Alfvénic localized structures in partially ionized plasmas

Jafar Borhanian and Arash Rezaei

Citation: Physics of Plasmas 24, 022302 (2017); doi: 10.1063/1.4975186

View online: http://dx.doi.org/10.1063/1.4975186
View Table of Contents: http://aip.scitation.org/toc/php/24/2
Published by the American Institute of Physics

Articles you may be interested in

Effects of ion motion on linear Landau damping
Physics of Plasmas 24, 022101022101 (2017); 10.1063/1.4975020

On the nonlinear dynamics of breathers waves in electronegative plasmas with Maxwellian negative ions
Physics of Plasmas 24, 022105022105 (2017); 10.1063/1.4975090

Ion temperature effects on its chaotic behavior in ion-sheath

Physics of Plasmas 24, 022301022301 (2017); 10.1063/1.4975316

Two-fluid biasing simulations of the large plasma device

Physics of Plasmas 24, 022303022303 (2017); 10.1063/1.4975616

Conservation laws for collisional and turbulent transport processes in toroidal plasmas with large mean flows
Physics of Plasmas 24, 020701020701 (2017); 10.1063/1.4975075

Power law spectra and intermittent fluctuations due to uncorrelated Lorentzian pulses
Physics of Plasmas 24, 020704020704 (2017); 10.1063/1.4975645
 PHYSICS OF PLASMAS 24, 022302 (2017)

nic localized structures in partially ionized plasmas

Alfve
Jafar Borhaniana) and Arash Rezaei
Department of Physics, Faculty of Science, University of Mohaghegh Ardabili, P. O. Box 179, Ardabil, Iran
(Received 11 December 2016; accepted 17 January 2017; published online 3 February 2017)
The existence and dynamics of Alfvenic localized structures are investigated in partially ionized
plasmas. We have employed the Hall magnetohydrodynamics model for partially ionized plasmas
and shown that the evolution of a weakly nonlinear and weakly dispersive Alfven wave is
governed by a derivative nonlinear Schr€odinger (DNLS) type equation. In the Hall effect
domination limit, this equation reduces to a standard DNLS equation that possesses localized
solutions in the form of solitons and rogue waves. The dependence of the profile of these structures
on the Hall parameter is addressed. When the ohmic and ambipolar effects are small but finite in
comparison to the Hall effect, the evolution equation takes the form of a perturbed DNLS equation.
In this limit, the dynamics of envelope soliton solution is examined by means of the soliton
perturbation method, the moment method, to be precise. Published by AIP Publishing.
[http://dx.doi.org/10.1063/1.4975186]

I. INTRODUCTION of numerous observations of large amplitude electromagnetic

waves in space plasmas,6,22 e.g., formation of magnetic
The Alfven waves are a kind of magnetohydrodynamic
holes23 and existence of numerous localized structures such
(MHD) wave that are of fundamental importance in the
as shock waves,24 solitons,25,26 and rogue waves.27,28
behavior of many laboratory and space plasmas. Alfven
In regard to nonlinear Alfven waves, all the studies are
waves have long been considered as an important candidate
limited to the fully ionized plasmas. Rogister used the kinetic
for the heating of, and the transport of energy in, laboratory,
model to obtain the derivative nonlinear Schr€odinger
space, and astrophysical plasmas.1–4 Propagation of linear
(DNLS) equation that governs the evolution of the low fre-
and nonlinear Alfven waves in fully ionized plasmas and
quency weakly nonlinear Alfven waves in a magnetized
related issues are a subject of intense study in succeeding
decades. The interested reader is refereed to Refs. 5 and 6 plasma.29 Subsequently, the two fluid approach is used to
for more information. derive the DNLS equation.30,31 A number of analytical and
Astrophysical and space plasmas often are partially ion- numerical studies have used the DNLS. We here mention
ized, such as solar atmosphere, Earth’s ionosphere, proto- few for the sake of reference. The modulational instability
planetary discs, and so on. Clearly, a partially ionized of parallel propagating Alfven waves was investigated in
plasma does not behave like fully ionized plasma and hence, Refs. 32–34. The one-solitons,35 two-solitons,36 stationary
may display new or different features. In a partially ionized solutions,37 N-solitons,38 breathers,39 and rogue waves40
plasma, due to the friction between charged particles and are various possible solutions of the DNLS equation. The
neutrals, a propagating Alfven wave also involves motion of papers18,22,41 address the physical and mathematical features
the neutrals. This leads to the damping of Alfven waves and of the DNLS equation and some of its solutions. The oblique
appearance of cut-off wavelength which arises due to colli- propagation of nonlinear Alfven waves in strongly magne-
sions between ions and neutrals and are vastly studied using tized plasmas was considered in Refs. 42 and 43. The effect
the single fluid model.7–11 These issues are recently recon- of Landau damping on dynamics of Alfven soliton and mul-
sidered using the two fluid approach.12,13 For example, the tisoliton is discussed in Ref. 44.
effects of neutral helium and stratification on torsional Although there is an immense value of investigations
Alfven waves is addressed in Ref. 14. related to nonlinear Alfven waves, to the best of our knowl-
Almost all the references cited above are dealing with edge, all of them concern with the fully ionized plasmas. As
linear Alfven waves and related issues in partially ionized mentioned in the above paragraphs, most part of the astro-
plasmas. The nonlinear propagation of MHD waves, espe- physical and space plasmas are partially ionized. Therefore,
cially Alfven waves, is of great interest both in basic plasma the study of the nonlinear effects related to Alfven waves is of
physics and in the fields of space and astrophysical problems. great interest, not only from the point of view of basic plasma
Finite amplitude Alfven waves can cause a number of non- physics, but also it is crucial to understanding and interpreta-
linear effects15 such as modulational and filamentation tion of the various phenomena in astrophysical, space, and
instabilities,16,17 existence of solitons,18 modification of the laboratory plasmas. Our aim in this paper is to employ the
background plasma number density,19 and parametric insta- recently proposed Hall magnetohydrodynamics model45 to
bilities.20,21 Understanding the nonlinear aspects of the explore the existence and dynamics of localized structures,
Alfven wave propagation is indispensable for interpretation i.e., solitons and rogue waves, in partially ionized plasmas.
This paper is organized as follows. Section II presents
a)
Electronic mail: [email protected] the basic model equations. In Section III, we carried out a

1070-664X/2017/24(2)/022302/9/$30.00 24, 022302-1 Published by AIP Publishing.

022302-2 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

 
survey on linear features of Alfven waves in partially ionized @v @v @B
plasmas. The equation governing the weakly nonlinear evo- q þu ¼ ; (6)
@t @z @z
lution of Alfven waves is obtained in Section IV. Then,    
Section V is devoted to the existence and dynamics of local- @u @u @ 1 2
q þu ¼ bq þ jBj ; (7)
ized structures in the form of solitons and rogue waves. @t @z @z 2
Finally, the summary and conclusion of this work are pre- 2
 
@B @ @ B @ 1 @B
sented in Section VI. ¼ ðv  uBÞ þ g 2  iagH
@t @z @z @z q @z
"   #
II. BASIC MODEL @ 1 @B 1 @jBj2
þ gA þ B ; (8)
We consider a partially ionized plasma consisting of @z q @z 2 @z
electrons, singly ionized ions, and neutrals immersed in
an ambient uniform magnetic field. As far as the propaga- where jBj2 ¼ B2x þ B2y , u ¼ vz and b ¼ c2s =v2A with vA ¼ B0 =
tion of low frequency waves are concerned, the single ð4pq0 Þ1=2 as the Alfven velocity. It’s worth to mention that
fluid Hall MHD model of Pandey and Wardle45 can be the parameter b, known as plasma beta, represents the ratio
used to describe the propagation of Alfven waves that are of the hydrostatic pressure to magnetic pressure. In the limit
given by of cold plasma approximation, this parameter approaches
zero. In the above system of equations, the magnetic field is
@q normalized by its ambient counterpart B0, the velocities are
þ r  ðqvÞ ¼ 0; (1)
@t normalized by Alfven velocity, the density is normalized by
 
@v 1 background bulk density q0; length and time have expressed
q þ ðv  rÞv ¼ rp þ J  B; (2) in units of vA/xci and x1
@t c ci , respectively, where the ion cyclo-
" tron frequency is defined through xci ¼ eB0 =mi . Moreover,
@B 4pg 4pgH J  B the parameters g, gH, and gA are respectively normalized by
¼r vB J
@t c c q v2A =xci ; q0 v2A =ðB0 xci Þ, and q0 v2A =ðB20 xci Þ. The system of
# Equations (5)–(8) can be used to study the one-dimensional
4pgA ðJ  BÞ  B propagation of arbitrary amplitude Alfven waves along the
þ : (3)
c q ambient magnetic field in a partially ionized plasma. By dint
of this model, the linear and nonlinear propagation of Alfven
This system of equations can be closed using the isother- waves will be addressed in the following.
mal equation of state p ¼ c2s q along with Ampere’s law
J ¼ cr  B=4p. In the equations above, q ¼ qi þ qn is the III. LINEAR DISPERSION RELATION
bulk fluid density and v ¼ ðqi vi þ qn vn Þ is the bulk veloc-
ity with qi, vi and qn, vn as the mass density and bulk In this section, we reexamine the linear propagation of
velocities of the ion and neutral fluids, respectively. The Alfven waves in partially ionized plasmas. Spatial and tem-
constant sound speed is defined through c2s ¼ cp=q, where poral Fourier transformation of the linearized form of
c is the isentropic factor. The electron bulk density is Equations (5)–(8) leads us to the following linear dispersion
neglected in the definition of bulk quantities. Moreover, relation (DR) for Alfven waves propagating along the mag-
the Ohmic (g), Hall (gH) and ambipolar (gA) parameters netic field10,45
are given by
x2  k2 ðig þ igA þ agH Þx  k2 ¼ 0; (9)
2 2
c D mi c
g¼ ; gA ¼ ; gH ¼ : (4) where the Alfven wave frequency x and wavenumber k are
4pr 4pin ð1  DÞ 4peð1  DÞ
normalized by the ion cyclotron frequency xci and the skin
Here, D ¼ qn/q is the neutral density fraction and the ion- depth xci/vA, respectively. Regarding this dispersion relation,
neutral collision frequency is given by in ¼ 1:9  109 nn two important issues can be considered. One is the damping
ðcm3 Þ. For weak ionization D ! 1, while for highly ionized of Alfven waves that makes contribution to warming the par-
plasma D ’ 0. tially ionized medium, and the other is related to the cut-off
We chose the coordinate system in a way that the ambi- wavenumber. It is well known that the presence of neutral
ent magnetic field B0 oriented along the z axis. In the one- atoms in partially ionized plasmas enhances the dissipation
dimensional approximation where all physical quantities of MHD waves.1,7–9,12,46–50 To investigate this issue, one
depend only on spatial variable z, the law of r  B ¼ 0 leads needs to obtain solutions of the dispersion relation (9) that is
to Bz ¼ B0 and Ampere’s law results in Jz ¼ 0. Therefore, given by
upon introducing B ¼ Bx þ iaBy and v ¼ vx þ iavy with  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
a ¼ 1 (a ¼ 1) for left circularly polarized (right circularly k
x ¼ ðigT þ agH Þk6 4 þ k2 ðigT þ agH Þ2 ; (10)
polarized) or LCP (RCP) waves, Eqs. (1)–(3) cast into the 2
following scalar equations:
with gT ¼ g þ gA . We can consider the wavenumber k to be
@q @ real and look for complex frequency x ¼ xr þ ixi (temporal
þ ðqvÞ ¼ 0; (5)
@t @z damping). Then, the dispersion relation (10) results in
022302-3 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

   
1 2 h 1 2 h If we go to the next order, the transverse momentum (6) and
xr ¼ ak gH 6kr cos ; xi ¼ k gT 6kr sin ; induction (8) equations yield
2 2 2 2
(11) @v2 @B2 @v1 @v1
þ ¼ þ ðu1  q1 Þ ; (19)
where the module r and the phase h are defined through @n @n @s @n
r ¼ ðe2r þ e2i Þ1=2 and h ¼ arctanðei =er Þ, respectively, with @v2 @B2 @B1 @ ðu1 B1 Þ @ 2 B1
er ¼ 4 þ k2 ðg2H  g2T Þ and ei ¼ 2ak2 gH gT . In the absence þ ¼ þ þ ðiagH  gT Þ 2 : (20)
@n @n @s @n @n
of the Hall effect, i.e., when xr  xH ¼ v2A =gH ,45 the real
and imaginary parts simplify to The compatibility condition for Eqs. (19) and (20), by the
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi help of (18), gives up the desired nonlinear evolution equation
1 1
xr ¼ k 4  k2 g2T ; xi ¼ k2 gT : (12)
2 2 @B1 1 @   1 @ 2 B1
þ B1 jB1 j2 þ ðiagH  gT Þ 2 ¼ 0:
One sees that there is a critical wavenumber, kc ¼ 2/gT, at @s 4ð1  bÞ @n 2 @n
which the real part of the frequency becomes zero. Alfven (21)
waves cannot propagate for wavenumbers higher than this
cut-off value.11,51 In the presence of the Hall effect, the sim- When gT ¼ 0, Equation (21) is called a derivative nonlinear
ple analytical expression for cut-off wavenumber cannot be Schr€odinger (DNLS) equation;35 otherwise, it is sometimes
obtained from (11) and, therefore, would be investigated called a perturbed DNLS equation.53 This equation describes
numerically. This subject is thoroughly discussed in Refs. 8, the propagation of weakly nonlinear and weakly dispersive
10, and 45. Alfven waves along the ambient magnetic field in partially
ionized plasmas in the presence of Hall, ambipolar, and
IV. NONLINEAR EVOLUTION EQUATION Ohmic effects.
To study small but finite amplitude Alfven waves, we V. LOCALIZED STRUCTURES
make the following stretched space and time variables:31,52
Our aim in this section is to report localized solutions
n ¼ ðz  tÞ; s ¼ 2 t; (13) for Eq. (21). The problem to be tractable analytically, we
have restricted the discussion to two limiting cases where the
where  is a small parameter measuring the weakness of the Hall effect is dominated or the ambipolar and ohmic effects
dispersion and of the nonlinear effects. We then expand the are small in comparison to the Hall effect.
dependent variables around their homogeneous equilibrium
values in powers of 31,33 A. Solitons and rogue waves

qðn; sÞ ¼ 1 þ q1 ðn; sÞ þ 2 q2 ðn; sÞ þ   ; (14) First, we consider the case where ambipolar and ohmic
effects can be neglected in comparison to the Hall effect.
uðn; sÞ ¼ u1 ðn; sÞ þ 2 u2 ðn; sÞ þ   ; (15) This condition can occur for example, in protoplanetary
vðn; sÞ ¼ 1=2 v1 ðn; sÞ þ 3=2 v2 ðn; sÞ þ   ; (16) discs.45 With this restriction, Eq. (21) reduces to the DNLS
equation, which belongs to the class of integrable equations
1=2 3=2
Bðn; sÞ ¼  B1 ðn; sÞ þ  B2 ðn; sÞ þ   : (17) and possesses several types of exact solutions under the dif-
ferent boundary conditions as mentioned in the Introduction.
Now, substituting expressions (13)–(17) into Equations We want to focus in particular, one, i.e., one-soliton solution.
(5)–(8) and equating terms with equal powers of  result in For a ¼ 1, and by imposing the vanishing boundary condi-
systems of equations that can be solved to obtain desired tion, the self-phase-modulated wavepacket solutions to the
results. In the lowest order approximation, system of equa- DNLS equation are given by30,34
tions leads to the following results:
B1 ðn; sÞ ¼ wðn; sÞ expðihÞ; (22)
jB1 j2
v1 ¼ B1 ; q1 ¼ u1 ¼ : (18) where the amplitude was obtained as
2ð1  bÞ

 
2 4 K02 þ X0
w ¼ " qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi !# ; (23)
 1=2 1 2K0
X0 þ 2K02 cosh 2 K02 þ X20 ðn  n0 Þ þ s þ K0
2ð1  bÞgH 8ð1  bÞ2 gH

and the wavenumber K and frequency X that are respectively defined through hn ¼ K=½2ð1  bÞgH ; hs ¼ X=½8ð1  bÞ2 gH 
are given by

K ¼ K0 þ 3w2 =4; X ¼ X0  3K0 w2 =2; (24)

022302-4 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

with arbitrary constants K0 and X0. Here, n0 denotes the ini-

tial position of the soliton. The soliton wavepacket solution
(22) belongs to the two parameter family solutions and is an
exact soliton in view of the inverse scattering transform
approach. It is obvious from (23) that the soliton amplitude
is independent of the plasma parameters but its width
depends on plasma parameter beta and Hall parameter gH. In
order to show the behavior we adopt the parameter values to
the protoplanetary disc condition. The variation of the profile
of the soliton solution (22) with Hall parameter gH is
depicted in Fig. 1. As one also can deduce from (23), the sol-
iton’s width increases when the Hall parameter increases. As
is clear from definitions and can also be deduced from physi-
cal points of view (Hall term is inversely proportional to
electron density), the Hall parameter gH as well as ambipolar
parameter gA increases when the ionization degree decreases.
This means that these parameters take high values in weakly
ionized plasmas. Therefore, the larger energy Alfven soliton
formation is more possible in a plasma with weaker ioniza-
tion. This is because of the fact that here the inertia of the FIG. 2. The variation of the profile of the envelope soliton with b. We have
taken gH ¼ 2.
Alfven waves is provided by the Hall effect. Hence, the
larger amount of Hall effect is equivalent to a stronger linear
r1 r2 r3
dispersion effect and the possibility for formation of wider B1 ¼ ; (25)
solitons. In Fig. 2, the profile of the soliton is plotted for dif- r 4 r5
ferent values of plasma beta b. The soliton’s width decreases
where we have chosen a ¼ 1 and rj’s are provided in the
when plasma beta increases as long as the condition b < 1 is
Appendix. The solution (25) represents a rogue wave with
satisfied. For b > 1, the trend is reversed, the soliton’s width
constant amplitude. As shown in Fig. 3, the variation of the
increases as plasma beta increases. This different behavior
plasma parameters changes the spatial as well as temporal
may be understood from (18) that when b < 1 the plasma
pulse width. We can found out that the spatial and temporal
density is enhanced in regions of strong magnetic field, while
width of rogue wave solution increases when the Hall param-
for b > 1, the plasma density is diminished. The effect of
eter increases. As discussed above, the Hall parameter
plasma beta on Alfven solitons in fully ionized plasmas is
increases when the ionization degree decreases. Therefore,
extensively studied in Refs. 41 and 54.
wider rogue waves can occur in a plasma with weaker ioniza-
Apart from the single-soliton solutions discussed above,
tion. The rogue wave spatial and temporal width decreases as
the DNLS equation possesses another solution in the form40
the plasma parameter b increases (see Fig. 4). The possibility
of formation of ion-acoustic kinetic and dust ion-acoustic
Alfven waves in superthermal plasmas is recently proposed in
Refs. 55–57. The present paper reports the possibility for exis-
tence of rogue waves in partially ionized plasmas.

B. Soliton perturbation
In this subsection, we want to suppress the limit of Hall
effect domination that is employed in Subsection V A.
Equation (21) is not integrable in the presence of ambipolar
and ohmic terms. However, in cases where these effects are
small in comparison with the Hall effect, the dynamics of
soliton solution can be addressed using the soliton perturba-
tion approach. This method has been reviewed in Ref. 58.
This condition will be fulfilled in Earth’s lower E region and
in some heights of solar atmosphere.45 The dynamics of per-
turbed DNLS soliton can be treated by means of various
methods including the inverse scattering transform formal-
ism,53 direct perturbation method,59 and the moment
method.44,60 A solitonic form in principle persists after add-
ing a small perturbation (thus breaking integrability),
although the evolution of a quasi-soliton may be more com-
FIG. 1. The variation of the profile of the envelope soliton with Hall param- plicated and delicate to model under the action of the pertur-
eter. We have taken b ¼ 2. bation. In the following, we shall adopt the latter method to
022302-5 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

FIG. 3. The rogue wave solution (25) for different values of Hall parameter (a) gH ¼ 0.1, (b) gH ¼ 0.7, and (c) gH ¼ 2. The plasma beta is taken as b ¼ 0.1 and
the wave parameters are chosen as a1 ¼ b1 ¼ 0.5.

ð1
derive a dynamical model for the soliton characteristics. The 1
E¼ jB1 j2 dn;
method of moments is a reduction of the complete problem 2ð1  bÞgH 1
(evolution of a field that has an infinite number of degrees of ð1 " #
  jB1 j4
freedom) to the evolution of a finite set of variables for the M¼ i B1 B1n  B1 B1n  dn: (27)
pulse characteristics. 1 2ð1  bÞgH
We proceed by assuming that the ambipolar and ohmic
effects are small in comparison with Hall effects and rewrite In the presence of perturbation, the integrability of the
Eq. (21) in the new form DNLS equation breaks down and the above quantities can no
longer be conserved, instead they are evolved with time
@B1 1 @   1 @ 2 B1 according to the following relations:44
þ B1 jB1 j2 þ iagH ¼ F½B1 ; (26) ð1
@s 4ð1  bÞ @n 2 @n2 dE 1   
¼ 3
B1 F þ B1 F dn ¼ GE ; (28)
with the small perturbation term F ¼ 4gT gH ð1  bÞ2 ds 16ð1  bÞ gH 1 2

ð@ 2 B1 =@n2 Þ. This equation is called perturbed DNLS equa- ð1 "

dM 1  
tion. It is known that the unperturbed DNLS equation, on the ¼ 2
i B1n F  B1n F
other hand (i.e., Eq. (26), upon formally setting F[B1] ¼ 0) ds 4ð1  bÞ gH 1
#
has an infinite set of integrals of motion. The first two of jB1 j2 ðB1 F þ B1 F Þ
them, namely, the energy E, and the momentum M,  dn ¼ GM : (29)
2ð1  bÞgH
read44,61,62 (for a ¼ 1)
022302-6 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

FIG. 4. The rogue wave solution (25) for different values of plasma beta (a) b ¼ 0.02, (b) b ¼ 0.2, and (c) b ¼ 2. The Hall parameter is taken as gH ¼ 1 and the
wave parameters are chosen as a1 ¼ 0.6 and b1 ¼ 0.4.

 
To proceed further, we will examine the dynamics of a two 2 l 1
E¼ tan1 ; M¼ g: (31)
parameter envelope soliton solution of the DNLS equation ð1  bÞgH g 2ð1  bÞgH
that is given by44,63
Using (31) in (28) and (29) results in
2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
8g2 6 lv l2 þ g2 dg
B1 ¼ qei/ ¼ 2
exp i4   t ¼ 2ð1  bÞgH GM ; (32)
jkj cosh v  l 2g 2ð1  bÞ2 gH ds
 
0 13 dl l ð1  bÞgH GE l2 þ g2
v ¼ ½2ð1  bÞgH GM  þ : (33)
gtanh ds g 2 g
@ 2 A7
þ 3 tan1 5; (30)
jk2 j  l
Now if one inserts F into Eqs. (28) and (30) and uses (30)
and the fact that @/=@n ¼ ð3q2 =4  2lÞ=2ð1  bÞgH , the
with k2 ¼ l þ ig and v ¼ 4g½n=ð2ð1  bÞgH Þ  4ls= expressions for GE and GM can be calculated by virtue of the
ð8ð1  b2 ÞgH Þ. In the absence of perturbation, the parame- symbolic computation
ters l and g are constant; on the contrary, they change slowly   1 l 
8gT 2 2
with time in the presence of small perturbation. Substituting GE ¼ 5g þ 9l tan þ 4lg ; (34)
(30) into (27) and evaluating the integrals lead to44 g2 g2H ð1  bÞ2 g
022302-7 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

FIG. 5. Temporal evolution of the soli-

ton parameters (a) g and (b) l for dif-
ferent values of Hall parameter gH. We
have taken b ¼ gT ¼ 0.1.

"
gT   neutrals are more frequent in a plasma with weak ionization.
GM ¼ 2
496lg2 þ 752l3  3g2  30l2 Figure 8 shows the temporal evolution of the envelop soli-
2g2H ð1
 bÞ ton profile. It is clear that as time passes, the soliton gets
  #
l taller and narrower but its energy decreases from E(s ¼ 0)
 tan1 þ 96g3 þ 360gl2  15gl : (35) ¼ 53.023 to E(s ¼ 1) ¼ 48.09.
g

We integrate the system of ordinary equations (32) and VI. SUMMARY AND CONCLUSION
(33) using the Runge-Kutta scheme. The temporal evolution In the present paper, we have studied theoretically linear
of the soliton parameters l and g for different values of gH and nonlinear Alfven waves in a partially ionized plasma
and gT is shown in Figs. 5 and 6. We see that the effect of comprising electrons, ions, and neutrals, a configuration that
variation of gT on g and l is more profound than the varia- is relevant in many astrophysical, space, and laboratory
tion of gH. To clarify this statement, the soliton profile for plasmas.
different values of gH and gT is depicted in Fig. 7. It is appar- The recently proposed Hall magnetohydrodynamics
ent that at time s ¼ 1 the soliton profile does not change sub- model is used to deal with linear and nonlinear propaga-
stantially for different values of Hall parameter gH, while it tion of Alfven waves in a partially ionized plasma. By
is changed significantly when values of gT are varied. As is employing the Fourier transformation, the linear disper-
understandable from the definitions, when the ionization sion relation is obtained for Alfven waves and its complex
degree is growing, i.e., D is decreasing, the values of gH and solutions are derived. The main concern of this paper is
gT are decreasing as well. This means that the taller but nar- devoted to study of nonlinear Alfven waves. By virtue of
rower solitons can appear in a plasma with a weaker ioniza- the reductive perturbation technique, the derivative non-
tion degree. Moreover, the evolution of the soliton is faster linear Schr€odinger (DNLS) type equation is obtained that
in a plasma with a weaker ionization degree. It is from the governs the propagation of weakly nonlinear and weakly
fact that the collisions between charged particles and dispersive Alfven waves along with an ambient magnetic

FIG. 6. Temporal evolution of the soli-

ton parameters (a) g and (b) l for dif-
ferent values of the parameter gT. We
have taken b ¼ 0.1 and gH ¼ 3.
022302-8 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

FIG. 7. Variation of the soliton profile

with Hall and ambipolar coefficients.
We have taken (a) gT ¼ b ¼ 0.1 and
s ¼ 1 and (b) gH ¼ 1, b ¼ 0.1, and
s ¼ 0.3.

field. The Hall, ohmic, and ambipolar effects manifest degree decreases the spatially and temporally wide rogue
itself via various terms in this equation. waves can be excited in partially ionized plasmas. It is
To explore localized structures in the framework of found that the effect of plasma beta parameter b on soliton
the derived DNLS type equation, we have restricted our- and rogue waves is slightly different. If plasma beta
selves to two limiting cases where the Hall effect is domi- increases but the restriction b < 1 is satisfied, then the soli-
nated or the ambipolar and ohmic effects are small in ton’s width decreases. On the contrary, for b > 1 the soli-
comparison to the Hall effect. In the Hall effect domination ton’s width increases when beta increases. In the case of
regime, the evolution equation reduces to a standard DNLS rogue waves, spatial and temporal width decreases when
equation that comprises numerous solutions. The soliton plasma beta increases.
and rogue wave solutions are reported and the effect of On the other hand, when ohmic and ambipolar effects
plasma parameters on their profiles is investigated. It is are present but are small in comparison to the Hall effect, the
found that the larger energy solitons can occur in a plasma evolution equation takes the form of the perturbed DNLS
with a weaker ionization degree. Moreover, although equation. In this limit, we used the moment method to inves-
the amplitude of the rogue waves does not depend on sys- tigate the dynamics of two parameter envelope soliton.
tem parameters, its spatial and temporal width changes Using this method, two coupled evolution equations for soli-
with changing parameters. Basically, when the ionization ton parameters are derived and solved numerically using the
Runge-Kutta scheme. It is shown that the evolution of the
soliton parameters is essentially affected by ambipolar and
ohmic effects rather than the Hall effect. This fact comes
from the collision of the charged particles, especially ions
with neutrals.
In this work, we have restricted ourselves to study of the
existence and dynamics of one soliton and rogue wave solu-
tions in partially ionized plasmas. In the same way, the anal-
ysis can be outlined for multi-soliton solution and higher
order rational solutions of the DNLS equation, although the
mathematical procedure is more complex. The results of this
work are relevant for interpretation of some nonlinear
effects, which may arise in astrophysical, space, and labora-
tory partially ionized plasmas.

APPENDIX: ROGUE WAVE PARAMETERS

The parameters for rj’s in Eq. (25) are given by40

r1 ¼ 2 exp 2iða21 þ b21 Þð2sa21 þ n  2sb21 Þ ; (A1)

h
r2 ¼ b1 16b21 a21 ð4sa21 þ nÞ2 þ 16b41 ð4sb21  nÞ2
i
FIG. 8. Temporal evolution of the soliton profile for gT ¼ 0.2, b ¼ 0.01, and þ 8ib21 ðn þ 4sa21  8sb21 Þ þ 1 ; (A2)
gH ¼ 2.5, and for initial conditions l(0) ¼ 0.08 and g(0) ¼ 0.04.
022302-9 J. Borhanian and A. Rezaei Phys. Plasmas 24, 022302 (2017)

23
r3 ¼ 2 16b21 a21 ð4sa21 þ nÞ2 þ 16b41 ð4sb21  nÞ2 K. Baumg€artel, J. Geophys. Res. 104, 28295, doi:10.1029/1999JA900393
(1999).
 8ia1 b1 ðn þ 4sa21  8sb21 Þ þ 1 24
P. Shukla, B. Eliasson, and L. Stenflo, Phys. Lett. A 375, 2371 (2011).
25
K. Stasiewicz, P. K. Shukla, G. Gustafsson, S. Buchert, B. Lavraud, B.
  a1 þ 16b1 ðb41  a41 Þs  4b1 ða21 þ b21 Þn Thide, and Z. Klos, Phys. Rev. Lett. 90, 085002 (2003).
26
K. Stasiewicz, Phys. Rev. Lett. 93, 125004 (2004).
þ 16ia1 b21 ða21 þ b21 Þs  ib1  16b21 a21 ð4sa21 þ nÞ2 27
D. Laveder, T. Passot, P. Sulem, and G. Sanchez-Arriaga, Phys. Lett. A
þ 16b41 ð4sb21  nÞ2 þ 8ib1 b1 ðn þ 4sa21  8sb21 Þ þ 1 28
375, 3997 (2011).
P. K. Shukla and W. M. Moslem, Phys. Lett. A 376, 1125 (2012).
 a1 þ 16b1 ðb41  a41 Þs  4b1 ða21 þ b21 Þn 29
30
A. Rogister, Phys. Fluids 14, 2733 (1971).
E. Mjølhus, J. Plasma Phys. 16, 321 (1976).
þ 16ia1 b21 ða21 þ b21 Þs þ ib1 ; 31
K. Mio, T. Ogino, K. Minami, and S. Takeda, J. Phys. Soc. Jpn. 41, 265
(1976).
(A3) 32
K. Mio, T. Ogino, K. Minami, and S. Takeda, J. Phys. Soc. Jpn. 41, 667
(1976).
r4 ¼ a1 þ 16b1 ðb41  a41 Þs  4b1 ða21 þ b21 Þn 33
J.-I. Sakai and B. U. O. Sonnerup, J. Geophys. Res. 88, 9069, doi:10.1029/
þ 16ia1 b21 ða21 þ b21 Þs þ ib1 ; (A4) 34
JA088iA11p09069 (1983).
E. Mjølhus, J. Plasma Phys. 19, 437 (1978).
35
D. J. Kaup and A. C. Newell, J. Math. Phys. 19, 798 (1978).
r5 ¼  16b21 a21 ð4sa21 þ nÞ2  16b41 ð4sb21  nÞ2 36
T. Kawata, N. Kobayashi, and H. Inoue, J. Phys. Soc. Jpn. 46, 1008
2 (1979).
þ 8ib21 ðn þ 4sa21  8sb21 Þ  1 : (A5) 37
Y. Ichikawa, K. Konno, M. Wadati, and H. Sanuki, J. Phys. Soc. Jpn. 48,
279 (1980).
Here, s stands for s=ð8ð1  bÞ2 gH Þ and n for n=ð2ð1  bÞgH Þ. 38
G.-Q. Zhou and N.-N. Huang, J. Phys. A: Math. Theor. 40, 13607 (2007).
39
The a1 and b1 are arbitrary parameters. 40
X. Chen and W. Lam, Phys. Rev. E 69, 066604 (2004).
S. Xu, J. He, and L. Wang, J. Phys. A: Math. Theor. 44, 305203 (2011).
41
1
E. Mjølhus and J. Wyller, J. Plasma Phys. 40, 299 (1988).
J. H. Piddington, Mon. Not. R. Astron. Soc. 116, 314 (1956). 42
B. Deconinck, P. Meuris, and F. Verheest, J. Plasma Phys. 50, 445
2
B. D. Pontieu and G. Haerendel, Astron. Astrophys. 338, 729 (1998). (1993).
3
B. D. Pontieu, P. C. Martens, and H. S. Hudson, Astrophys. J. 558, 859 43
B. Deconinck, P. Meuris, and F. Verheest, J. Plasma Phys. 50, 457
(2001). (1993).
4
L. Chen and F. Zonca, Rev. Mod. Phys. 88, 015008 (2016). 44
G. Sanchez-Arriaga, Phys. Plasmas 17, 082313 (2010).
5
N. F. Cramer, The Physics of Alfven Waves (Wiley-VCH, Berlin, 2001). 45
B. P. Pandey and M. Wardle, Mon. Not. R. Astron. Soc. 385, 2269 (2008).
6
T. Passot and P.-L. Sulem, Nonlinear MHD Waves and Turbulence 46
M. L. Khodachenko, T. D. Arber, H. O. Rucker, and A. Hanslmeier,
(Springer, Berlin, Heidelberg, 1999). Astron. Astrophys. 422, 1073 (2004).
7
J. E. Leake, T. D. Arber, and M. L. Khodachenko, Astron. Astrophys. 442, 47
J. Vranjes, S. Poedts, B. P. Pandey, and B. D. Pontieu, Astron. Astrophys.
1091 (2005). 478, 553 (2008).
8 48
P. Forteza, R. Oliver, J. L. Ballester, and M. L. Khodachenko, Astron. M. Carbonell, P. Forteza, R. Oliver, and J. L. Ballester, Astron. Astrophys.
Astrophys. 461, 731 (2007). 525, A60 (2010).
9 49
P. Forteza, R. Oliver, and J. L. Ballester, Astron. Astrophys. 492, 223 S. Barcelo, M. Carbonell, and J. L. Ballester, Astron. Astrophys. 525, A60
(2008). (2011).
10 50
B. P. Pandey, J. Vranjes, and V. Krishan, Mon. Not. R. Astron. Soc. 386, T. V. Zaqarashvili, M. L. Khodachenko, and H. O. Rucker, Astron.
1635 (2008). Astrophys. 534, A93 (2011).
11 51
J. Vranjes and M. Kono, Phys. Plasmas 21, 012110 (2014). T. V. Zaqarashvili, M. Carbonell, J. L. Ballester, and M. L. Khodachenko,
12
T. V. Zaqarashvili, M. L. Khodachenko, and H. O. Rucker, Astron. Astron. Astrophys. 544, A143 (2012).
52
Astrophys. 529, A82 (2011). Y. H. Ichikawa, Phys. Scr. 20, 296 (1979).
13 53
R. Soler, M. Carbonell, J. L. Ballester, and J. Terradas, Astrophys. J. 767, J. Wyller and E. Mjølhus, Physica D 13, 234 (1984).
54
171 (2013). S. R. Spangler and J. P. Sheerin, J. Plasma Phys. 27, 193 (1982).
14 55
T. V. Zaqarashvili, M. L. Khodachenko, and R. Soler, Astron. Astrophys. A. S. Bains, B. Li, and L.-D. Xia, Phys. Plasmas 21, 032123 (2014).
56
549, A113 (2013). N. S. Saini, J. Plasma Phys. 81, 905810316 (2015).
15 57
P. K. Shukla, L. Stenflo, R. Bingham, and B. Eliasson, Plasma Phys. N. Kaur and N. Saini, Astrophys. Space Sci. 361, 331 (2016).
58
Controlled Fusion 46, B349 (2004). Y. S. Kivshar and B. A. Malomed, Rev. Mod. Phys. 61, 763 (1989).
16 59
C. Lashmore-Davies, Phys. Fluids 19, 587 (1976). C. Xue-Ping, L. Ji, and H. Ping, Commun. Theor. Phys. 50, 501 (2008).
17 60
S. Chapeaux, T. Passot, and P. L. Sulem, J. Plasma Phys. 58, 665 (1997). A. Biswas, Phys. Plasmas 12, 022306 (2005).
18 61
E. Mjølhus and J. Wyller, Phys. Scr. 33, 442 (1986). A. Maimistov and A. Basharov, Nonlinear Optical Waves (Springer,
19
P. K. Shukla and L. Stenflo, Phys. Scr. T60, 32 (1995). Netherlands, 1999), p. 352.
20 62
G. Brodin and L. Stenflo, Phys. Scr. 37, 89 (1988). V. M. Lashkin, Phys. Rev. E 74, 016603 (2006).
21 63
S. Dorfman and T. A. Carter, Phys. Rev. Lett. 116, 195002 (2016). T. Hada, R. L. Hamilton, and C. F. Kennel, Geophys. Res. Lett. 20, 779,
22
E. Mjølhus, Phys. Scr. T122, 135 (2006). doi:10.1029/93GL00073 (1993).