1: %
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%LATEX (AASMS4 Style) %%%%%%%%%%%%%%%%%%%%%%%%
3: %
4: \documentstyle[12pt,aasms4]{article}
5: %\documentstyle[12pt,aaspp4]{article}
6: %\documentstyle[11pt,aas2pp4]{article}
7: %\documentstyle[11pt,eqsecnum,aaspp4]{article}
8: %\documentstyle[12pt,amssym,aasms4]{article}
9: %\received{4 August 1988}
10: %\accepted{23 September 1988}
11: %\journalid{337}{15 January 1989}
12: %\articleid{11}{14}
13: %\slugcomment{Submitted to the Astrophysical Journal Letters }
14: \lefthead{SENGUPTA \& KRISHAN}
15: \righthead{POLARIZATION FROM BROWN DWARF}
16: \begin{document}
17:
18: \title{PROBING DUST IN THE ATMOSPHERE OF BROWN DWARFS THROUGH POLARIZATION}
19: \author{SUJAN SENGUPTA\footnote{sujan@iiap.ernet.in} AND VINOD KRISHAN }
20: %\altaffilmark{1}}
21: \affil{ Indian Institute of Astrophysics, Koramangala, Bangalore 560 034,
22: India }
23:
24: \begin{abstract}
25:
26: Theoretical analysis and observational evidences indicate that
27: a brown dwarf with effective temperature greater than 1300 K would
28: have dust cloud in its atmosphere. In this letter, we show
29: that dust scattering should yield polarized continuum radiation
30: from the relatively warm brown dwarfs and the polarized flux profile
31: could be a potential diagnosis tool for the optical and the physical
32: properties of dust grains. The degree of polarization due to multiple
33: scattering will be more in the optical region if the particle size
34: is small while significant polarization should be detected in the
35: infra-red region if the particle size is large. It is pointed out
36: that the departure from sphericity in the shape of the object due to
37: rapid rotation and due to tidal effect by the companion in a binary
38: system ensures the disc integrated polarization to be non-zero.
39:
40: \end{abstract}
41:
42: \keywords{radiative transfer - scattering - polarization - stars:atmosphere
43: - stars:brown dwarfs}
44:
45: \section{INTRODUCTION}
46:
47: The synthetic continuum spectra of the brown dwarf Gliese 229B has been obtained
48: by several authors.
49: Marley et al. (1996), Griffith, Yelle \& Marley (1998), Saumon et al (2000),
50: obtained an overall good fit of the observed spectra of Gliese 229B over a wide
51: range of wavelengths and attributed the very rapid decline of the observed
52: continuum flux in the optical region to the dust particulates in its atmosphere.
53: On the other hand, Tinney et al. (1998); Kirkpatrick et al. (1999);
54: Tsuji, Ohnaka, \& Aoki (1999), Burrows, Marley, \& Sharp
55: (2000) argued
56: that the pressure broadened red wing of the 0.77 $\mu m$ K I doublet and
57: the 0.5891, 0.5897 $\mu m$ Na I D
58: could account for the observed features of the continuum flux shortward
59: of 1.1 $\mu m$ (for a recent review, see Burrows, Hubbard, Lunine \& Liebert 2001).
60: The first optical spectrum of a T Dwarf SDSS 1624+0029 favors
61: the latter explanation(Liebert et al. 2000) and at present
62: there is enough evidence that dust particulates cannot exist in the
63: visible atmospheres of comparatively cool T dwarf such as Gl 229B. However,
64: dust can be present at the visible height of the atmosphere of
65: warmer brown dwarfs and L dwarfs with effective temperature
66: greater than 1300 K.
67:
68: Very recently, Allard et al. (2001) have presented a detail theoretical investigation
69: on the effects of dust in the atmospheres of brown dwarfs by considering two
70: limiting cases: (a) inefficient gravitational settling of the dust grains
71: and (b) efficient gravitational settling. In the former case the dust is distributed
72: according to the chemical equilibrium predictions and provides the maximum impact
73: upon the atmosphere while in the latter case the dust forms and depletes refractory
74: elements from the gas yielding a minimal effect on the brown dwarf atmosphere.
75: When the effective temperature of the object is bellow 1300 K complete
76: gravitational settling of dust grains should occur whereas gravitational
77: settling cannot occur when the effective temperature is above 1800 K.
78: In the intermediate case ($T_{eff}=1400$ K) an optically thick dust cloud
79: forms in the visible atmosphere (Marley and Ackerman 2001).
80:
81: In this letter, we show that significant polarization can arise due to
82: dust scattering in the atmosphere of relatively warmer brown dwarfs where
83: dust grains have not been settled completely. In the presence of dust
84: particulates of small size, the continuum flux in the near optical region
85: could be polarized significantly while non-zero polarization can occur in
86: the infra-red if the particle size is large. Hence observation of polarized
87: radiation from brown dwarfs could be a potential probe for the properties
88: of dust grains.
89:
90: \section{NONSPHERICITY OF RAPIDLY ROTATING BROWN DWARFS AND
91: INTRINSIC POLARIZATION BY DUST SCATTERING}
92:
93: Polarization of radiation due to scattering in stellar atmospheres or
94: envelopes has been considered by many authors. Numerous
95: polarimetric data and their interpretation have been
96: presented by Gehrels (1974). Since noticeable deviation from
97: sphericity are to be expected from rapidly rotating stars, much
98: attention has been paid to calculations of intrinsic polarization
99: of Be stars. Degree of polarization varying from 0.1 \% in the
100: visible region to 2\% in the far infra-red has been observed in Be stars
101: by Rucinski (1970), Collins (1970), Harrington and Collins (1968), Haisch
102: and Cassinelli (1978). Extensive observational data for polarization
103: from cool stars are given by Serkowskii (1967, 1968). It has been
104: established by many observations that the intrinsic polarization of radiation
105: is a common behavior for most young stars (see for example Menard \&
106: Bastien 1992, Yudin \& Evans 1998 and references therein).
107:
108: Observation of non-zero polarization from unresolved objects
109: indicates non-sphericity of the stars. The non-sphericity of an
110: object, which leads to incomplete cancellation of the polarization
111: of radiation from different areas of the surface, may result from
112: several causes. The rotation of a stellar object imparts to it the shape
113: of an oblate ellipsoid. This is evident in
114: the outer solar planets as well. The eccentricities of Jupiter,
115: Saturn and Uranus are 0.35, 0.43, 0.21 respectively at 1 bar pressure
116: level. Tidal interaction with the companion in a binary
117: system imposes an ellipsoidal shape extended towards the companion.
118:
119: Spectroscopic studies by Basri (1999) indicate rapid rotation of
120: brown dwarfs along their axis. A general trend to higher velocities
121: as one looks to objects of lower luminosity is found in this study.
122: The brown dwarf Kelu 1 is found to be the fastest rotator having
123: the projected angular velocity $v\sin i$ as high as 80 km/s corresponding to
124: an angular velocity of $1.16\times 10^{-3} s^{-1}$.
125: The surface gravity of brown dwarf varies from $10^5 cms^{-2}$
126: to $3\times 10^5 cms^{-2}$ implying a mass range from 36 $M_J$ to
127: 73 $M_J$ (Marley et al 1996).
128: As a consequence, a fast rotating brown dwarf will have significant
129: departure from sphericity.
130:
131: The maximum possible oblateness of a stable rotating fluid body
132: having polytropic equation of state with different indices
133: has been derived by Chandrasekhar (1933). The relation between the
134: eccentricity $e$ and the rotational velocity $\omega$ for a
135: rigidly rotating body (Maclaurine spheroid) with uniform density is given by :
136:
137: \begin{equation}
138: \omega=\left[2\pi G \rho \left\{\frac{(1-e^2)^{1/2}}{e^3}(3-
139: 2e^2)\sin^{-1}e-\frac{3}{e^2}(1-e^2)\right\}\right]^{1/2},
140: \end{equation}
141: where $\rho$ is the density of the fluid. Here the eccentricity is defined
142: as $e=(1-r_p^2/r_e^2)^{1/2}$ where $r_p$ and $r_e$ are the polar and the
143: equatorial radius respectively. The eccentricity `e' and
144: the oblateness `q' are related by the expression $e^2=1-(1-q)^2$.
145:
146: Adopting the empirical relationship given in Marley et al. (1996)
147: for mass and radius of brown dwarf, the mean density of a brown
148: dwarf with $g=10^5 cms^{-2}$ and $T_{eff}=1500 K$ would be $51 gmcm^{-3}$.
149: For $\omega=1.16\times 10^{-3} s^{-1}$, the above formula yields
150: $e=0.48$. However, in reality the density is not uniform and
151: the above formula yields slightly higher value of e compared to
152: the same for non-uniform density distribution as can be
153: verified by applying to the case of Jupiter, Saturn and Uranus.
154: Nevertheless, the rapid rotation of brown dwarfs would certainly
155: impose deviation from sphericity in its shape and hence the disc
156: integrated polarization will not be cancelled out.
157:
158: The dependence of polarization due to single scattering by grains on the
159: oblateness of an object has been discussed by Dolginov, Gnedin \&
160: Silant'ev (1995) and by Simmons (1982). Following Simmons (1982), the
161: analytical expression (first order approximation) for the degree of
162: polarization due to single scattering from a spheroid with
163: uniform density can be written as :
164: \begin{equation}
165: p=\lambda^{2}(R_1-R_2)n_0K_2(\frac{5}{96\pi})^{1/2}P^2_2\cos(i)\exp
166: (-2i\phi)F_{22}
167: \end{equation}
168: where $F_{22}$ depends on the scattering phase function, $R_1$ and $R_2$
169: are the outer and the inner equatorial axis lengths, $n_0$ is the
170: particle number density, $P^2_2$ is the associated Legendre
171: polynomial of order 2 and $\lambda$ is the wavelength. The expression for
172: $K_2$ depends on the density distribution and hence on the eccentricity
173: of the spheroid. The values of $F_{22}$ for different values of $ x=2\pi a/\lambda$,
174: $a$ being the radius of the scattering particle, as well as the expression
175: for $K_2$ are given in Simmons
176: (1982). The first order approximation is valid when $x\leq 2$.
177: We have calculated the degree of polarization as a function of eccentricity $e$
178: by considering Mie scattering. Figure~1 presents the degree of
179: polarization as a function of eccentricity viewed edge on
180: i.e., when $i=\pi/2$ and $\phi=0$.
181:
182: \placefigure{figure1}
183:
184: Figure~1 shows that the qualitative feature of polarization
185: does not change with the change in the eccentricity. This is
186: found by Simmons (1982) as well. Polarization increases with
187: the increase in the eccentricity. For $x\geq 2$ the change
188: in the degree of polarization is small. For all the values of
189: $x$ the variation in polarization is not too high when $e$ varies
190: from 0.2 to 0.4. In our calculation we have taken $n_0a^2=100$.
191: The choice is arbitrary as the main purpose of Figure~1 is to show
192: the effect of eccentricity on the degree of polarization. In the
193: realistic case, the density must be non-uniform and the polarization
194: would occur due to multiple scattering. Therefore one has to solve
195: the equations for the transfer of polarized radiation with the
196: incorporation of Mie theory of scattering.
197:
198: \section{TRANSFER EQUATIONS FOR POLARIZED RADIATION}
199:
200: The Stokes parameters representing a linearly polarized beam are given by
201: $I=I_l+I_r$ and $Q=I_l-I_r$ where $I_l$ and $I_r$ represent two mutually orthogonal
202: states of linear polarization. The degree of polarization is given by $p=Q/I$.
203:
204: For a plane parallel atmosphere the equation of transfer can be written as
205: (Chandrasekhar 1960) :
206: \begin{equation}
207: \mu\frac{d{\bf I}(\tau,\mu)}{d\tau}={\bf I}(\tau,\mu)-\frac{\omega_0}{2}\int_{-1}^{1}
208: {\bf{P}(\mu,\mu')\bf{I}(\tau,\mu')d\mu'},
209: \end{equation}
210: where
211: $$\bf{I}(\tau,\mu)=\left(\begin{array}{c}I_l(\tau,\mu) \\ I_r(\tau,\mu)\end{array}\right),$$
212: $\mu$ is the cosine of the angle made by the ray to the normal, $\tau$ is the optical depth, $\omega_0$ is the albedo for single scattering and
213: $\bf{P}(\mu,\mu')$ is the azimuth-independent phase matrix.
214:
215: \section{THE ATMOSPHERE MODEL}
216:
217: The values of the effective
218: temperature $T_{eff}$ and the surface gravity $g$ of brown dwarf
219: are constrained by the bolometric luminosity
220: and the evolutionary sequence of the object (Saumon et al 1996).
221: In the present letter, we have adopted a model atmosphere
222: with $T_{eff}=1500K$, $log(g)=5.0$ and $[M/H]=-0.3$ (K band). The
223: temperature-pressure profile for model with $log(g)=5.0$ and $T_{eff}=1030 K$
224: has kindly been provided by M. Marley (private communication) and
225: the opacity data by D. Saumon. Using an iterative process for temperature
226: correction we have obtained the temperature profile for $T_{eff}=1500 K$. In
227: this process we have used the temperature profile for $T_{eff}=1030K$ as
228: the initial temperature profile. We have ignored any change in the
229: density profile with the increase in the temperature. We have incorporated
230: the dust opacity by using the Mie theory of scattering. Following Griffith,
231: Yelle and Marley (1998), and Marley \& Ackerman (2001) we have considered
232: a lognormal spherical particle size distribution. The mean radius of the
233: silicate grain (real part of the refractive index 1.65) has been considered
234: as $ 0.1 \mu m$ and $1.0 \mu m$. Although it is difficult to infer
235: the grain sizes by direct observation, atmospheres of the outer solar
236: planets indicate that the mean radius
237: of the grains is expected to be as high as $10 \mu m$ to $100 \mu m$
238: Marley \& Ackerman 2001). However in the
239: present investigation we have considered smaller particles as the effect of
240: particle size on the degree of polarization can well be visualized without
241: going for larger particle size.
242:
243: \section{RESULTS AND DISCUSSION}
244:
245: We solve the transfer equations for the polarized radiation by using discretization
246: method. The numerical procedure is described in detail by Sengupta (1993).
247:
248: In the absence of a magnetic field, a radiation field could be polarized only
249: by scattering. The extent of scattering is determined by the albedo $\omega_0$
250: and the angular distribution of the scattered photon is governed by the phase
251: function.
252: The albedo for single scattering is
253: defined as $\omega_0=(\sigma_R+\sigma_M)/\chi_{\nu}$ where $\sigma_R$ and
254: $\sigma_M$ are the cross-section for the Rayleigh and the Mie scattering respectively
255: and $\chi_{\nu}$ is the total extinction co-efficient. While the Rayleigh phase function
256: is symmetric in the sense that the amount of backward and forward scattering
257: are the same, the Mie phase function is asymmetric. The larger the size of the dust
258: particulate, the greater the effect of the Mie scattering.
259:
260: When there are dust clouds at or above the photosphere, the albedo
261: is high. Conversely, when clouds are absent, the albedo in the mostly absorbing
262: atmosphere is low (Burrows et al. 2001). As a consequence,
263: in the atmosphere of cool brown dwarf such as Gl229B, contribution of scattering
264: is negligible and so there will be no polarization.
265: If the mean particle size is small, say $0.1 \mu m$,
266: the albedo beyond near infra-red
267: is zero as the Mie scattering cross section goes over to that of the Rayleigh and
268: the scattering cross section of various molecules present in the atmosphere
269: is negligibly small or zero. Therefore scattering polarization can arise only up to the optical
270: and near infra-red region where the contribution of scattering cross section to the
271: total extinction co-efficient is significant.
272: It should be mentioned that grains with small size cannot
273: contribute significantly to the opacities in the near-infrared where water
274: bands still dominate the spectra of brown dwarf from J to K band.
275:
276: On the other hand if we consider larger particle size, then scattering
277: will contribute to the far infra-red region as well.
278:
279: \placefigure{figure2}
280: The degree of polarization due to multiple scattering is presented
281: in figure~2. Here we notice that the
282: degree of polarization is almost zero at $1.34 \mu m$ and onwards when the
283: mean particle size is $0.1 \mu m$.. Shortwards
284: of $1.3 \mu m$ the degree of polarization increases.
285: This is because of the fact that scattering is negligible at wavelengths greater
286: than $1.3 \mu m$. As the wavelength decreases, the effect of dust scattering
287: increases. As a consequence the radiation field gets polarized due to
288: scattering and the degree of polarization increases with the decrease in
289: the wavelength. The degree of polarization always remains negative.
290:
291: If we consider dust particles with larger mean radius, say $1 \mu m$, then
292: most of the flux get blocked in the optical region. As a result
293: the degree of polarization decreases significantly in the optical region.
294: However, when the size of the particle is comparable to the wavelength,
295: degree of polarization increases significantly. Figure~2 shows that
296: the degree of polarization decreases in the optical region if the
297: the particle size is comparatively large. In the infra-red and far infra-red,
298: degree of polarization increases if the grain
299: size is increased. It's worth mentioning that multiple scattering
300: decreases the degree of polarization as compared to that with single
301: scattering.
302:
303: \section{CONCLUSION}
304:
305: The important message that is conveyed in this letter is that the
306: continuum radiation from the atmosphere of relatively warm brown
307: dwarfs where the dust particulates are not gravitationally
308: settled should be polarized and the polarized radiation could provide
309: a lot of information on the properties of dust particulates.
310: Since, no polarization will occur in the absence of dust grains,
311: observation of polarized radiation will also help in deciding
312: if a particular brown dwarf contains dust in its atmosphere.
313: Rapid rotation would impose non-sphericity in the shape of
314: brown dwarfs. As a result, a net non-zero polarization integrated
315: over the stellar disc should be obtained.
316: The results strongly suggest that the polarimetric
317: observations posses a great diagnostic potential for
318: the understanding of the properties of dust particulates.
319: Furthermore, if no polarization is observed from any particular
320: brown dwarf then that will indicate the absence of dust in
321: its atmosphere.
322:
323: \acknowledgments
324: We thank M. Marley for kindly providing the temperature-pressure profile
325: and D. Saumon for the continuum opacity data. We are indebed to A. B.
326: Ravindran for many valuable discussions and to the referee for
327: constructive comments, suggestions and criticisms. Thanks are due to N. K. Rao,
328: P. Bhattacharya, H. C. Bhatt and T. P. Prabhu.
329:
330:
331:
332: \begin{thebibliography}{}
333: \bibitem {} Allard, F., Hauschildt, P. H., Alexander, D. R., Tamanai, A., \& Schweitzer, A.
334: 2001, astro-ph/0104256
335: \bibitem {} Basri, G. 1999, Proc. American Astron Soc. 194, 8208
336: \bibitem {} Burrows, A., Hubbard, W. B., Lunine, J. I., \& Liebert, J. 2001, preprint
337: \bibitem {} Burrows, A., Marley, M. S., \& Sharp, C. 2000, \apj, 531, 438
338: \bibitem {} Chandrasekhar, S. 1960, Radiative Transfer (Dover, New York)
339: \bibitem {} Chandrasekhar, S. 1933, \mnras, 93, 390
340: \bibitem {} Collins, G. W. 1970, \apj, 159, 583
341: \bibitem {} Collins, G. W. 1972, \apj, 175, 147
342: \bibitem {} Dolginov, A. Z., Gnedin, Yu. N. \& Silant'ev, N. A. 1995,
343: Propagation and Polarization of Radiation in Cosmic Media (Gordon and Breach, Basel)
344: \bibitem {} Gehrels, T. Ed. 1974, Planets, Stars and Nebulae Studied with
345: Photopolarimetry (University of Arizona, Tucson)
346: \bibitem {} Griffith, C. A., Yelle, R. A., \& Marley, M. S. 1998, Science, 282, 2063
347: \bibitem {} Haisch, B. M. \& Cassinelli, J. P. 1978, \apj, 220, 101
348: \bibitem {} Harrington, J. P. \& Collins, G. W. 1968, \apj, 151 1051
349: \bibitem {} Kirkpatrick, J. D., et al. 1999, \apj, 519, 834
350: \bibitem {} Liebert, J., Reid, I. N., Burrows, A., Burgasser, A. J., Kirkpatrick,
351: J. D., \& Gizis, E. 2000, \apj, 533, L155
352: \bibitem {} Marley, M., Saumon, D., Guillot, T., Freedman, R., Hubbard, W. B.,
353: Burrows, A., \& Lunine, J. I. 1996, Science, 272, 1919
354: \bibitem {} Marley, M. S. \& Ackerman, A. S. 2001, astro-ph/0103269
355: \bibitem {} Menard, F. \& Bastien, P. 1992, \aj, 103, 564
356: \bibitem {} Rucinski, S. M. 1970, Acta Astron. 20, 1
357: \bibitem {} Saumon, D., Geballe, T. R., Leggett, S. K., Marley, M. S., Freedman, R. S.,
358: Lodders, K., Fegley Jr. B., \& Sengupta, S. K. 2000, \apj, 541, 374
359: \bibitem {} Sengupta, S. 1993, \mnras, 265, 513
360: \bibitem {} Serkowskii, K. 1967, \apj, 144, 3
361: \bibitem {} Serkowskii, K. 1968, \aj, 73, 8
362: \bibitem {} Simmons, J. F. L. 1982, \mnras, 200, 91
363: \bibitem {} Tinney, C. G., Delfosse, X., Forveille, T., \& Allard, F. 1998,
364: A \& A, 338, 1066
365: \bibitem {} Tsuji, T., Ohnaka, K., \& Aoki, W. 1999, ApJ, 520, L119
366: \bibitem {} Yudin, R. V. \& Evans, A. 1998, A \& AS, 131, 402
367: \bibitem {} Yudin, R. V. 2000, A \& AS, 144, 285
368: \end{thebibliography}
369:
370: \clearpage
371:
372: \figcaption[f1.eps]{Degree of polarization as a function of
373: eccentricity for different values of $x=2\pi a/\lambda$. Single scattering on a
374: spherical grain in an oblate spheroid with uniform density is considered.
375: \label{figure1}}
376:
377: \figcaption[f2.eps]{Degree of polarization by multiple scattering.
378: \label{figure2}}
379:
380:
381: \end{document}
382:
383:
384: