astro-ph0606177/MS.TEX
1: 
2: %\documentclass[12pt,preprint]{aastex}
3: %\documentstyle[12pt,aasms4]{article}
4: %\documentstyle[emulapj,draft]{article}
5: %\documentstyle[emulapj]{article}
6: %\documentstyle[emulapj,epsfig]{article}
7: 
8: %\documentclass{aastex}
9: %\usepackage{emulateapj5,epsfig}
10: %\input epsf.def
11: %\input{standard-defs.tex}
12: 
13: \font\tenbg=cmmib10 at 10pt
14: \def \rvecmu{{\hbox{\tenbg\char'026}}}
15: \def \rvecphi{{\hbox{\tenbg\char'036}}}
16: \def \Omegabold {{\hbox {\tenbg\char'012}}}
17: \def \rvecOmega {{\hbox {\tenbg\char'012}}}
18: 
19: 
20: %\usepackage{emulateapj5}
21: %\usepackage{emulapj}
22: \documentstyle[emulapj,epsfig]{article}
23: \begin{document}
24: 
25: \bigskip
26: 
27: %\magnification=\magstep1
28: %\baselineskip=12pt \vbadness=10000
29: %\font\one=cmbx10 scaled\magstep2
30: \font\two=cmbx10 scaled \magstep1
31: 
32: 
33: %\def\idem{\smallskip\noindent
34: %\hangindent 2 pc }
35: 
36: 
37: 
38: \title{\bf Magnetospheric Gap and Accumulation of Giant Planets
39:  Close  to the Star}
40: 
41: \author{M.M.~Romanova}
42: \affil{Department of Astronomy, Cornell University, Ithaca, NY
43: 14853-6801; ~ romanova@astro.cornell.edu}
44: 
45: \author{R.V.E.~Lovelace}
46: \affil{Departments of Astronomy and Applied Physics, Cornell
47: University, Ithaca, NY 14853-6801; ~RVL1@cornell.edu }
48: 
49: 
50: \keywords{accretion, dipole
51: --- plasmas --- magnetic
52: fields --- stars: magnetic fields --- X-rays: stars}
53: 
54: 
55: 
56: \begin{abstract}
57: 
58: 
59: The bunching of giant planets at a  distance of several stellar
60: radii may be explained by the disruption of the inner part of the
61: disk by the magnetosphere of the star during the T Tauri stage of
62: evolution. The rotating magnetic
63: field of the star gives rise to a low
64: density magnetospheric gap where
65: stellar migration is strongly suppressed.
66:    We performed full 3D magnetohydrodynamic simulations of the
67: disk-magnetosphere interaction and examined conditions for which
68: the magnetospheric gap is ``empty", by changing the misalignment
69: angle between magnetic and rotational axes of the star, $\Theta$,
70: and by lowering the adiabatic index $\gamma$, which mocks up the
71: effect of heat conductivity and cooling.
72:     Our simulations  show that for a wide range
73: of plausible conditions the gap is essentially empty.
74:   However, in  the case
75: of  large misalignment angles $\Theta$,
76: part of the funnel stream is located
77: in the equatorial plane and the
78: gap is not empty.
79:    Furthermore, if the adiabatic index is small
80: ($\gamma \sim 1.1$) and the rotational and magnetic axes are almost
81: aligned, then matter penetrates through the magnetosphere
82: due to 3D instabilities
83: forming high-density equatorial funnels.
84:   For these two limits there is appreciable matter
85: density  in the equatorial plane of the disk so that a planet may
86: migrate into the star.
87: 
88: 
89: 
90: \end{abstract}
91: 
92: \section{Introduction}
93: 
94:    More than  $170$  giant planets have been
95: discovered around solar-type stars (see catalog of planets in
96: http://vo.obspm.fr/exoplanetes/encyclo/catalog-main.php).
97:   About $24\%$ of them are located very close to the star,
98: at $r \lesssim  0.1 {\rm AU}$ (see also reviews by Marcy et al.
99: 2003; Papaloizou \& Terquem 2006). There is a prominent peak in the
100: radial distribution of the planets at $r\sim 0.04 - 0.05~ {\rm AU}
101: \approx (7-10) R_\odot$ which corresponds to periods of 3 days (see
102: Figure 1).
103: 
104:   The radial distribution and other properties
105: of the giant planets have stimulated work  on  models of planet
106: formation and migration.
107:     According to the presently favored  interpretation,
108: planets form far away from the star either through  core
109: accretion (Mizuno 1980; Pollack  et al. 1996), or through
110: instabilities in the disk (Boss 2001).
111:    Subsequently they migrate inward
112: due to their gravitational interaction
113: with the disk (Lin \& Paploizou 1986;
114: Lin, Bodenheimer \& Richardson
115: 1996; Ward 1997; Nelson \& Papaloizou 2003).
116: 
117: 
118: 
119: 
120:   Some planets may migrate close to the star
121: where the disk properties are strongly
122: influenced by the star and/or
123: the star's magnetic field.
124:    In particular, the inner regions of the disk
125: may be truncated as a result of
126: heating by the star (e.g., Kuchner \&
127: Lecar 2002) and
128: photoevaporation of the region of the disk
129: $\lesssim 1$ AU
130: (Matsuyama, Johnstone \& Murray 2003).
131:   Or, the inner part of the disk may be disrupted by  the strong
132: magnetic field of the protostar (Lin et al.
133: 1996).
134: 
135: 
136: 
137:       There are strong
138: observational arguments that young solar-type stars (T Tauri type
139: stars) have  strong magnetic fields.
140:   They are thousands of times
141: brighter than the Sun in the X-ray
142: (see review by Feigelson \&
143: Montmerle 1999) which is a sign
144: of their high magnetic activity.
145:     In a number of cases, direct
146: measurements of the Zeeman broadening imply a
147: $B$ field  (averaged over the surface of the star)
148: of the order of several kilo-Gauss, which is much stronger
149: than that the average field on the
150: surface of the  Sun (e.g., Basri et
151: al. 1992; Johns-Krull et al. 1999).
152:   The strongest magnetic fields are
153: probably associated with the
154: multipolar component (e.g. Safier 1998;
155: Johns-Krull et al. 1999; Smirnov et al. 2003).
156:    However, a significant
157: dipole component is also expected.
158:   It gives many observational
159: signs of magnetospheric accretion in the T Tauri stage (see review
160: by Bouvier et al. 2006) and also is observed in some circular
161: polarization measurements  (e.g., Valenti \& Johns-Krull 2004;
162: Symington et al. 2005).
163: 
164: 
165: A sufficiently strong dipole magnetic
166: field will truncate the disk at
167: a distance of several stellar radii forming a low-density
168: magnetospheric gap as shown in Figure 2.
169:    The migration rate
170: of planets in this gap will be greatly reduced.
171: 
172: 
173:  This work  analyzes the conditions where a low-density gap exists
174: between the inner edge of the disk and the
175: star's surface.
176:    Our analysis is based on 3D MHD simulations.
177: In particular, we investigate properties of the gaps
178: for different misalignment angles
179: $\Theta$ between the rotation axis of the star $\Omegabold_*$ and its
180: magnetic moment $\rvecmu$.
181:   We investigate conditions where part of the funnel
182: stream is in the equatorial plane close to the star.
183:    This  may
184: give an appreciable matter density in the magnetospheric
185: gap.
186:     Also we consider the possibility of
187: direct equatorial accretion from
188: the disk to the star due to 3D
189: instabilities.
190: 
191: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
192: \begin{figure*}[t]
193: \epsscale{1.1} \plotone{f1.eps}
194:  \caption{Distribution of extrasolar planets in the  vicinity of
195: the star.} \label{Figure 1}
196: \end{figure*}
197: 
198: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
199: 
200: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
201: \begin{figure*}[t]
202: \epsscale{1.3} \plotone{f2.eps}
203:  \caption{Sketch of an accretion  disk which is
204: disrupted by the star's dipole magnetic field.
205:   The rate of migration of a planet is greatly
206: slowed  once it enters the gap.}
207: \label{Figure 2}
208: \end{figure*}
209: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
210: 
211: 
212: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
213: \begin{figure*}[t]
214: \epsscale{0.9}
215: \plotone{f3.eps}
216: \caption{Radial distribution of the
217: density (solid line) and the angular velocity of the disk
218: $\omega_d=v_\phi/r$ (dashed line) in the vicinity of the star for a
219: misalignment angle of $\Theta=30^\circ$. Thin dashed line shows
220: Keplerian angular velocity $\omega_K$.}
221: \label{Figure 3}
222: \end{figure*}
223: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
224: 
225: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
226: \begin{figure*}[t]
227: \epsscale{1.6} \plotone{f4.eps} \caption{Result of 3D simulations of
228: disk accretion to a rotating star with a dipole moment $\rvecmu$
229: misaligned with the star's rotation axis $\Omegabold_*$ by
230: $\Theta=30^\circ$.
231:    The color background shows the density
232: distribution in the equatorial region.
233:  Density varies from $\rho\approx 0.003$
234: (blue color) to $\rho\approx 1$ (red
235: color).
236:   The red lines are magnetic field lines.
237: The black arrow shows the direction of the magnetic moment
238: $\rvecmu$.}
239: \label{Figure 4}
240: \end{figure*}
241: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
242: 
243: 
244: 
245: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
246: \begin{figure*}[t]
247: \epsscale{1.8} \plotone{f5.eps} \caption{The top panels show the
248: density distributions and sample magnetic field lines in the
249: $\Omegabold_*-\rvecmu$ plane for different misalignment angles
250: $\Theta = 30^\circ, 45^\circ, 60^\circ$.
251:    The bottom panels show the equatorial distribution of
252: density.}
253: \label{Figure 5}
254: \end{figure*}
255: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
256: 
257: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
258: \begin{figure*}[t]
259: \epsscale{1.7} \plotone{f6.eps} \caption{Result of simulations for
260: $\gamma=1.1$ for $\Theta = 5^\circ$ after $P=12$ rotations.
261:   The left panel
262: shows the density distribution
263: and sample field lines in the
264: $\Omegabold_* - \rvecmu$  plane.
265:   The middle panel shows the density distribution in the
266: equatorial plane, and the right
267: panel shows a 3D view  of the density levels and
268: sample field lines. }
269: \label{Figure 6}
270: \end{figure*}
271: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
272: 
273: 
274: \section{Where Does the Planet Migration Stop?}
275: 
276: 
277: The accretion disk is disrupted at the
278: Alfv\'en radius
279: where the dynamic pressure  of the
280: disk matter is comparable to the
281: magnetic pressure in the star's dipole field,
282: $r_A=[\mu^2/(\dot{M}\sqrt{GM})]^{2/7}$ or
283: \begin{equation}
284: r_A \approx 7.2\times 10^{11}~ {{B_3^{4/7} R_{2.5}^{12/7}}
285: \over{M_{0.8}^{1/7} \dot M_{-7}^{2/7}}}~{\rm cm}\approx 0.05~{\rm
286: AU}~,
287: \end{equation}
288: where  $M_{0.8}\equiv M_*/{\rm 0.8 M_\odot}$, $R_{2.5}\equiv
289: R_*/{2.5~R_\odot}$, and $B_3\equiv B_*/{10^3~{\rm G}}$ are the
290: normalized mass, radius, and magnetic field of the protostar, and
291: $\dot M_{-7}\equiv \dot M/{10^{-7}~{\rm M_\odot/{\rm yr}}} $ is the
292: accretion rate in the disk
293: (Ghosh \& Lamb 1979; Camenzind 1990;
294: K\"onigl 1991).
295:    For the typical parameters of T Tauri stars used in
296: this formula, this radius  coincides
297: approximately with the peak of
298: the distribution shown at Figure 1.
299:    Thus this peak may result from
300: the greatly reduced rate of migration inside
301: the magnetospheric gap.
302: 
303: Numerical simulations show that the disk is disrupted at the
304: distance $r\approx r_A$, where
305:   the plasma is
306: lifted out of the disk plane by the vertical
307:  pressure force and
308: it then flows along the star's dipole field lines in a funnel flow
309: (Romanova et al. 2002; 2003; 2004).
310:    As a consequence, the density of matter
311: in the equatorial plane is greatly reduced for $r < r_A$.
312:   Figure 3 shows the
313: equatorial density distribution obtained from our 3D
314: simulations.
315:   The density is large in the disk, and it often increases
316: as $r_A$ is approached.
317: However, for $r<r_A$ the density drops by
318: a factor of $\sim 100-300$
319: in the magnetically-dominated magnetosphere.
320:   A protoplanet which migrates inward to radii  $<r_A$
321: enters a region of greatly reduced density.
322: 
323: 
324:   For typical conditions  planets migrate inward as
325: a result of interaction of the planet with the disk matter. The
326: planet loses part of its orbital angular momentum by overtaking
327: collisions with the disk outside its orbit and it gains a smaller
328: part by overtaking collisions of the disk matter inside its orbit.
329: The rate of migration, or radial speed, $V_{pr}$, depends on a number
330: of parameters such as  mass of the planet, $M_p$, surface density of
331: the disk, $\Sigma$, viscosity in the disk, $\nu$, and others, and
332: also it is different in cases when a planet opens a gap in the disk
333: or not.
334: 
335:           If the planet's mass is
336: relatively small ($M_p\lesssim 10
337: M_\oplus$) it
338: does not open a gap in the accretion disk.
339:   The migration in this
340: case is referred to as ``type I,''    and the planet's inward drift
341: speed is $V_{pr} \propto -M_p (\Sigma r^2)$ (Ward 1997; Papaloizou
342: \& Terquem 2006).
343: 
344:            Planets of sufficiently large mass
345: open a gap in the disk of
346: width of the order of the disk thickness.
347:   The migration in
348: this case is referred to as ``type II,''  and it tends to ``lock''
349: the planet's migration to that of the disk matter if
350: the local disk mass, $M_d =4\pi r^2 \Sigma$ ,
351: is larger than the planets mass $M_p$.
352:   The disk matter moves inward with a
353: radial speed $v_r = - 3 \nu/(2r)$, where $\nu =
354:  \alpha c_s^2/ \omega_K$ is the usual Shakura-Sunyaev
355: turbulent viscosity with $\alpha = 10^{-3}-10^{-2}$.
356:     However, if the local disk mass is small compared
357: with the planet's mass, then the planets migration is
358: slower than that of the disk matter.
359:    The angular momentum lost by the planet
360: $d{J}_p/dt =M_p v_K V_{pr}/2$ in a second is equal to
361: the angular momentum transported outward
362: by the viscous stress in the disk, $d{J}_d/dt = \dot{M} r v_K$,
363: where $v_K=(GM/r)^{1/2}$ is the Keplerian velocity
364: and $\dot{M}$ mass accretion rate of the disk.
365:   This gives a migration speed of the planet
366: $V_{pr} = -(M_d/M_p)|v_r|$.
367:     We can write $\dot{M}=2\pi r \Sigma |v_r|$  so
368: that $M_d =4\pi r^2 \Sigma \approx 1.3 \times 10^{29}{\rm g}
369: (r/0.1{\rm AU})^{3/2}(\dot{M}/10^{-7}M_\odot/{\rm yr})$ for $h/r =0.1$ and
370: $\alpha = 10^{-3}$.
371: 
372: 
373:       For $M_d < M_p$,the time scale
374: for the planet's migration at
375: $r$ is simply
376: \begin{equation}
377: \tau_p = {r \over V_{pr} }= {M_p\over M_d}{2r^2 \over 3\nu}~,
378: \end{equation}
379: which is independent of $r$ and inversely
380: proportional to both $\dot{M}$
381: and $\alpha$.
382:   For a Jupiter mass planet, $\tau_p \approx 4700{\rm yr}
383: (10^{-7}M_\odot/{\rm yr}/\dot{M})$ for $\alpha =10^{-3}$ and $h/r=0.1$.
384:       However, equation(2) does not include the influence of
385: the star's magnetic field.
386:     Inside the magnetospheric gap the matter density is reduced
387: by a factor $\gtrsim 10^2$.
388:    The migration time will be increased by a corresponding
389: factor.
390: 
391: 
392: 
393:   Figure 3 also shows that the angular velocity of
394: the disk plasma in the equatorial plane $\omega_d$ within the
395: magnetospheric gap is much lower
396: than the Keplerian angular velocity of the
397: planet $\omega_K$ for stars with periods $P\gtrsim 2$ days.
398:   Thus, the disk
399: matter passing close by the planet will have a large relative
400: velocity $r(\omega_K-\omega_d)$.
401:   The relative velocity  is larger than in the non-magnetic
402: case by a factor $F\sim (r/h)(\omega_K-\omega_d)/\omega_K\gg1$ for
403: the type II migration.
404:     One can readily show that the rate of exchange of
405: angular momentum between the planet and the
406: disk is reduced by a factor $1/F^2 \ll 1 $.
407:     Thus the migration time inside the magnetospheric
408: gap will be increased further by a factor $F^2$.
409: We estimate $F^2 \gtrsim 10$.
410:    Once the planet is inside the magnetospheric gap
411: it may continue to migrate slowly inward owing to resonances
412: it has with the disk matter at larger radii but a
413: treatment of this is beyond the scope of the present
414: work.
415: 
416: 
417: Thus, in both cases the migration
418:  will be greatly reduced  if the
419: density in magnetospheric gap is
420: much lower than in the disk.
421: Conditions may be changed however
422: if the magnetic axis $\rvecmu$ is
423: misaligned relative to the rotational axis of the star
424: $\Omegabold_*$ by an angle $\Theta$.
425:    For high misalignment
426: angles $\Theta$, the funnel streams
427: may be partially located  in the
428: equatorial region (Romanova et al. 2003).
429:   From the other side, even in
430: the aligned case, some matter may accrete to the star in the
431: equatorial plane due to instabilities (e.g., Arons \& Lea 1976).
432:   We performed 3D simulations to investigate both factors.
433: 
434: 
435: 
436: \section{3D Simulations of the Disk-Magnetosphere Interaction
437: and Magnetospheric Gaps}
438: 
439: We performed a set of 3D simulations
440: using our code based on the
441: ``cubed sphere" grid  (Koldoba et al. 2002) with the main goal
442: of analyzing
443: the density distribution in the magnetospheric gap.
444:   Simulations were set up in
445: a way similar to those of Romanova et
446: al. (2003, 2004).
447:   Namely,  quasi-stationary initial conditions
448: were used which permitted slow viscous
449: accretion from the disk to a star.
450: An $\alpha-$viscosity  was incorporated to the code with
451: typical values of $\alpha-$parameter: $\alpha=0.02$ and $0.04$.
452: The magnetic axis $\rvecmu$ is
453: misaligned relative to rotational axis
454: of the star $\Omegabold_*$ by an angle $\Theta$.
455:    The rotational axis of the
456: star coincides with that of the disk.
457: 
458: Simulations were done for parameters
459: typical for T Tauri type stars:
460:  $M_*=0.8 M_\odot$, $R_*= 2.5 R_\odot$,
461: $B_*=10^3 G$, $\dot M\approx 3\times 10^{-8}
462: {M_\odot/{\rm yr}}$.
463:    Compared to our previous 3D runs,
464: we changed parameters so as to
465: increase the size of magnetospheric
466: gap to $r_A=(4-5) R_*$ versus
467: $r_A=(2-3) R_*$ in our previous papers.
468:   Figure 4 shows the magnetospheric gap in a
469: test case with an  even larger
470: magnetosphere $r_A=(6-7) R_*$.
471:    One can see that the low-density
472: magnetospheric gap
473: can be quite large.
474:    One can also see that inside the
475: magnetospheric radius (which
476: corresponds approximately to the edge of the disk),
477: the magnetic field
478: lines are closed, while outside
479: of this radius they are carried by
480: the matter of the disk or corona.
481:  Our simulations show that the density
482: inside the magnetospheric gap is about
483: $100-300$ times smaller than
484: the density in the nearby disk.
485: 
486: 
487: 
488: 
489: \noindent{\bf Magnetospheric Gap at Different $\Theta$.} We
490: performed 3D simulations for different misalignment angles $\Theta$
491: from $\Theta=0^\circ$ to $90^\circ$.
492:   We investigated the
493: magnetospheric gaps in the equatorial plane. Simulations have shown
494: that matter flow is different at small and large misalignment
495: angles.
496:   For  angles $\Theta\lesssim 45^\circ$, matter
497: flows to the star along  funnel streams which are
498: above and below the equatorial plane.
499:    Thus within the magnetospheric gap $r<r_A$ the
500: matter density in the equatorial plane is greatly
501: reduced (see Figure 5, left two panels).
502:     For larger angles, matter also
503: accretes to a star through the funnel streams.
504:     However,  part of the funnel
505: streams is  located in the equatorial plane
506: and the magnetospheric gap
507: is not empty (Figure 5, right panel).
508:   Thus at large $\Theta$, the
509: planets orbiting in the equatorial
510: plane will interact with the
511: dense gas of the streams and may
512: continue to migrate inward to the
513: star.
514: 
515: 
516: 
517: 
518: \noindent{\bf Accretion through Equatorial Funnels at Low $\gamma$}.
519:   There is another possible reason why
520: the magnetospheric gap may  have some matter density.
521:    There are possible instabilities which may
522: lead to the direct accretion of
523: matter through the magnetosphere
524: in the equatorial plane.
525:    To investigate such instabilities,
526: we took the almost aligned case,
527: $\Theta=5^\circ$, and decreased
528: the adiabatic index from $\gamma=5/3$ to $\gamma=1.1$.
529:   The adiabatic index may be significantly lower
530: than its  ideal value in the case of  high electron
531: heat conductivity which may occur in a highly
532: ionized plasma.
533:   In  our simulations, the low value of $\gamma$
534: acts to give a low temperature
535: in the disk and the funnel flow.
536:   We observed that  matter partially accretes in
537: the equatorial plane.
538:    Figure 6 shows that matter accreted
539: through funnels which are located inside the
540: magnetosphere.
541:    They penetrate inwards through
542: the Rayleigh-Taylor type instability
543:    (e.g. Arons \& Lea 1976)
544:    up to some distance $r_{1}$,
545: and then form regular funnel streams along
546: the field lines.
547:    The distance of penetration
548: depends on the ratio $r_A/R_*$.
549:  At relatively small values $r_A/R_*$, the equatorial funnels may
550: penetrate almost to the surface of the star, as shown in
551: Figure 6.
552:  At larger values of $r_A/R_*$,
553: the funnels move inward only part of the way.
554:   Thus, in the case of a
555: weak magnetic field and/or high accretion rate, the magnetospheric
556: radius $r_A\approx (1-3 )R_*$, and a  small adiabatic indexes
557: $\gamma$, the gap will not be empty and planets will continue to
558: migrate inward unless the tidal interaction or some other force will
559: prevent them against falling to the star.
560:    In the opposite case of a larger
561: magnetosphere, the equatorial funnels will occupy only a part of the
562: gap and planet may survive longer inside the innermost gap.
563: 
564: 
565: 
566: \section{Conclusions}
567: 
568: We conclude that for  typical parameters of solar-type
569: protostars a very low-density magnetospheric gap forms.
570:    We show that the rate of inward migration of protoplanets
571: is greatly reduced within this gap.
572:    This gap may explain the
573: observed accumulation of planets
574: at the distances $\sim (0.04-0.05)$
575: AU from the star.
576:    Note however that there may
577: be significant matter density
578: within $r_A$ in  cases of high misalignment angles
579: $\Theta \gtrsim 45^\circ$ where the part of the  funnel stream is
580: located in the equatorial plane.
581:    Further,  for low values of the
582: adiabatic index, $\gamma=1.1$, and  small misalignment angles,
583: $\Theta \lesssim 5^\circ$, matter may accrete in the equatorial
584: plane due to instabilities.  However, the azimuthal velocity
585: of the matter within $r_A$ differs substantially from
586: the Keplerian angular velocity of the planet.
587:  This difference in the
588: velocities acts to slow the planet's migration.
589: 
590: 
591: 
592: 
593: \acknowledgments{We thank Drs. G.V. Ustyugova
594: and A.V. Koldoba for valuable contributions to
595: our MHD simulation codes.  Also we thank the referee
596: for thoughtfull comments.
597: This work was supported in part by NASA grant
598: NNG05GL49G and  by NSF grant AST-0507760.}
599: 
600: 
601: \begin{references}
602: 
603: 
604: \reference{} Arons, J. \& Lea, S.M. 1976, ApJ, 207, 914
605: 
606: 
607: \reference{} Basri, G., Marcy, G.W., \& Valenti, J.A. 1992, ApJ,
608: 390, 622
609: 
610: \reference{} Boss, A.P. 2001, ApJ, 563, 367
611: 
612: 
613: 
614: \reference{}  Bouvier, J., Alencar, S.,  Harries, T.J.,
615:  Johns-Krull, C.M., \&  Romanova, M.M. 2006,
616: Protostars and Planets V, accepted
617: \smallskip
618: 
619: \reference{} Camenzind, M. 1990, Rev. Modern Astron., 3, 234
620: 
621: \reference{} Feigelson, E.D., \& Montmerle, T.  1999, Ann. Rev.
622: Astron. Astrophys., 37, 363
623: 
624: \reference{} Ghosh, P., \& Lamb, F.K. 1979, ApJ, 234, 296
625: 
626: 
627: \reference{} Johns-Krull, C.,  Valenti, J.A., \& Koresko, C. 1999,
628: ApJ, 516, 900
629: 
630: 
631: \reference{} Koldoba, A.V., Romanova, M.M., Ustyugova, G.V., \&
632: Lovelace, R.V.E. 2002, ApJ, 576, L53
633: 
634: 
635: \reference{} K\"onigl, A. 1991, ApJ, 370, L39
636: 
637: \reference{} Kuchner, M.J., \& Lecar, M. 2002, ApJ, 574, L87
638: 
639: 
640: \reference{} Lin, D.N.C., Bodenheimer, P., \& Richardson, D.C.
641: 1996, Nature, 380, 606
642: 
643: \reference{} Lin, D.N.C., \& Papaloizou, J. 1986, ApJ, 309, 846
644: 
645: 
646: \reference{} Marcy, G.W.,  Butler, R.P., Fischer, D.A., \& Vogt,
647: S.S. 2003, Scientific Frontiers in Research on Extrasolar Planets,
648: ASP Conference Series, Vol. 294, Eds. Deming, D. \& Seager, S.
649: 
650: \reference{} Matsuyama,I.,  Johnstone, D.,  \& Murray, N. 2003,
651: ApJ, 585, L143
652: 
653: \reference{} Mizuno, H. 1980, Prog. Theor. Phys., 64, 544
654: 
655: \reference{} Nelson, R.P., \& Papaloizou, J.C.B. 2003, MNRAS, 339,
656: 993
657: 
658: \reference{} Papaloizou, J.C.B., \& Terquem, C. 2006, Rep. Prog.
659: Phys., 69, 119
660: 
661: 
662: \reference{} Pollack, J.B., Hubickyj, O., Bodenheimer, P., et al.
663: 1996, Icarus, 124, 62
664: 
665: 
666: \reference{} Romanova, M.M., Ustyugova, G.V., Koldoba, A.V., \&
667: Lovelace, R.V.E. 2002, ApJ, 578, 420
668: 
669: \reference{} Romanova, M.M., Ustyugova, G.V., Koldoba, A.V., Wick,
670: J.V., \& Lovelace, R.V.E. 2003, ApJ, 595, 1009
671: 
672: \reference{} Romanova, M.M., Ustyugova, G.V., Koldoba, A.V., \&
673: Lovelace, R.V.E. 2004, ApJ, 610, 920
674: 
675: \reference{} Safier, P.N. 1998, ApJ, 494, 336
676: 
677: \reference{} Smirnov, D.A., Lamzin, S.A.,  Fabrika, S.N. , \&
678: Valyavin, G.G. 2003, A\&A, 401, 1057
679: 
680: \reference{} Symington, N.U., Harries, T.J., Kurosawa, R., \&
681: Naylor, T. 2005, MNRAS, 358, 977
682: 
683: \reference{} Valenti, J.A. \& Johns-Krull, C. 2004, Astrophys. Sp.
684: Sci. Ser., 292,619
685: 
686: \reference{} Ward, W.R. 1997, ApJ, 482, L211
687: 
688: 
689: \end{references}
690: 
691:  \end{document}
692: