en}

1510:  \frac{\partial}{\partial t} \left( \frac{1}{2}\rho_p U_p^2 + \frac{P_p}{\gamma-1}+\frac{B^2}{8\pi}\right)

1511: &+& \nabla \cdot \left( \frac{1}{2}\rho_p U_p^2{\bf U}_p 

1512: + \frac{\gamma}{\gamma-1}\frac{P_p}{\rho_p}\rho_p{\bf U}_p

1513: + \frac{c}{4\pi}{\bf E} \times {\bf B}  \right) \nonumber \\

1514: &=&Q_E({\bf U}_p,{\bf V}_n,\rho_p, \rho_n, T_n, T_p).

1515: \eeq

1516: 

1517: 

1518: 

1519: \be

1520: \label{hd:cont}

1521:  \frac{\partial \rho_n}{\partial t} + \nabla \cdot (\rho_n{\bf V}_n)=0;

1522: \ee

1523: \be

1524: \label{hd:mom}

1525: \rho_n \left(  \frac{\partial }{\partial t} + {\bf V}_n \cdot \nabla \right) {\bf V}_n

1526: = -\nabla P_n + {\bf Q}_M({\bf V}_n,{\bf U}_p,\rho_p, \rho_n, T_n, T_p);

1527: \ee

1528: \be

1529: \label{hd:en}

1530:  \frac{\partial}{\partial t} \left( \frac{1}{2}\rho_n V_n^2 + \frac{P_n}{\gamma-1} \right)

1531: + \nabla \cdot \left( \frac{1}{2}\rho_n V_n^2{\bf V}_n 

1532: + \frac{\gamma}{\gamma-1}\frac{P_n}{\rho_n}\rho_n{\bf V}_n \right)

1533: =Q_E({\bf V}_n,{\bf U}_p,\rho_p, \rho_n, T_n, T_p).

1534: \ee

1535: 

1536: 

1537: 

1538: \bibitem[Zank et al (1998)]{zank1998}

1539: Zank, G. P.; Matthaeus, W. H.; Bieber, J. W.; Moraal, H., 1998,

1540: JGR., 103, 2085.

1541: