1: \documentclass{article}
2:
3: \usepackage{amssymb}
4: \usepackage{amsmath}
5: \usepackage{epsfig}
6: \newenvironment{proof}[1][Proof]{\noindent\textbf{#1.}}{\
7: \rule{0.5em}{0.5em}} \topmargin=-1cm \textwidth=17cm
8: \textheight=23cm \evensidemargin=0cm \oddsidemargin=0cm
9: \newcommand{\newsubsection}[1]{ \setcounter{equation}{0} \section{ #1} }
10: \renewcommand{\theequation}{ \arabic{section}. \arabic{equation}}
11:
12: \begin{document}
13:
14: \title{\bf Cold Plasma Dispersion Relations in the Vicinity of a Schwarzschild Black Hole Horizon}
15:
16: \author{M. Sharif \thanks{msharif@math.pu.edu.pk} and Umber Sheikh\\
17: %EndAName
18: \textit{\ {\small Department of Mathematics,}}\\
19: \textit{{\small University of the Punjab, Lahore 54590,
20: Pakistan}}}
21: \date{}
22: \maketitle
23: \begin{abstract}
24: We apply the ADM 3+1 formalism to derive the general relativistic
25: magnetohydrodynamic equations for cold plasma in spatially flat
26: Schwarzschild metric. Respective perturbed equations are
27: linearized for non-magnetized and magnetized plasmas both in
28: non-rotating and rotating backgrounds. These are then Fourier
29: analyzed and the corresponding dispersion relations are obtained.
30: These relations are discussed for the existence of waves with
31: positive angular frequency in the region near the horizon. Our
32: results support the fact that no information can be extracted from
33: the Schwarzschild black hole. It is concluded that negative phase
34: velocity propagates in the rotating background whether the black
35: hole is rotating or non-rotating.
36: \end{abstract}
37:
38: {\bf Keywords }: 3+1 formalism, perturbations, dispersion
39: relations.
40:
41: \section{Introduction}
42:
43: General Relativity (GR) serves both as a theory of space and time
44: as well as a theory of gravitation. This theory has tremendous
45: philosophical implications and has given rise to exotic new
46: physical concepts like black holes and dark matter. All the
47: predictions and applications by different scientists make the
48: theory a spectacularly successful one. The theory of GR is not
49: complete unless it includes the hard reality of experiments and of
50: astronomical observations. To explain the phenomena observed by
51: astronomers in physical terms we have astrophysical relativity.
52: This serves as a tool to gauge the properties of large scale
53: structures for which gravitation plays a significant role.
54:
55: It is believed that the compact objects have strong gravitational
56: fields near their surfaces [1]. The study of general relativistic
57: effects on electromagnetic processes which take place in the
58: vicinity of compact objects is of theoretical interest in its own
59: right. However, it is also important to interpret and understand
60: the still puzzling signals we receive from these astronomical
61: objects, including neutron stars and black holes. All the massive
62: stars, neutron stars and black holes carry energy flux. Due to
63: this flux, a relatively large magnetic field is produced. In order
64: to understand the phenomena of the compact objects and many more
65: like these, we have a developed theory of magnetized plasma called
66: theory of magnetohydrodynamics (MHD). To study the magnetosphere
67: of massive black holes, the strength of gravity demands the theory
68: of general relativistic magnetohydrodynamics (GRMHD). GRMHD
69: equations help us to study stationary configurations and dynamic
70: evolution of conducting fluid in a magnetosphere. These equations
71: (Maxwell's equations, Ohm's law, mass, momentum and energy
72: conservation equations) are required to investigate various
73: aspects of the interaction of relativistic gravity with plasma's
74: magnetic field. The successful study of plasmas in the black hole
75: environment is important because a successful study of waves will
76: be of great value in aiding the observational identification of
77: black hole candidates. It is well-known [2-4] that an isolated
78: black hole cannot have an electromagnetic field unless it is
79: endowed with a net electric charge. Since a collapsed object can
80: have a very strong effect on an electromagnetic field, it is of
81: interest to determine this effect when a black hole is placed in
82: an external electromagnetic field using GRMHD equations.
83:
84: The 3+1 formalism (also called ADM formalism) was originally
85: developed to study the quantization of gravitational field by
86: Arnowitt et al. [5]. Tarfulea [6] worked on numerical solutions of
87: Einstein field equations by applying constraint preserving
88: boundary conditions. Thorne and Macdonald [7,8] extended the
89: formulation to electromagnetic fields. Thorne et al. [9] described
90: its further applications for the black hole theory. The black hole
91: theory was considered by Holcomb and Tajima [10], Holcomb [11] and
92: Dettmann et al. [12] to investigate some properties of wave
93: propagation in Friedmann universe. Khanna [13] derived the MHD
94: equations describing the two component plasma theory of Kerr black
95: hole in 3+1 split. Ant\'{o}n et al. [14] used 3+1 formalism to
96: investigate various test simulations and discussed
97: magneto-rotational instability of accretion disks. Anile [15]
98: worked on propagation and stability of relativistic shocks and
99: relativistic simple waves in magneto-fluids. He also linearized
100: the electromagnetic wave theory in cold relativistic plasma.
101: Komissarov [16] discussed Blendford-Znajec monopole solution using
102: 3+1 formalism in black hole electrodynamics. In another paper
103: [17], he formulated and solved time dependent, force free,
104: degenerate electrodynamics as a hyperbolic system of conservations
105: laws for pulsars. He also discussed the waves mode and one
106: dimensional numerical scheme based on linear and exact Riemann
107: solvers. Zhang [18] formulated the black hole theory for
108: stationary symmetric GRMHD using $3+1$ formalism. The same author
109: [19] applied this formulation to the linearized waves propagating
110: in two dimensions using perfect GRMHD fluid for the cold plasma
111: state in the vicinity of Kerr black hole magnetosphere. He
112: investigated the behavior of perturbations in graphic form which
113: described the response of the magnetosphere to oscillatory driving
114: forces near the plasma-injection plane. Tsikarishvili et al. [20]
115: considered 3+1 general relativistic hydrodynamical equations
116: describing strongly magnetized collisionless plasma with an
117: anisotropic pressure and obtained the equations of state in
118: ultra-relativistic limit. In his recent paper, Mofiz [21]
119: discussed gravitational waves in magnetized plasmas.
120:
121: The formalism for gravitational perturbations away from a
122: Scwarzschild background was developed by Regge and Wheeler [22].
123: It was extended by Zerilli [23] who proved that the perturbations
124: corresponding to a change in mass, the angular momentum and
125: charge, of the Schwarzschild black hole are well-behaved. The
126: decay of non well-behaved perturbations was investigated by Price
127: [24]. The quasi-static electric problem was solved by Hanni and
128: Ruffini [25] who showed that the lines of force diverge at the
129: horizon for the observer at infinity. Wald [26] derived the
130: solution for electromagnetic field occurring when a stationary
131: axisymmetric black hole is placed in a uniform magnetic field
132: aligned along the symmetry axis of black hole. A linearized
133: treatment of plasma waves for special relativistic formulation of
134: Schwarzschild black holes was developed by Sakai and Kawata [27].
135: This treatment was extended to waves in general relativistic two
136: component plasma in 3+1 ADM formalism, propagating in radial
137: direction by Buzzi et al. [28]. They investigated the one
138: dimensional radial propagation of transverse and longitudinal
139: waves close to the Schwarzschild horizon.
140:
141: This paper is devoted to study the dynamical magnetosphere of the
142: Schwarzschild black hole. The model spacetime is the Schwarzschild
143: metric in Rindler coordinates which preserves the key features of
144: the Schwarzschild geometry. This investigation supports the
145: well-known fact that the information from the black hole cannot be
146: extracted. We apply 3+1 form of the GRMHD equations to this
147: spacetime by using the cold plasma state. Afterwards, we linearize
148: the equations to study perturbations of the magnetosphere. These
149: equations are then Fourier analyzed to get the sets of ordinary
150: differential equations which are comparatively easier to handle
151: than the original GRMHD partial differential equations. The
152: determinant of these equations are solved [29] to get dispersion
153: relations. Since the determinant of the coefficients of the
154: Fourier analyzed equations cannot be solved analytically, we solve
155: it with the help of software \emph{Mathematica} for special cases
156: and wave numbers are obtained in terms of angular frequency and
157: lapse function. We use wave number to calculate phase and group
158: velocities, refractive index and change in refractive index w.r.t.
159: angular frequency. The graphs of each of these quantities are
160: given and discussed.
161:
162: The paper has been organized as follows. In the next section, we
163: shall review the general results for the 3+1 formalism. The GRMHD
164: equations in the Schwarzschild black hole magnetosphere for the
165: cold plasma are also given. In section $3$, we restrict the
166: problem only to the non-rotating background. Section $4$ is
167: devoted to the rotating non-magnetized plasma and section $5$ is
168: furnished with the dispersion relations for the rotating
169: magnetized plasma. In the last section, we shall summarize and
170: discuss the results.
171:
172: \section{3+1 Spacetime Modelling}
173:
174: In the $3+1$ formalism, the line element of the spacetime can be
175: written as [19]
176: \begin{equation}
177: ds^2=-\alpha^2dt^2+\gamma_{ij}(dx^i+\beta^idt)(dx^j+\beta^jdt),
178: \end{equation}
179: where $\alpha$ (lapse function), $\beta^i$ (shift vector) and
180: $\gamma_{ij}$ (spatial metric) are functions of the coordinates
181: $t,~x^i$. A natural observer, associated with this spacetime
182: called fiducial observer (FIDO), has four velocity \textbf{n}
183: perpendicular to the hypersurfaces of constant time $t$ and is
184: given by
185: \begin{equation}
186: \bf{n}=\frac{1}{\alpha}(\frac{\partial}{\partial
187: t}-\beta^i\frac{\partial}{\partial x^i}).
188: \end{equation}
189: Throughout the paper, geometrized units with $G=c=1$ will be used.
190: Vectors and tensors living in four-dimensional spacetime will be
191: denoted by boldface italic letters such as FIDO's four-velocity
192: \textbf{n}. The three-dimensional tensors are distinguished from
193: vectors by a dyad over the letter, such as three-dimensional
194: metric $\overleftrightarrow{\gamma}$. All vector analysis
195: notations such as gradient, curl and vector product will be those
196: of the three dimensional absolute space with three metric
197: $\overleftrightarrow{\gamma}$. The determinant of three metric is
198: denoted as $g$: $g\equiv det|\gamma_{ij}|.$ Latin letters $i, j,
199: k,...$ represent indices in absolute space and run from 1 to 3.
200:
201: The Schwarzschild line element in Rindler coordinates [9]
202: \begin{equation}
203: ds^2=-\alpha^2(z)dt^2+dx^2+dy^2+dz^2
204: \end{equation}
205: is the best approximation of this spacetime in Cartesian
206: coordinates. Since the Schwarzschild black hole is non-rotating,
207: hence the shift vector $\beta=0$. This model is an analogue of the
208: Schwarzschild spacetime with $z$, $x$ and $y$ as radial $r$, axial
209: $\phi$ and poloidal $\theta$ directions respectively. $\alpha$ is
210: the lapse function which depends on $r$ in the Shwarzschild
211: metric, whereas it depends on $z$ in this analogue. In Rindler
212: coordinates, $\alpha=\frac{z}{2r_H}$, where $r_H$ is the radius of
213: the Schwarzschild black hole. This function vanishes at the
214: horizon which we can place at $z=0$ and it increases monotonically
215: as $z$ increases from $0$ to $\infty.$ The associated FIDO's
216: four-velocity vector reduces to
217: $\textbf{n}=\frac{1}{\alpha}\frac{\partial}{\partial t}$. The
218: constant time surfaces have spatial geometry $ds^2=dx^2+dy^2+dz^2$
219: which gives a good approximation to the Schwarzschild geometry
220: near the horizon.
221:
222: \subsection{The 3+1 Split: Laws of Physics}
223:
224: The real power of 3+1 viewpoint lies in its strong similarity to a
225: physicists with non-general relativistic experience. The 3+1 laws
226: of electrodynamics are formulated with an eye towards emphasizing
227: this similarity. However, there are real unavoidable and crucial
228: new features of these familiar laws that enter electrodynamics in
229: the vicinity of a black hole. The most important of these is the
230: issue of two times, the universal time $t$ and the FIDO's time
231: $\tau$. Both times are related to each other by
232: $\alpha\equiv\frac{d\tau}{dt}$. For the Schwarzschild black hole,
233: the Maxwell equations in 3+1 formalism take the form [9]
234: \begin{eqnarray}
235: \nabla.\textbf{B}&=&0,\\
236: \frac{\partial\textbf{B}}{\partial
237: t}&=&-\nabla\times(\alpha\textbf{E}),\\
238: \nabla.\textbf{E}&=&4\pi\rho_e,\\
239: \frac{\partial\textbf{E}}{\partial
240: t}&=&\nabla\times(\alpha\textbf{B})-4\pi \alpha \textbf{j},
241: \end{eqnarray}
242: where $\textbf{B}$ and $\textbf{E}$ are the FIDO measured magnetic
243: and electric fields, $\textbf{j}$ and $\rho_e$ are electric
244: current and electric charge density respectively. Under the
245: perfect MHD condition, i.e.,
246: \begin{equation}
247: \textbf{E}+\textbf{V}\times\textbf{B}=0,
248: \end{equation}
249: where {\textbf V} is the velocity of the fluid measured by the
250: FIDO, there can be no electric field in fluid's rest frame. For
251: perfect MHD, the equations of evolution of magnetic field [18]
252: become
253: \begin{equation}
254: \frac{D\textbf{B}}{D\tau}+\frac{1}{\alpha}(\textbf{B}.\nabla)(\beta-\alpha\textbf{V})
255: +\left\{\theta+\frac{1}{\alpha}\nabla.(\alpha\textbf{V})\right\}\textbf{B}=0,
256: \end{equation}
257: where
258: \begin{equation*}
259: \frac{D}{D\tau}\equiv\frac{d}{d\tau}+\textbf{V}.\nabla
260: =\frac{1}{\alpha}\left\{\frac{\partial}{\partial
261: t}+(\alpha\textbf{V}-\beta).\nabla\right\}.
262: \end{equation*}
263: The expansion rate of FIDO's four-velocity is
264: \begin{equation*}
265: \theta=\frac{1}{\alpha}(\frac{\dot{g}}{2g}-\nabla.\beta).
266: \end{equation*}
267: Here dot indicates partial derivative with respect to time $t$.
268: The FIDO's measured rate of change of any three dimensional vector
269: in absolute space (orthogonal to \textbf{n}) is
270: \begin{equation*}
271: \frac{d}{d \tau}\equiv \frac{1}{\alpha}(\frac{\partial}{\partial t
272: }-\beta.\nabla).
273: \end{equation*}
274: Local conservation law of rest-mass according to FIDO is given by
275: [18]
276: \begin{equation*}
277: \frac{D\rho_0}{D\tau}+\rho_0\gamma^2\textbf{V}.\frac{D\textbf{V}}{D\tau}+
278: \frac{\rho_0}{\alpha}\left\{\frac{\dot{g}}{2g}+\nabla.(\alpha\textbf{V}-\beta)\right\}=0,
279: \end{equation*}
280: or
281: \begin{equation}
282: \frac{D\rho_0}{D\tau}+\rho_0\gamma^2\textbf{V}.\frac{D\textbf{V}}{D\tau}+
283: \rho_0\left\{\theta+\frac{1}{\alpha}\nabla.(\alpha\textbf{V})\right\}=0
284: \end{equation}
285: with $\rho_0$ as the rest mass density. FIDO measured law of force
286: balance equation can be written as
287: \begin{eqnarray}
288: &&\left\{\left(\rho_0\gamma^2\mu+\frac{\textbf{B}^2}{4\pi}\right)\gamma_{ij}
289: +\rho_0\gamma^4\mu
290: V_iV_j-\frac{1}{4\pi}B_iB_j\right\}\frac{DV^j}{D\tau}
291: +\rho_0\gamma^2V_i\frac{D\mu}{D\tau}
292: -\left(\frac{\textbf{B}^2}{4\pi}\gamma_{ij}
293: -\frac{1}{4\pi}B_iB_j\right){V^j}_{|k}V^k\nonumber\\
294: && =-\rho_0\gamma^2\mu \left\{a_i-\frac{1}{\alpha}\beta_{j|i}V^j
295: -(\pounds_t\gamma_{ij})V^j\right\}-p_{|i}+\frac{1}{4\pi}(\textbf{V}\times
296: \textbf{B})_i \nabla.(\textbf{V}\times \textbf{B})
297: -\frac{1}{8\pi\alpha^2}(\alpha
298: \textbf{B})^2_{|i}\nonumber\\
299: &&+\frac{1}{4\pi\alpha}(\alpha
300: B_i)_{|j}B^j-\frac{1}{4\pi\alpha}(\textbf{B}\times\{\textbf{V}\times
301: [\nabla \times(\alpha \textbf{V} \times
302: \textbf{B})-(\textbf{B}.\nabla ) \beta]+(\textbf{V} \times
303: \textbf{B}). \nabla \beta \})_i
304: \end{eqnarray}
305: with $p$, the pressure. Notice that the acceleration,
306: $\textbf{a}=\frac{\nabla \alpha}{\alpha}$, is a function of time
307: lapse in 3+1 formalism.
308:
309: For the Schwarzschild black hole with perfect MHD assumption given
310: by Eq.(2.8), the equations of evolution of magnetic field (2.5)
311: and (2.9), the mass conservation law (2.10) and the momentum
312: conservation law (2.11) take the following forms respectively
313: \begin{eqnarray}
314: &&\frac{\partial \textbf{B}}{\partial t}=-\nabla
315: \times(\alpha \textbf{V}\times \textbf{B}),\\
316: &&\frac{\partial \textbf{B}}{\partial t}+(\alpha
317: \textbf{V}.\nabla)\textbf{B}-(\textbf{B}.\nabla)(\alpha \textbf{V}
318: )+(\nabla.(\alpha \textbf{V}))\textbf{B}=0,\\
319: &&\frac{\partial (\rho_0 \mu)}{\partial t}+\{(\alpha
320: \textbf{V}).\nabla\}(\rho_0 \mu)+\rho_0 \mu
321: \gamma^2\textbf{V}.\frac{\partial \textbf{V}}{\partial t}+\rho_0
322: \mu \gamma^2\textbf{V}.(\alpha \textbf{V}.\nabla)\textbf{V}+
323: \rho_0 \mu \{\nabla.(\alpha\textbf{V})\}=0,\\
324: &&\left\{\left(\rho_0\mu
325: \gamma^2+\frac{\textbf{B}^2}{4\pi}\right)\delta_{ij}+\rho_0\mu
326: \gamma^4
327: V_iV_j-\frac{1}{4\pi}B_iB_j\right\}\left(\frac{1}{\alpha}\frac{\partial}{\partial
328: t}+\textbf{V}.\nabla\right)V^j-\left(\frac{\textbf{B}^2}{4\pi}\delta_{ij}
329: -\frac{1}{4\pi}B_iB_j\right){V^j}_{,k}V^k\nonumber\\
330: &&+\rho_0 \gamma^2 V_i\left\{\frac{1}{\alpha}\frac{\partial
331: \mu}{\partial t}+(\textbf{V}.\nabla)\mu\right\}=-\rho_0\mu
332: \gamma^2 a_i-p_{,i}+ \frac{1}{4\pi}(\textbf{V}\times \textbf{B})_i
333: \nabla.(\textbf{V}\times \textbf{B})
334: -\frac{1}{8\pi\alpha^2}(\alpha \textbf{B})^2_{,i}\nonumber\\
335: &&+\frac{1}{4\pi\alpha}(\alpha
336: B_i)_{,j}B^j-\frac{1}{4\pi\alpha}[\textbf{B}\times\{\textbf{V}
337: \times (\nabla \times (\alpha \textbf{V} \times \textbf{B}))\}]_i.
338: \end{eqnarray}
339: Eqs.(2.12)-(2.15) are the perfect GRMHD equations for the
340: Schwarzschild black hole. In the rest of the paper, we shall
341: analyze these equations using perturbation and Fourier analysis
342: procedures.
343:
344: \subsection{Cold Plasma and GRMHD Equations}
345:
346: Theoretical modelling of the moving plasmas neglects the thermal
347: and gravity effects (such as the pressure forces acting on plasma
348: may not be as important as electromagnetic and centrifugal
349: forces). This plasma is called cold relativistic plasma. This is
350: the simplest closed system and contains only the equations of
351: conservation of mass and momentum. The highest moment of this
352: system, i.e., the kinematic pressure dyad is taken to be zero.
353: This model can be used in the study of small amplitude
354: electromagnetic waves propagating in plasmas with phase velocity
355: much larger than the thermal velocity of the particles.
356:
357: In hydrodynamic treatment, the plasma is represented as a perfect
358: fluid which has no viscosity and no heat conduction. In Rindler
359: analogue of the Schwarzschild spacetime, we assume that the system
360: of equations for perfect MHD is enclosed by the cold plasma. This
361: state can be expressed by the following equation [19]
362: \begin{equation}
363: \mu=\frac{\rho}{\rho_0}=constant.
364: \end{equation}
365: We note that the cold plasma has vanishing thermal pressure and
366: vanishing thermal energy. Here $\rho$ is the mass density of the
367: fluid.
368:
369: Using Eq.(2.16), the GRMHD Eqs.(2.12)-(2.15) become
370: \begin{eqnarray}
371: &&\frac{\partial \textbf{B}}{\partial t}=\nabla \times(\alpha
372: \textbf{V} \times \textbf{B}),\\
373: &&\frac{\partial \textbf{B}}{\partial t}+(\alpha
374: \textbf{V}.\nabla)\textbf{B}-(\textbf{B}.\nabla)(\alpha
375: \textbf{V})+\textbf{B}\nabla.(\alpha \textbf{V})=0,\\
376: &&\frac{\partial \rho}{\partial t}+(\alpha \textbf{V}.\nabla)\rho
377: +\rho\gamma^2\textbf{V}.\frac{\partial \textbf{V}}{\partial
378: t}+\rho\gamma^2\textbf{V}.(\alpha
379: \textbf{V}.\nabla)\textbf{V}+ \rho \nabla.(\alpha\textbf{V})=0,\\
380: &&\left\{\left(\rho \gamma^2+\frac{\textbf{B}^2}{4
381: \pi}\right)\delta_{ij} +\rho
382: \gamma^4V_iV_j-\frac{1}{4\pi}B_iB_j\right\}\left(\frac{1}{\alpha}\frac{\partial}{\partial
383: t}+\textbf{V}.\nabla
384: \right)V^j-\left(\frac{\textbf{B}^2}{4\pi}\delta_{ij}-\frac{1}{4
385: \pi}B_i B_j\right)V^j_{,k} V^k+\rho \gamma^2 a_i \nonumber\\
386: &&\frac{1}{4 \pi} (\textbf{V}\times \textbf{B})_i
387: \nabla.(\textbf{V}\times \textbf{B}) -\frac{(\alpha
388: \textbf{B})^2_{,i}}{8 \pi \alpha^2} +\frac{(\alpha B_i)_{,j}
389: B^j}{4 \pi \alpha}-\frac{1}{4 \pi
390: \alpha}[\textbf{B}\times\{\textbf{V} \times (\nabla \times (\alpha
391: \textbf{V} \times \textbf{B}))\}]_i.
392: \end{eqnarray}
393: The perturbed flow in the magnetosphere shall be characterized by
394: its velocity \textbf{V}, magnetic field \textbf{B} (as measured by
395: the FIDO) and fluid density $\rho$. The first order perturbations
396: in the above mentioned quantities are denoted by
397: $\delta\textbf{V},~\delta \textbf{B}$ and $\delta \rho$.
398: Consequently, the perturbed variables will take the following
399: form:
400: \begin{equation}
401: \textbf{B}=\textbf{B}^0+B\textbf{b},~
402: \textbf{V}=\textbf{V}^0+\textbf{v},~ \rho=\rho^0+\rho
403: \tilde{\rho},
404: \end{equation}
405: where $\textbf{B}^0,~\textbf{V}^0$ and $\rho^0$ are unperturbed
406: quantities. Since waves can propagate in $z$-direction due to
407: gravitation with respect to time $t$, the perturbed quantities
408: depend on $z$ and $t$.
409:
410: \section{Non-Rotating Background}
411:
412: In this section, we shall solve GRMHD equations by taking
413: non-rotating background. Notice that this makes no difference
414: whether we take magnetic field zero or non-zero. The relative
415: assumptions for this background are given. The perturbations and
416: Fourier analyze techniques are used to reduce GRMHD equations to
417: ordinary differential equations. The numerical solutions of
418: dispersion relations are given in the form of graphs.
419:
420: \subsection{Relative Assumptions}
421:
422: For non-rotating background, the magnetosphere has the perturbed
423: flow only along $z$-axis. The fluid four-velocity measured by FIDO
424: is described by a spatial vector field lying along the $z$-axis
425: and is given by
426: \begin{equation}{\setcounter{equation}{1}}
427: \textbf{V}=u(z)\textbf{e}_\textbf{z}.
428: \end{equation}
429: The Lorentz factor $\gamma=\frac{1}{\sqrt{1-\textbf{V}^2}}$ which
430: takes the form
431: \begin{equation}
432: \gamma=\frac{1}{\sqrt{1-u^2}}.
433: \end{equation}
434: FIDO measured magnetic field is given by
435: \begin{equation}
436: \textbf{B}=B(z)\textbf{e}_\textbf{z}.
437: \end{equation}
438: The following notations for the perturbed quantities will be used
439: \begin{eqnarray}
440: \textbf{b}&\equiv& \frac{ \delta \textbf{B}}{B}
441: =\textbf{b}_\textbf{z}(t,z)\textbf{e}_\textbf{z},\nonumber\\
442: \textbf{v}&\equiv& \delta \textbf{V}=\textbf{v}
443: _\textbf{z}(t,z)\textbf{e}_\textbf{z},\nonumber\\
444: \tilde{\rho}&\equiv&\frac{\delta \rho}{\rho}=\tilde{\rho}(t,z).
445: \end{eqnarray}
446: We also assume that the perturbations have harmonic space and time
447: dependence, i.e.,
448: \begin{equation}\tilde{\rho},~\textbf{b},~
449: \textbf{v}\sim e^{-i (\omega t-k z)}.
450: \end{equation}
451: Thus the perturbed variables can be expressed as follows:
452: \begin{eqnarray}
453: \tilde{\rho}(t,z)=c_1e^{-\iota (\omega t-kz)},\nonumber\\
454: \textbf{v}_\textbf{z}(t,z)=c_2e^{-\iota (\omega t-kz)},\nonumber\\
455: \textbf{b}_\textbf{z}(t,z)=c_3e^{-\iota (\omega t-kz)},
456: \end{eqnarray}
457: where $c_1,~c_2,~c_3$ are arbitrary constants.
458:
459: \subsection{Perturbation Equations}
460:
461: We use linear perturbation to write down the GRMHD
462: Eqs.(2.17)-(2.20) by using Eq.(2.21)
463: \begin{eqnarray}
464: &&\frac{\partial (\delta \textbf{B})}{\partial t}=\nabla
465: \times(\alpha \textbf{v} \times \textbf{B})
466: +\nabla \times (\alpha \textbf{V} \times \delta \textbf{B}),\\
467: &&\nabla. (\delta \textbf{B})=0,\\
468: &&(\frac{1}{\alpha}\frac{\partial}{\partial t}+\textbf{V}.\nabla
469: )\delta \rho + \rho \gamma^2 \textbf{V}.(\frac{1}{\alpha}
470: \frac{\partial}{\partial t}+ \textbf{V}.\nabla)\textbf{v} -
471: \frac{\delta \rho}{\rho}(\textbf{V}.\nabla)\rho +\rho (\nabla.\textbf{v})\nonumber\\
472: &&=-2\rho \gamma^2 (\textbf{V}.\textbf{v})( \textbf{V}. \nabla) ln
473: \gamma -\rho \gamma^2 (\textbf{V}.\nabla \textbf{V}).\textbf{v}
474: +\rho (\textbf{v}.\nabla ln u),\\
475: &&\{(\rho\gamma^2+\frac{\textbf{B}^2}{4
476: \pi})\delta_{ij}+\rho\gamma^4 V_i V_j-\frac{1}{4 \pi}B_i
477: B_j\}\frac{1}{\alpha} \frac{\partial v^j}{\partial t} +\frac{1}{4
478: \pi}[\textbf{B} \times \{\textbf{V} \times
479: \frac{1}{\alpha}\frac{\partial(\delta
480: \textbf{B})}{\partial t}\}]_i\nonumber\\
481: &&+\rho \gamma^2v_{i,j}V^j +\rho \gamma^4 V_i v_{j,k} V^j V^k
482: -\frac{1}{4 \pi \alpha}\{(\alpha
483: \delta B_i)_{,j}-(\alpha \delta B_j)_{,i}\} B^j\nonumber\\
484: &&=-\gamma^2\{\delta \rho+2 \rho \gamma^2
485: (\textbf{V}.\textbf{v})\}a_i +\frac{1}{4 \pi \alpha}\{(\alpha
486: B_i)_{,j}-(\alpha B_j)_{,i}\}\delta B^j \nonumber\\
487: &&-\rho\gamma^4 (v_i V^j + v^j V_i)V_{k,j}V^k- \gamma^2\{\delta
488: \rho V^j +2 \rho \gamma^2 (\textbf{V}.\textbf{v}) V^j +\rho v^j\} V_{i,j}\nonumber\\
489: &&-\gamma^4 V_i\{\delta \rho V^j+4 \rho \gamma^2
490: (\textbf{V}.\textbf{v})V^j+\rho v^j\}V_{j,k}V^k.
491: \end{eqnarray}
492: It is to be noted that the law of conservation of rest-mass [18]
493: in three dimensional hypersurface $$ \alpha \rho \gamma
494: u=constant$$ is used to obtain Eq.(3.9). This will also be used in
495: deriving the component form of these equations.
496:
497: The component form of Eqs.(3.7)-(3.10) is given as follows:
498: \begin{eqnarray}
499: &&\frac{1}{\alpha}\frac{\partial b_z}{\partial t}=0,\\
500: &&b_{z,z}=0,\\
501: &&\frac{1}{\alpha}\frac{\partial \tilde{\rho}}{\partial t}+u
502: \tilde{\rho}_{,z} +\gamma^2 u\frac{1}{\alpha} \frac{\partial
503: v_z}{\partial t}+(1+\gamma^2 u^2)v_{z,z}
504: =(1-2\gamma^2u^2)(1+\gamma^2 u^2)\frac{u'}{u}v_z,\\
505: &&\gamma^2(1+\gamma^2 u^2)\frac{1}{\alpha} \frac{\partial
506: v_z}{\partial t}+\gamma^2(1+\gamma^2 u^2)u v_{z,z}
507: =-\tilde{\rho}\gamma^2\{a_z+u(1+\gamma^2u^2)u'\}\nonumber\\
508: &&-\gamma^2\{u'(1+\gamma^2 u^2)(1+4\gamma^2 u^2)
509: +2u\gamma^2a_z)\}v_z.
510: \end{eqnarray}
511: When the above equations are Fourier analyzed by substituting
512: values from Eq.(3.6), these take the form
513: \begin{eqnarray}
514: &&-\frac{\iota \omega}{\alpha} c_3=0,\\
515: &&\iota k c_3=0,\\
516: &&c_1\left(\frac{-\iota \omega}{\alpha}+\iota k
517: u\right)+c_2\left\{(1+\gamma^2 u^2)\iota
518: k-(1-2\gamma^2u^2)(1+\gamma^2u^2)\frac{u'}{u}
519: -\frac{i\omega}{\alpha}\gamma^2 u\right\}=0,\\
520: &&c_1\gamma^2 \{a_z+uu'(1+\gamma^2 u^2)\}
521: +c_2[\gamma^2(1+\gamma^2u^2)\left(\frac{-\iota \omega}{\alpha}
522: +\iota u k\right)+\nonumber\\
523: &&+\gamma^2\{u'(1+\gamma^2 u^2)(1+4\gamma^2 u^2) +2u\gamma^2
524: a_z\}]=0.
525: \end{eqnarray}
526: Eqs.(3.15) and (3.16) show that $c_3$ is zero which implies that
527: there are no perturbations occurring in magnetic field of the
528: fluid. Thus we are left with Eqs.(3.17) and (3.18). It is
529: mentioned here that for the non-magnetized plasma, we also obtain
530: the same two equations.
531:
532: \subsection{Numerical Solutions}
533:
534: We consider time lapse $\alpha=z$ and $\rho$ as a constant. Using
535: mass conservation law in three dimensions, we obtain the value of
536: $u=\frac{1}{\sqrt{1+z^2}}$. When we use these values, the
537: determinant of the coefficients of constants $c_1$ and $c_2$ in
538: Eqs.(3.17) and (3.18) give us a complex number. Equating it to
539: zero, we obtain three values for $k$, two from the real part and
540: one from the imaginary part. Using the value of $k$, one can
541: calculate the following quantities:
542: \begin{enumerate}
543: \item \textbf{Phase Velocity ($v_p$)}: The velocity with which
544: the carrier wave in a modulated signal moves and can be
545: calculated by $\frac{\omega}{k}$.
546: \item \textbf{Refractive Index ($n$)}: A property of a material that
547: changes the speed of light, computed as the ratio of the speed of light
548: in a vacuum to the speed of light through the material.
549: Refractive index is the inverse of the phase velocity.
550: \item \textbf{Change in refractive index with respect to angular frequency
551: ($\frac{dn}{d\omega}$)}:
552: This term helps to find out whether the dispersion is normal
553: or not.
554: \item \textbf{Group Velocity ($v_g$)}: This is the speed of transmission
555: of information and/or
556: energy in a wave packet and can be calculated by the formula
557: $\frac{1}{n+\omega\frac{dn}{d\omega}}$.
558: \end{enumerate}
559: We would like to mention here that only longitudinal oscillations
560: are calculated in this case as we are using the longitudinal part
561: of the momentum equation. Thus we would discuss the longitudinal
562: waves propagating parallel to the magnetic field $\textbf{B}$ in
563: the remaining part of this section.
564:
565: The dispersion equation from real part of the matrix is of the
566: form $A(z)k^2+B(z)k\omega+C(z)\omega^2+D(z)=0.$ This leads to two
567: values of $k$ in terms of $z$ and angular frequency shown in the
568: Figures 1 and 2. These are conjugate roots of quadratic equation.
569: The real wave numbers are shown in the graph lying in the region
570: $3.5\times 10^{-7}\leq z\leq 1.41.$
571: \begin{figure}
572: \center \epsfig{file=p1a.eps,width=0.4\linewidth} \center
573: \epsfig{file=p1b.eps,width=0.8\linewidth} \center
574: \epsfig{file=p1c.eps,width=0.8\linewidth} \caption{The waves
575: decrease as they move away from the event horizon. The phase and the
576: group velocities are equal except for the waves with less angular
577: frequencies. The region shows normal dispersion.}
578: \end{figure}
579:
580: We see from the Figure 1 that angular frequency $\omega$ and wave
581: number $k$ are directly proportional to each other, i.e., the
582: change in angular velocity gives the respective change in wave
583: number. Also, the wave number decreases as $z$ increases. This
584: means that waves lose energy as we go away from the black hole
585: horizon. At the horizon, the wave number becomes very large and
586: the waves vanish due to the effect of gravity. There is an abrupt
587: increment in phase velocity when the angular frequency is near to
588: zero. The region $3.5\times 10^{-7}\leq
589: z\leq0.5,~1.5\leq\omega\leq10$ has refractive index greater than
590: one and its variation with respect to omega greater than zero.
591: This corresponds to the region of normal dispersion [30]. The
592: group and wave velocities go parallel and are increasing as we go
593: away from the horizon.
594: \begin{figure}
595: \center \epsfig{file=p2a.eps,width=0.4\linewidth} \center
596: \epsfig{file=p2b.eps,width=0.8\linewidth} \center
597: \epsfig{file=p2c.eps,width=0.8\linewidth} \caption{The waves damp as
598: they go away from the event horizon. The phase and the group
599: velocities are equal except for the waves with negligible angular
600: frequencies. The region shows normal dispersion.}
601: \end{figure}
602:
603: The Figure 2 shows that the waves are growing with the increase in
604: angular velocity, but damping occurs with as $z$ increases. For
605: $0\leq z<3.5\times10^{-7}$, the wave number goes to infinity and
606: hence there is no wave at the horizon. This shows that the event
607: horizon is the interface at which they are formed and as we go
608: away from the event horizon, the intensity of these waves decay.
609: Since the refractive index is greater than one and
610: $\frac{dn}{d\omega}$ is greater than zero, the whole region shows
611: normal dispersion.
612:
613: The dispersion equation obtained from the imaginary part of the
614: matrix is of the form $A(z)k+B(z)\omega=0.$ Hence we are with only
615: one value of $k$ given by the Figure 3.
616: \begin{figure}
617: \center \epsfig{file=p3a.eps,width=0.4\linewidth} \center
618: \epsfig{file=p3b.eps,width=0.8\linewidth} \center
619: \epsfig{file=p3c.eps,width=0.8\linewidth} \caption{The waves damp as
620: they move away from the event horizon. The phase and the group
621: velocities are same. The dispersion is not normal.}
622: \end{figure}
623:
624: The Figure 3 shows that the waves gain energy with the increase in
625: angular velocity, but lose as $z$ increases. The wave number goes
626: to infinity and hence there is no wave for $0\leq
627: z<3.5\times10^{-7}$. The group and phase velocities are showing
628: the same behavior i.e. they are increasing with an increase in
629: $z$. The refractive index is not greater than one and hence the
630: dispersion is not normal.
631:
632: \section{Rotating Non-Magnetized Background}
633:
634: When we consider non-magnetized cold plasma in rotating
635: background, i.e., $\textbf{B}=0$, the GRMHD Eqs.(2.17) and (2.18)
636: vanish. Eqs.(2.19) and (2.20) change into general relativistic
637: hydrodynamical equations given below:
638: \begin{eqnarray}{\setcounter{equation}{1}}
639: &&\frac{\partial \rho}{\partial t}+(\alpha \textbf{V}.\nabla)\rho
640: +\rho\gamma^2\textbf{V}.\frac{\partial \textbf{V}}{\partial
641: t}+\rho\gamma^2\textbf{V}.(\alpha \textbf{V}.\nabla)\textbf{V}+
642: \rho \nabla.(\alpha\textbf{V})=0,\\
643: &&(\rho \gamma^2\delta_{ij}+\rho
644: \gamma^4V_iV_j)\left(\frac{1}{\alpha}\frac{\partial}{\partial
645: t}+\textbf{V}.\nabla\right)V^j=-\rho \gamma^2 a_i.
646: \end{eqnarray}
647: For rotating background, the fluid four-velocity measured by FIDO
648: can be described in $xz$-plane as follows
649: \begin{equation}
650: \textbf{V}=V(z)\textbf{e}_\textbf{x}+u(z)\textbf{e}_\textbf{z}
651: \end{equation}
652: and the Lorentz factor $\gamma$ takes the form
653: \begin{equation}
654: \gamma=\frac{1}{\sqrt{1-u^2-V^2}}.
655: \end{equation}
656: Here we shall use the following notations for the perturbed
657: quantities
658: \begin{eqnarray}
659: \tilde{\rho}&\equiv&\frac{\delta \rho}{\rho}=\tilde{\rho}(t,z)\nonumber\\
660: \textbf{v}&\equiv& \delta \textbf{V}=\textbf{v}
661: _\textbf{x}(t,z)\textbf{e}_\textbf{x}+\textbf{v}
662: _\textbf{z}(t,z)\textbf{e}_\textbf{z}
663: \end{eqnarray}
664: and the perturbations have harmonic space and time dependence,
665: i.e.,
666: \begin{equation}\tilde{\rho},~
667: \textbf{v}\sim e^{-i (\omega t-k z)}.
668: \end{equation}
669: The perturbed variables take the form
670: \begin{eqnarray}
671: \tilde{\rho}(t,z)=c_1e^{-\iota (\omega t-kz)},\nonumber\\
672: \textbf{v}_\textbf{z}(t,z)=c_2e^{-\iota (\omega t-kz)},\nonumber\\
673: \textbf{v}_\textbf{x}(t,z)=c_3e^{-\iota (\omega t-kz)}.
674: \end{eqnarray}
675:
676: \subsection{Perturbation Equations}
677:
678: Introducing linear perturbations in Eqs.(4.1) and (4.2), it
679: follows that
680: \begin{eqnarray}
681: &&\left(\frac{1}{\alpha}\frac{\partial}{\partial
682: t}+\textbf{V}.\nabla \right)\delta \rho + \rho \gamma^2
683: \textbf{V}.\left(\frac{1}{\alpha} \frac{\partial}{\partial t}+
684: \textbf{V}.\nabla\right)\textbf{v} -
685: \frac{\delta \rho}{\rho}(\textbf{V}.\nabla)\rho +\rho (\nabla.\textbf{v})\nonumber\\
686: &&=-2\rho \gamma^2 (\textbf{V}.\textbf{v})( \textbf{V}. \nabla) ln
687: \gamma -\rho \gamma^2 (\textbf{V}.\nabla \textbf{V}).\textbf{v}
688: +\rho (\textbf{v}.\nabla ln u),\\
689: &&(\rho\gamma^2 \delta_{ij}+\rho\gamma^4 V_i V_j)\frac{1}{\alpha}
690: \frac{\partial v^j}{\partial t}+\rho \gamma^2v_{i,j}V^j +\rho
691: \gamma^4 V_i v_{j,k} V^j V^k=-\gamma^2\{\delta \rho+2 \rho
692: \gamma^2 (\textbf{V}.\textbf{v})\}a_i\nonumber\\&&
693: -\rho\gamma^4(v_i V^j + v^j V_i)V_{k,j}V^k- \gamma^2\{\delta
694: \rho V^j +2 \rho \gamma^2 (\textbf{V}.\textbf{v}) V^j +\rho v^j\} V_{i,j}\nonumber\\
695: &&-\gamma^4 V_i\{\delta \rho V^j+4 \rho \gamma^2
696: (\textbf{V}.\textbf{v})V^j+\rho v^j\}V_{j,k}V^k.
697: \end{eqnarray}
698: The component form of Eqs.(4.8) and (4.9) gives the following
699: three equations
700: \begin{eqnarray}
701: &&\frac{1}{\alpha}\frac{\partial \tilde{\rho}}{\partial t}+u
702: \tilde{\rho}_{,z}+\gamma^2 V\left(\frac{1}{\alpha} \frac{\partial
703: v_x}{\partial t}+uv_{x,z}\right)+\gamma^2 u\frac{1}{\alpha}
704: \frac{\partial v_z}{\partial t}+(1+\gamma^2 u^2)v_{z,z}\nonumber\\
705: &&=-\gamma^2 u\{(1+2\gamma^2 V^2)V'+2\gamma^2uVu'\}v_x
706: +\left\{(1-2\gamma^2u^2)(1+\gamma^2 u^2)\frac{u'}{u}-2\gamma^4u^2VV'\right\}v_z,\\
707: && \gamma^2(1+\gamma^2V^2)\left(\frac{1}{\alpha} \frac{\partial
708: v_x}{\partial t}+u v_{x,z}\right)+\gamma^4uV(\frac{1}{\alpha}
709: \frac{\partial v_z}{\partial t}+uv_{z,z})\nonumber\\
710: &&=-\tilde{\rho}\gamma^2u\{(1+\gamma^2V^2)V'+\gamma^2uVu'\}
711: -\gamma^4u\{(1+4\gamma^2V^2)uu'+4VV'(1+\gamma^2V^2)\}v_x\nonumber\\
712: &&-[\gamma^2\{(1+2\gamma^2u^2)(1+2\gamma^2V^2)-\gamma^2V^2\}V'
713: +2\gamma^4(1+2\gamma^2u^2)uVu']v_z,\\
714: &&\gamma^2(1+\gamma^2 u^2)\left(\frac{1}{\alpha} \frac{\partial
715: v_z}{\partial t}+u v_{z,z}\right)+\gamma^4uV\left(\frac{1}{\alpha}
716: \frac{\partial v_x}{\partial t}+u v_{x,z}\right)\nonumber\\
717: &&=-\tilde{\rho}\gamma^2\{a_z+(1+\gamma^2u^2)uu'+\gamma^2 u^2 VV'\}\nonumber\\
718: &&-[\gamma^4\{u^2V'(1+4\gamma^2V^2)+2V(a_z+(1+2\gamma^2u^2)uu')\}]v_x\nonumber\\
719: &&-\gamma^2[u'(1+\gamma^2 u^2)(1+4\gamma^2 u^2)
720: +2u\gamma^2\{a_z+(1+2\gamma^2u^2)VV'\}]v_z.
721: \end{eqnarray}
722: Using Eq.(4.7), the Fourier analyzed form of Eqs.(4.10)-(4.12)
723: become
724: \begin{eqnarray}
725: &&c_1\left(-\frac{\iota \omega}{\alpha}+\iota k
726: u\right)+c_2\left[-\frac{\iota
727: \omega}{\alpha}\gamma^2u+(1+\gamma^2 u^2)\iota
728: k-(1-2\gamma^2u^2)(1+\gamma^2u^2)\frac{u'}{u}+
729: 2\gamma^4u^2VV'\right]\nonumber\\
730: &&+c_3\gamma^2\left[-\frac{\iota \omega}{\alpha}V+\iota
731: kuV+\gamma^2u\{(1+2\gamma^2V^2)V'+2\gamma^2uVu'\}\right]=0,\\
732: &&c_1\gamma^2u\{(1+\gamma^2V^2)V'+\gamma^2uVu'\}+c_2\gamma^2[\left(-\frac{\iota
733: \omega}{\alpha}+\iota ku \right)\gamma^2uV\nonumber\\
734: &&+\{(1+2\gamma^2u^2)(1+2\gamma^2V^2)-\gamma^2V^2\}V'
735: +2\gamma^2(1+2\gamma^2u^2)uVu']\nonumber\\
736: &&+c_3[\left(-\frac{\iota \omega}{\alpha}+\iota ku\right)
737: \gamma^2(1+\gamma^2V^2)+\gamma^4u\{(1+4\gamma^2V^2)uu'+4VV'(1+\gamma^2V^2)\}]=0,\\
738: &&c_1\gamma^2\{a_z+(1+\gamma^2u^2)uu'+\gamma^2u^2VV'\}
739: +c_2[\gamma^2(1+\gamma^2u^2)\left(-\frac{\iota
740: \omega}{\alpha}+\iota uk\right)\nonumber\\
741: &&+\gamma^2\{u'(1+\gamma^2u^2)(1+4\gamma^2u^2)
742: +2u\gamma^2(a_z+(1+2\gamma^2u^2)VV')\}]\nonumber\\
743: &&+c_3\gamma^4\left[\left(-\frac{\iota \omega}{\alpha}+\iota u
744: k\right)uV+u^2V(1+4\gamma^2V^2)+2V\{a_z+uu'(1+2\gamma^2u^2)\}\right]=0.
745: \end{eqnarray}
746: The determinant of the coefficients of $c_1,~c_2$ and $c_3$ gives
747: a complex number. Equating it to zero yields a complex dispersion
748: relation in $k$.
749:
750: \subsection{Numerical Solutions}
751:
752: We consider time lapse $\alpha=z$ and assume that $V=u$ for the
753: sake of convenience. Using mass conservation law in three
754: dimensions (with mass density as a constant quantity) we obtain
755: $u=\frac{1}{\sqrt{2+z^2}}.$ In this section we are having modes
756: when $\textbf{B}=0$. The complex dispersion relation obtained from
757: the determinant of coefficients leads to two dispersion equations.
758: The real part gives an equation of the type
759: $A(z)k^2+B(z)k\omega+C(z)\omega^2+D(z)=0$ leading to two
760: dispersion relations shown in the Figures 4 and 5. The imaginary
761: part shows a cubic equation
762: $A(z)k^3+B(z)k^2\omega+C(z)k\omega^2+D(z)\omega^3=0.$ This
763: equation has one real and two complex conjugate solutions. The
764: Figure 6 shows the real value of $k$ obtained from this equation.
765: \begin{figure}
766: \center \epsfig{file=p4a.eps,width=0.4\linewidth} \center
767: \epsfig{file=p4b.eps,width=0.8\linewidth} \center
768: \epsfig{file=p4c.eps,width=0.8\linewidth} \caption{The waves damp
769: when they move away from the event horizon. The phase velocity is
770: less than the group velocity in most of the region. Most of the
771: region shows normal dispersion.}
772: \end{figure}
773:
774: The Figure 4 indicates that the region for which real waves exist
775: lies in $1.5\times 10^{-8}<z\leq10,~0.045<\omega\leq10$. It is
776: clear from the Figure that the region $0<\omega\leq 0.045$
777: contains evanescent waves. We note that at the horizon and nearby
778: values of $z$ i.e. in the region $0\leq z\leq 1.5\times 10^{-8}$,
779: the wave number is infinite and hence no wave exists there. The
780: wave number is decreasing as we go away from the horizon and hence
781: damping occurs. Also $n>1,~\frac{dn}{d\omega}>0$ and hence the
782: region is of normal dispersion. The phase velocity is less than
783: group velocity of waves except at some points where the refractive
784: index $n$ is positive but is less than one, as shown in the graph.
785: This means that the dispersion is not normal but anomalous at
786: these points [31], normal otherwise. At $\omega=0$ and very small
787: angular frequencies, the phase velocity is high but the group
788: velocity become infinite which is not physical.
789: \begin{figure}
790: \center \epsfig{file=p5a.eps,width=0.4\linewidth} \center
791: \epsfig{file=p5b.eps,width=0.8\linewidth} \center
792: \epsfig{file=p5c.eps,width=0.8\linewidth} \caption{The waves damp
793: while they move away from the event horizon. The dispersion is not
794: normal.}
795: \end{figure}
796:
797: The Figure 5 contains the infinite wave number in the region
798: $0\leq z<1.5\times 10^{-8}$, that is no wave present there. The
799: wave number possesses real values in the region $1.5\times
800: 10^{-8}<z\leq10,~0.05\leq\omega\leq10$. On increasing the value of
801: $z$, the waves are damping but are growing with the increase of
802: the angular frequency. In this region, refractive index $n<1$ and
803: $\frac{dn}{d\omega}<0$ which implies that the region is not of
804: normal dispersion. At $1.5\times
805: 10^{-8}<z\leq10,~0\leq\omega\leq4$, the wave number becomes
806: imaginary at several points leading to the result that evanescent
807: waves present there.
808: \begin{figure} \center
809: \epsfig{file=p6a.eps,width=0.4\linewidth} \center
810: \epsfig{file=p6b.eps,width=0.8\linewidth} \center
811: \epsfig{file=p6c.eps,width=0.8\linewidth} \caption{The waves damp as
812: they move away from the event horizon. The group velocity is greater
813: than the phase velocity. The region is of anomalous dispersion.}
814: \end{figure}
815:
816: In the Figure 6, real waves exist in the region $1.5\times
817: 10^{-8}\leq z\leq 10.$ The refractive index is greater than one,
818: $\frac{dn}{d\omega}$ is negative and thus the region is not of
819: normal dispersion. No wave is present on the event horizon because
820: the wave number is infinite there. As we go away from the horizon,
821: the wave number decreases. On increasing the angular frequency, we
822: find growing waves. Also, group velocity is greater than phase
823: velocity which indicates that the region is of anomalous
824: dispersion.
825:
826: \section{Rotating Magnetized Background}
827:
828: The GRMHD equations for magnetized cold plasma in rotating
829: background will remain the same as given by Eqs.(2.17)-(2.20).
830: Here the four-velocity of the fluid measured by FIDO can be
831: described in the same way as given in the previous section. The
832: rotating magnetic field can be expressed in the $xz$-plane as
833: \begin{equation}{\setcounter{equation}{1}}
834: \textbf{B}=B[\lambda(z)\textbf{e}_\textbf{x}+\textbf{e}_\textbf{z}],
835: \end{equation}
836: where $\lambda,~u$ and $V$ are related to each other by
837: \begin{equation}
838: V=\frac{V_F}{\alpha}+\lambda u,
839: \end{equation}
840: $V_F$ as an integration constant. The following notations for the
841: perturbed magnetic field will be used in addition to the notations
842: given by Eq.(4.5).
843: \begin{eqnarray}
844: \textbf{b}&\equiv& \frac{\delta \textbf{B}}{B}=\textbf{b}
845: _\textbf{x}(t,z)\textbf{e}_\textbf{x}+\textbf{b}
846: _\textbf{z}(t,z)\textbf{e}_\textbf{z}.
847: \end{eqnarray}
848: Since the perturbations have harmonic space and time dependence,
849: i.e., $\textbf{b}\sim e^{-i (\omega t-k z)}$, the perturbed
850: variable can be expressed as follows
851: \begin{eqnarray}
852: \textbf{b}_\textbf{x}(t,z)=c_4e^{-\iota (\omega t-kz)},\nonumber\\
853: \textbf{b}_\textbf{z}(t,z)=c_5e^{-\iota (\omega t-kz)}.
854: \end{eqnarray}
855:
856: \subsection{Perturbation Equations}
857:
858: Using Eq.(2.21) in Eqs.(2.17)-(2.20), we have
859: \begin{eqnarray}
860: &&\frac{\partial (\delta \textbf{B})}{\partial t}=\nabla
861: \times(\alpha \textbf{v} \times \textbf{B})
862: +\nabla \times (\alpha \textbf{V} \times \delta \textbf{B}),\\
863: &&\nabla. (\delta \textbf{B})=0,\\
864: &&\left(\frac{1}{\alpha}\frac{\partial}{\partial
865: t}+\textbf{V}.\nabla \right)\delta \rho + \rho \gamma^2
866: \textbf{V}.\left(\frac{1}{\alpha} \frac{\partial}{\partial t}+
867: \textbf{V}.\nabla\right)\textbf{v}
868: -\frac{\delta \rho}{\rho}(\textbf{V}.\nabla)\rho +\rho (\nabla.\textbf{v})\nonumber\\
869: &&=-2\rho \gamma^2 (\textbf{V}.\textbf{v})( \textbf{V}. \nabla) ln
870: \gamma -\rho \gamma^2 (\textbf{V}.\nabla \textbf{V}).\textbf{v}
871: +\rho(\textbf{v}.\nabla ln u),\\
872: &&\left\{\left(\rho\gamma^2+\frac{\textbf{B}^2}{4\pi}\right)\delta_{ij}+\rho\gamma^4
873: V_i V_j-\frac{1}{4 \pi}B_i B_j\right\}\frac{1}{\alpha}
874: \frac{\partial v^j}{\partial t}+\frac{1}{4 \pi}\left[\textbf{B}
875: \times \left\{\textbf{V} \times
876: \frac{1}{\alpha}\frac{\partial(\delta \textbf{B})}
877: {\partial t}\right\}\right]_i\nonumber\\
878: &&+\rho \gamma^2v_{i,j}V^j +\rho \gamma^4 V_i v_{j,k} V^j V^k
879: -\frac{1}{4 \pi \alpha}\{(\alpha \delta B_i)_{,j}-(\alpha \delta
880: B_j)_{,i}\} B^j=-\gamma^2\{\delta \rho+2 \rho \gamma^2
881: (\textbf{V}.\textbf{v})\}a_i\nonumber\\
882: &&+\frac{1}{4 \pi \alpha}\{(\alpha B_i)_{,j}-(\alpha
883: B_j)_{,i}\}\delta B^j-\rho\gamma^4 (v_i V^j + v^j V_i)V_{k,j}V^k-
884: \gamma^2\{\delta
885: \rho V^j +2 \rho \gamma^2 (\textbf{V}.\textbf{v}) V^j +\rho v^j\} V_{i,j}\nonumber\\
886: &&-\gamma^4 V_i\{\delta \rho V^j+4 \rho
887: \gamma^2(\textbf{V}.\textbf{v})V^j+\rho v^j\}V_{j,k}V^k.
888: \end{eqnarray}
889: The component form of the Eqs.(5.5)-(5.8) can be written, after a
890: tedious algebra, as follows
891: \begin{eqnarray}
892: &&\frac{1}{\alpha}\frac{\partial b_x}{\partial t}+ub_{x,z}=\nabla
893: ln \alpha(v_x-\lambda v_z +V b_z-u
894: b_x)+(v_{x,z}-\lambda v_{z,z}-\lambda' v_{z}+V'b_z-u'b_x),\\
895: &&\frac{1}{\alpha}\frac{\partial b_z}{\partial
896: t}+ub_{z,z}=0,\\
897: &&b_{z,z}=0,\\
898: &&\frac{1}{\alpha}\frac{{\partial \tilde{\rho}}}{\partial t}+u
899: \tilde{\rho}_{,z} +\gamma^2 V\frac{1}{\alpha}\frac{\partial
900: v_x}{\partial t}+\gamma^2 u\frac{1}{\alpha} \frac{\partial
901: v_z}{\partial t}+(1+\gamma^2 u^2)v_{z,z}+\gamma^2uVv_{x,z}\nonumber\\
902: &&=-\gamma^2u[(1+2\gamma^2 V^2)V'+2\gamma^2uVu']v_x
903: +[(1-2\gamma^2u^2)(1+\gamma^2u^2)\frac{u'}{u}-2\gamma^4u^2VV']v_z,\\
904: &&\left\{\rho
905: \gamma^2(1+\gamma^2V^2)+\frac{B^2}{4\pi}\right\}\frac{1}{\alpha}
906: \frac{\partial v_x}{\partial t}+\left(\rho \gamma^4 u
907: V-\frac{\lambda B^2}{4\pi}\right)\frac{1}{\alpha}\frac{\partial
908: v_z}{\partial t}
909: +\left\{\rho \gamma^2(1+\gamma^2 V^2)-\frac{B^2}{4\pi}\right\}u v_{x,z}\nonumber\\
910: &&+\left(\rho \gamma^4uV+\frac{\lambda
911: B^2}{4\pi}\right)uv_{z,z}-\frac{B^2}{4\pi}(1-u^2)b_{x,z}
912: -\frac{B^2}{4\pi\alpha}b_x \{\alpha'(1-u^2)-\alpha uu'\}\nonumber\\
913: &&+\tilde{\rho}\rho\gamma^2 u\{(1+\gamma^2
914: V^2)V'+\gamma^2uVu'\}+\left[\rho\gamma^4u\{(1+4\gamma^2 V^2)u
915: u'+4(1+\gamma^2
916: V^2)VV'\}+\frac{B^2 u \alpha'}{4\pi \alpha}\right]v_x\nonumber\\
917: &&+\left[\rho\gamma^2 [\{(1+2 \gamma^2 u^2)(1+2 \gamma^2 V^2) -
918: \gamma^2 V^2\}V'+2\gamma^2(1+2\gamma^2 u^2)u V u'] +\frac{B^2 u}{4
919: \pi\alpha}(\lambda\alpha)'\right]v_z=0,\\
920: &&\left\{\rho \gamma^2(1+\gamma^2 u^2)+ \frac{\lambda^2
921: B^2}{4\pi}\right\}\frac{1}{\alpha}\frac{\partial v_z}{\partial
922: t}+\left(\rho \gamma^4 u V-\frac{\lambda
923: B^2}{4\pi}\right)\frac{1}{\alpha}\frac{\partial v_x}{\partial t}
924: +\left\{\rho \gamma^2(1+\gamma^2 u^2)
925: -\frac{\lambda^2 B^2}{4\pi}\right\}u v_{z,z}\nonumber\\
926: &&+\left(\rho \gamma^4 u V+\frac{\lambda
927: B^2}{4\pi})uv_{x,z}\right)+\frac{\lambda B^2}{4
928: \pi}(1-u^2)b_{x,z}+\frac{B^2}{4\pi \alpha}\{(\alpha
929: \lambda)'+\alpha'\lambda-u\lambda(u\alpha'+u'\alpha)\}b_x\nonumber\\
930: &&+\tilde{\rho} \gamma^2[a_z+u\{(1+\gamma^2 u^2)u'+\gamma^2uVV'\}]\nonumber\\
931: &&+\left[\rho\gamma^4\{u^2V'(1+4\gamma^2V^2)+2V(a_z+uu'(1+2\gamma^2
932: u^2))\}+\frac{\lambda B^2 \alpha' u}{4\pi\alpha}\right]v_x\nonumber\\
933: &&+\left[\rho\gamma^2\{u'(1+\gamma^2 u^2)(1+4\gamma^2 u^2)
934: +2u\gamma^2\{(1+2\gamma^2 u^2)VV'+a_z\}\}-\frac{\lambda B^2
935: u}{4\pi\alpha}(\alpha\lambda)'\right]v_z=0.
936: \end{eqnarray}
937: When the above equations are Fourier analyzed, they take the
938: following form
939: \begin{eqnarray}
940: &&c_3(\alpha'+\iota k \alpha)-c_2\{(\alpha \lambda)' +\iota
941: k\alpha\lambda\}+c_5(\alpha V)'
942: -c_4\{(\alpha u)'-\iota \omega +\iota k \alpha u \}=0,\\
943: &&c_5\left(-\frac{\iota \omega}{\alpha}+\iota k u\right)=0,\\
944: &&c_5 \iota k=0,\\
945: &&c_1\left(-\frac{\iota \omega}{\alpha}+\iota k
946: u\right)+c_2\left\{-\frac{\iota \omega}{\alpha}\gamma^2 u+\iota
947: k(1+\gamma^2u^2)-(1-2\gamma^2u^2)(1+\gamma^2u^2)\frac{u'}{u}+2\gamma^4
948: u^2VV'\right\}\nonumber\\
949: &&+c_3\gamma^2\left[\left(-\frac{\iota\omega}{\alpha}+\iota
950: ku\right)V+u\{(1+2\gamma^2V^2)V'+2\gamma^2uVu'\}\right]=0,\\
951: &&c_1\rho\gamma^2u\{(1+\gamma^2V^2)V'+\gamma^2uVu'\}-\frac{B^2}{4\pi}c_4\{(1-u^2)\iota
952: k+\frac{\alpha'}{\alpha}(1-u^2)-uu'\}\nonumber\\
953: &&+c_2\left[-\left(\rho\gamma^4 u V- \frac{\lambda
954: B^2}{4\pi}\right)\frac{\iota \omega}{\alpha}+\iota k
955: u\left(\rho\gamma^4 u V+ \frac{\lambda
956: B^2}{4\pi}\right)+\rho\gamma^2\{(1+2\gamma^2 u^2)(1+2\gamma^2V^2)
957: -\gamma^2V^2\}V'\right.\nonumber\\
958: &&\left.+2\rho\gamma^4(1+2\gamma^2 u^2)u V
959: u'+\frac{B^2u}{4\pi\alpha}(\lambda \alpha)'\right]
960: +c_3\left[-\left\{\rho\gamma^2(1+\gamma^2 V^2)+
961: \frac{B^2}{4\pi}\right\}\frac{\iota \omega}{\alpha}\right.\nonumber\\
962: &&\left.+\iota ku\left\{\rho\gamma^2(1+\gamma^2
963: V^2)-\frac{B^2}{4\pi}\right\}+\rho\gamma^4u\{(1+4\gamma^2V^2)uu'+4(1+\gamma^2V^2)VV'\}
964: -\frac{B^2u\alpha'}{4\pi\alpha}\right]=0,\\
965: &&c_1\rho\gamma^2[a_z +u\{(1+\gamma^2 u^2)u'+\gamma^2VuV'\}]\nonumber\\
966: &&+c_2\left[-\left\{\rho \gamma^2(1+\gamma^2 u^2)+\frac{\lambda^2
967: B^2}{4\pi}\right\}\frac{\iota \omega}{\alpha}+\left\{\rho
968: \gamma^2(1+\gamma^2u^2)-\frac{\lambda^2 B^2}{4\pi}\right\}\iota
969: ku\right.\nonumber\\
970: &&\left.+\left\{\rho\gamma^2\{u'(1+\gamma^2 u^2)(1+4\gamma^2 u^2)
971: +2u\gamma^2 ((1+2\gamma^2 u^2)VV'+a_z)\}-\frac{\lambda B^2
972: u}{4\pi\alpha}(\alpha \lambda)'\right\}\right]\nonumber\\
973: &&+c_3\left[-\left(\rho \gamma^4 u V- \frac{\lambda
974: B^2}{4\pi}\right)\frac{\iota \omega}{\alpha}+\left(\rho \gamma^4 u
975: V+\frac{\lambda B^2}{4\pi}\right)\iota ku\right.\nonumber\\
976: &&\left.+\left\{\rho\gamma^4 \{u^2V'(1+4\gamma^2
977: V^2)+2V(a_z+uu'(1+2\gamma^2 u^2))\}+\frac{\lambda B^2 \alpha'
978: u}{4\pi\alpha}\right\}\right]\nonumber\\
979: &&+\frac{B^2}{4 \pi}c_4\left[\lambda (1-u^2)\iota
980: k+\lambda\frac{\alpha'}{\alpha}(1-u^2)-\lambda uu'+\frac{(\lambda
981: \alpha)'}{\alpha}\right]=0.
982: \end{eqnarray}
983: Equation (5.17) implies that $c_5$ is zero which gives that $b_z$
984: is zero.
985:
986: \subsection{Numerical Solutions}
987:
988: We consider the same circumstances as we have considered in the
989: rotating non-magnetized plasma and assume that
990: $u=\frac{1}{\sqrt{z^2+2}}=V$. Substituting this value in Eq.(5.2)
991: with $V_F =1$, it follows that $\lambda=1-\frac{\sqrt{z^2+2}}{z}$
992: which shows that the magnetic field diverges near the horizon (as
993: discussed in [25]). For the sake of convenience, we take
994: $B=\sqrt{\frac{176}{7}}$ so that $\frac{B^2}{4\pi}=2$. Making use
995: of $c_5=0$ (Eq.(5.17)), we get a $4\times4$ matrix of the
996: coefficients of constants. The complex dispersion relation has two
997: parts. The real part gives a dispersion relation of the type
998: $A(z)k^4+B(z)k^3\omega+C(z)\omega^2+D(z)\omega^4+E(z)k^2
999: +F(z)k^2\omega^2+G(z)k\omega+H(z)k\omega^2+L(z)=0$ whereas the
1000: imaginary part gives a relation of the form
1001: $A(z)k^3+B(z)k^2\omega+C(z)\omega+D(z)\omega^3+E(z)k\omega^2
1002: +F(z)k=0.$ The first equation gives four dispersion relations out
1003: of which two are real and interesting while the other two roots
1004: are complex conjugates of each other. The second equation gives
1005: three dispersion relations. The respective wave numbers are shown
1006: in the Figures 9-11. These relations give different types of modes
1007: when $\textbf{B}>0$ and wave number is in arbitrary direction to
1008: $\textbf{B}.$ The two dispersion relations obtained from the real
1009: part are shown in the Figures 7 and 8.
1010: \begin{figure}
1011: \center \epsfig{file=p7a.eps,width=0.4\linewidth} \center
1012: \epsfig{file=p7b.eps,width=0.8\linewidth} \center
1013: \epsfig{file=p7c.eps,width=0.8\linewidth} \caption{The waves grow
1014: as they move away from the event horizon. The medium admits the
1015: properties of metamaterials.}
1016: \end{figure}
1017:
1018: The Figure 7 shows that at the horizon the wave number is infinite
1019: and hence no wave is present there. $k$ has real values in the
1020: region $1\leq z\leq10,~10^{-7}\leq\omega\leq10$. No wave is
1021: present in the region $0\leq z<1$, evanescent waves are present in
1022: $0\leq\omega<10^{-7}.$ All the quantities, the wave number, wave
1023: velocity, group velocity and refractive index are negative. This
1024: leads to the existence of metamaterial in the region [32]. The
1025: wave number is decreasing with an increase in angular frequency
1026: and increasing with an increase in time lapse.
1027: \begin{figure}
1028: \center \epsfig{file=p8a.eps,width=0.4\linewidth} \center
1029: \epsfig{file=p8b.eps,width=0.8\linewidth} \center
1030: \epsfig{file=p8c.eps,width=0.8\linewidth} \caption{The waves damp as
1031: they move away from the event horizon. The dispersion is not normal
1032: in the region.}
1033: \end{figure}
1034:
1035: The Figure 8 shows that the wave number increases with the
1036: increase in angular velocity and decreases with the increase of
1037: lapse function. The horizon possesses no wave due to infinite wave
1038: number. The region $0<z\leq1$ has no wave and $0\leq\omega<5\times
1039: 10^{-6}$ is region of evanescent waves. The value of refractive
1040: index is less than one and hence it is not a normal dispersion.
1041:
1042: The dispersion relation obtained from the imaginary part of the
1043: matrix determinant are shown in the Figures 9, 10 and 11.
1044: \begin{figure}
1045: \center \epsfig{file=p9a.eps,width=0.4\linewidth} \center
1046: \epsfig{file=p9b.eps,width=0.8\linewidth} \center
1047: \epsfig{file=p9c.eps,width=0.8\linewidth} \center \caption{The
1048: waves damp and then grow as they move away from the event horizon.
1049: The medium admits the properties of metamaterials.}
1050: \end{figure}
1051:
1052: The Figure 9 shows that the wave number is infinite in $0\leq
1053: z<0.5.$ The wave number is positive in the region $0.5\leq z<1.6$,
1054: it decreases abruptly and then it smoothly rises but is negative.
1055: The phase and group velocities are negative. The refractive index
1056: is negative in the domain $1.6\leq z<10$ so the region possesses
1057: the properties of metamaterials.
1058: \begin{figure}
1059: \center \epsfig{file=p10a.eps,width=0.4\linewidth} \center
1060: \epsfig{file=p10b.eps,width=0.8\linewidth} \center
1061: \epsfig{file=p10c.eps,width=0.8\linewidth} \center \caption{The
1062: waves suddenly grow and then damp on going away from the event
1063: horizon. The dispersion is not normal.}
1064: \end{figure}
1065:
1066: In the Figure 10, real waves exist in the region $0.5\leq z\leq
1067: 10.$ No wave is present on the event horizon as well as in the
1068: region $0<z\leq0.5$ because the wave number goes to infinity
1069: there. The waves are increasing in $0.5\leq z<0.7$ then there is a
1070: sudden decrease. The region $0.5\leq z\leq 0.7,~0\leq\omega<4$
1071: contains evanescent waves. The refractive index is less than one
1072: in the whole domain, hence the waves are not normally disperse.
1073:
1074: \begin{figure} \center
1075: \epsfig{file=p11a.eps,width=0.4\linewidth} \center
1076: \epsfig{file=p11b.eps,width=0.8\linewidth} \center
1077: \epsfig{file=p11c.eps,width=0.8\linewidth} \caption{The waves damp
1078: when they move away from the event horizon. The phase and the group
1079: velocities are equal. The dispersion is not normal.}
1080: \end{figure}
1081:
1082: The Figure 11 shows that real waves exist in the region $0.5\leq
1083: z\leq 10,~0.5\leq\omega\leq10.$ The phase and group velocities are
1084: same in this domain. The refractive index is less than one and the
1085: dispersion is not normal. No wave is present on the event horizon
1086: and the region near the horizon because the wave number is
1087: infinite there. As we go away from the horizon, the wave number
1088: decreases. On increasing the angular frequency, we find growing
1089: waves.
1090:
1091: \section{Conclusion\label{Conclusion}}
1092:
1093: We have reformulated the GRMHD equations for the Schwarzschild
1094: black hole magnetosphere to take account of gravitational effects
1095: due to event horizon. We discuss one dimensional perturbations in
1096: perfect MHD condition. This has been explored for the cold plasma
1097: case. These equations are written in component form to make them
1098: easier for the Fourier analysis. The equations are Fourier
1099: analyzed by using the assumption of plane wave. The determinant of
1100: the coefficients is solved numerically for the following
1101: backgrounds in the magnetosphere.
1102: \par\noindent
1103: \begin{enumerate}
1104: \item Non-Rotating Background (either non-magnetized or magnetized)
1105: \item Rotating Non-Magnetized Background
1106: \item Rotating Magnetized Background.
1107: \end{enumerate}
1108: \par\noindent
1109: The non-rotating background shows pure Schwarzschild geometry
1110: outside the event horizon whereas the rotating background
1111: demonstrates the restricted Kerr geometry in the vicinity of the
1112: event horizon which admits a variable lapse function with negligible
1113: rotation. We have taken Schwarzschild black hole with rotating
1114: background due to the well-known fact that the Schwarzschild black
1115: hole is the minimum configuration of the Kerr black hole.
1116:
1117: In the case of non-rotating background, we have found that the
1118: magnetospheric fluid is normally disperse. The wave number becomes
1119: indefinite at the horizon and for the values of $z$ near the
1120: horizon and hence no wave is present there. This shows that no
1121: signal can pass the event horizon or near to it. The third subcase
1122: shows a mode which is not normally dispersive.
1123:
1124: In the case of rotating non-magnetized background we have found
1125: one case (Figure 4) in which the dispersion is normal. The regions
1126: with zero angular frequency or very less angular frequency are
1127: evanescent in all the three cases. The wave number becomes
1128: infinite at event horizon and in the neighborhood of the event
1129: horizon and no wave exists there.
1130:
1131: In rotating magnetized background, we obtain region with negative
1132: wave number, phase velocity, group velocity and refractive index
1133: (Figures 7 and 9). This fact is deduced by the analysis of two of
1134: the dispersion relations obtained in section 5. The refractive
1135: index is negative in both the cases which implies that the region
1136: possesses all the qualities of left-handed metamaterials. The
1137: other three dispersion relations give that the region under
1138: discussion is of non-normal dispersion which restrict the real
1139: signals pass through that region.
1140:
1141: It is worth mentioning that the wave number becomes infinite at
1142: the event horizon and no wave exists there. This supports the
1143: well-known point of view that no information can be extracted from
1144: a black hole. It is interesting to mention that this work extends
1145: the results given by Mackay et al. in [33] according to which
1146: rotation of a black hole is required for negative phase velocity
1147: propagation. It is also observed that the waves of less angular
1148: velocity are evanescent. Our numerical results indicate that
1149: negative phase velocity propagates in the rotating background
1150: whether the black hole is rotating or non-rotating.
1151:
1152: We know that the MHD waves in cold plasma are non-dispersive.
1153: However, the dispersion is noted in the above figures. This factor
1154: comes due to the formalism used and the equations which provides
1155: different equations from the usual MHD equations. Since the 3+1
1156: split of GR is used in a preliminary investigation of waves
1157: propagating in a plasma influenced by the gravitational field. Thus
1158: internal gravity waves which interrupt the MHD waves imply the cases
1159: of dispersion in each of the hypersurface. We would like to mention
1160: that graphs of the waves are given in a particular hypersurface
1161: where time is constant, not in the whole Schwarzschild background
1162: and consequently this is justified locally, not globally. Finally,
1163: it is mentioned that some of the Figures have patches missing which
1164: is due to the existence of complex numbers there. Mathematica cannot
1165: plot complex numbers with real numbers. These complex numbers are
1166: shown by gaps in the Figures.
1167:
1168: This analysis has been done for the cold plasma state. It would be
1169: interesting to extend this analysis for the isothermal state of
1170: plasma. This would provide the effects of pressure in the current
1171: work. Currently, it is in progress.
1172:
1173: \newpage
1174: {\bf Acknowledgment}
1175:
1176: \vspace{0.5cm}
1177:
1178: We acknowledge the enabling role of the Higher Education
1179: Commission Islamabad, Pakistan, and appreciate its financial
1180: support through the {\it Indigenous PhD 5000 Fellowship Program
1181: Batch-II}.
1182:
1183: \vspace{0.5cm}
1184:
1185: \begin{thebibliography}{99}
1186:
1187:
1188: \bibitem{Petterson}Petterson, J.A.: Phys. Rev.
1189: \textbf{D10}(1974)3166.
1190:
1191: \bibitem{Israel} Israel, W.: Phys. Rev. \textbf{164}(1967)1776; Commun. Math. Phys. \textbf{8}(1968)245.
1192:
1193: \bibitem{Hawking} Hawking, S.: Commun. Math. Phys. \textbf{25}(1972)152.
1194:
1195: \bibitem{Robinson} Robinson, D.: Phys. Rev. \textbf{D10}(1974)458.
1196:
1197: \bibitem{ADM}Arnowitt, R., Deser, S. and Misner, C.W.: \textit{Gravitation: An
1198: Introduction to Current Research} (Wiley, New York, 1962).
1199:
1200: \bibitem{Tarfulea}Tarfulea, N.: Ph.D. Thesis.
1201: (University of Minnesota, 2004).
1202:
1203: \bibitem{TM}Thorne, K.S. and Macdonald, D.A.: Mon. Not. R. astr. Soc.
1204: \textbf{198}(1982)339.
1205:
1206: \bibitem{TM}Thorne, K.S. and Macdonald, D.A.: Mon. Not. R. astr.
1207: Soc.\textbf{198}(1982)345.
1208:
1209: \bibitem{TPM}Thorne, K.S., Price, R.H. and Macdonald, D.A.: \textit{Black Holes:
1210: The Membrane Paradigm} (Yale University Press, New Haven, 1986).
1211:
1212: \bibitem{Holcomb et al.}Holcomb, K.A. and Tajima, T.: Phys. Rev. \textbf{D40}(1989)3809.
1213:
1214: \bibitem{Holcomb et al.}Holcomb, K.A.: Astrophys. J. \textbf{362}(1990)381.
1215:
1216: \bibitem{Dettman et al.}Dettman, C.P., Frankel, N.E. and Kowalenko, V.: Phys. Rev.
1217: \textbf{D48}(1993)5655.
1218:
1219: \bibitem{Khanna}Khanna, R.: Mon. Not. R. Astr. Soc. \textbf{294}(1998)673.
1220:
1221: \bibitem{A et al.} Ant\'{o}n, L., Zanotti, O., Miralles, J.A.,
1222: Mart\'{i}, J.A., $M^{\textbf{A}}$ Ib\'{a}$\tilde{n}$ez, J., Font,
1223: J.A. and Pons, J.A.: Astrophys. J. \textbf{637}(2006)296.
1224:
1225: \bibitem{Anile}Anile, A.M.: \textit{Relativistic Fluids and
1226: Magneto-Fluids with Applications in Astrophysics and Plasma
1227: Physics} (Cambridge University Press, 1989).
1228:
1229: \bibitem{S.S} Komissarov, S.S.: Mon. Not. R. Astron. Soc.
1230: \textbf{336}(2002)759.
1231:
1232: \bibitem{Komissarov} Komissarov, S.S.: Mon. Not. R. Astron. Soc.
1233: \textbf{350}(2004)407.
1234:
1235: \bibitem{Zhang}Zhang, X.-H.: Phys. Rev. \textbf{D39}(1989)2933.
1236:
1237: \bibitem{Zhang}Zhang, X.-H.: Phys. Rev. \textbf{D40}(1989)3858.
1238:
1239: \bibitem{Tsikarishvili} Tsikarishvili, E.G., Rogava, A.D., Lominadze, J.G.
1240: and Javakhishvili, J.I.: Phys. Rev. \textbf{A46}(1992)1078.
1241:
1242: \bibitem{Mofiz} Mofiz, U.A.: BRAC University Journal, \textbf{II}(2005)83.
1243:
1244: \bibitem{Regge} Regge, T. and Wheeler, J.A.: Phys. Rev.
1245: \textbf{108}(1957)1063.
1246:
1247: \bibitem{Zerilli}Zerilli, F.: Phys. Rev. \textbf{D2}(1970)2141; J. Math. Phys.
1248: \textbf{11}(1970)2203; Phys. Rev. Lett. \textbf{24}(1970)737.
1249:
1250: \bibitem{Price} Price, R.H.: Phys. Rev. \textbf{D5}(1972)2419;
1251: \textit{ibid} 2439.
1252:
1253: \bibitem{Hanni}Hanni, R.S. and Ruffini R.: Phys. Rev.
1254: \textbf{D8}(1973)3259.
1255:
1256: \bibitem{Wald}Wald, R.M.: Phys. Rev. \textbf{D10}(1974)1680.
1257:
1258: \bibitem{SK}Sakai, J. and Kawata, T.: J. Phys. Soc. Jpn. \textbf{49}(1980)747.
1259:
1260: \bibitem{Buzzi et al.}Buzzi, V., Hines, K.C. and Treumann, R.A.: Phys. Rev.
1261: \textbf{D51}(1995)6663; \textit{ibid} 6677.
1262:
1263: \bibitem{Das}Das, A.C.: \textit{Space Plasma Physics: An Introduction}
1264: (Narosa Publishing House, New Delhi, 2004).
1265:
1266: \bibitem{Jackson} Jackson, J.D.: \textit{Classical Electrodynamics}
1267: (John Wiley \& Sons, 1999).
1268:
1269: \bibitem{Achenbach} Achenbach, J.D.: \textit{Wave Propogation in Elastic Solids}
1270: (North-Holland Publishing Company, Oxford, 1973).
1271:
1272: \bibitem{Veselago}Veselago, V.G.: Sov. Phys. Usp. \textbf{10}(1968)509.
1273:
1274: \bibitem{Mackay}Mackay, T.G., Lakhtakia, A. and Setiawan, S.: New J. Phys.
1275: \textbf{7}(2005)171.
1276:
1277: \end{thebibliography}
1278:
1279: \end{document}
1280: