- Email: [email protected]
The polarization of Titan
ICARUS
18,
661-664
(1973)
The
Polarization
of Titan
BEN ZELLNER Lunar
and
Planetary
Laboratory, Received
The
University September
of Arizona,
Tucson,
Arizona
85721
16, 1972
New polarization observations of Titan in three spectral regions are presented. The results are not consistent with scattering from either an ordinary planetary surface or a pure molecular atmosphere. Apparently an opaque cloud layer with strongly uv-absorbing constituent is needed.
Titan, the largest satellite of the solar unique in system, is photometrically several respects. These include limb darksmall variations of ening, vanishingly brightness and color with orbital revolution and solar phase angle (Harris 1961; Veverka 1970), a steep linear drop in albedo from 0.64 to 0.38p.m (McCord, Johnson and Elias, 1971) and suppressed thermal emission at 20p.m (Morrison, Cruikshank and Murphy 1972). While these peculiarities are presumably produced by Titan’s atmosphere, none of them have been adequately explained. Veverka (1970, 1973) has demonstrated that the linear polarization of Titan is positive, reaching +0.2% to +0.3% at phase angle 6” in his white-light observations. All other small solar-system bodies, by contrast, show negative polarizations ranging from -0.2% to -1.7% at small phase angles (e.g., Dollfus 1971). The absence of a negative polarization branch for Titan might conceivably be due to a layer of pure snow (Lyot 1929) or to a microscopically smooth surface of bare rock such as a glassy lava. However the albedo of Titan is too low for snows, and dust-free surfaces are otherwise unknown in the solar system. Thus we are led to attribute Titan’s polarimetric behavior to scattering in its atmosphere. A new polarimetric survey of the satellites of Saturn was begun in 197 1. The twochannel photoelectric polarimeter described by Coyne and Gehrels (1967) was used at the 154-cm Catalina reflector of Copyright 0 1973 by Academic Press, Inc. All rights of reproduction in any form reserved.
a
the Lunar and Planetary Laboratory, at the 229-cm reflector of Steward Observatory, and at the 61- and 224-cm instruments of the Institute for Astronomy, University of Hawaii. Observations of Iapetus have been reported elsewhere (Zellner 1972); preliminary results for Titan are given in this paper. Table I gives polarization observations of Titan in three wavelength bands. Since the polarizations are generally only a few times larger than the random probable errors, careful control of systematic errors is essential. The Wollaston prism was oriented at only the two precomputed position angles corresponding to polarization perpendicular to the Sun-TitanEarth plane (positive) and parallel to that plane (negative). Usually two unpolarized standard stars were observed at the same Wollaston angles, immediately before and after the Titan observations; the instrumental polarization was typically found to be less than 0.02%. Halation from Saturn and its rings presented a minor problem only in the ultraviolet filter. As illustrated in Fig. 1, the new measurements are in qualitative agreement with Veverka’s results; however the rather strong wavelength dependence, with negative polarization in the ultraviolet near 4” phase, was previously unsuspected. Figure 2 gives theoretical results from the Rayleigh-Chandrasekhar theory for the polarization produced by a pure molecular atmosphere above a (nonpolarizing) Lambert surface, as a function 661
662
ZELLNER TABLE POLARIZATION
Date
UT
71 71 72 72 72 72 71 71 71 71 71 72 72 72 72 72 72
Oct. Oct. Jan. Feb. Feb. Aug. Oct. Oct. Oct. Nov. Dec. Jan. Feb. Feb. Feb. Feb. Sep.
72 72 72 72
Jan. Mar. Sep. Sep.
I
OBSERVATIONS
OF TITAN Solar phase angle
AIP*
T’
9.42 19.32 2.12 7.18 24.14 31.57 9.32 19.29 20.29 18.30 30.10 28.10 1.26 7.26 22.13 24.20 5.57
Cl54 5229 Cl54 Cl54 Cl54 H224 Cl54 5229 S229 Cl54 Cl54 Cl54 Cl54 Cl54 5229 Cl54 H 61
64 45 80 73 53 30 20 20 18 130 34 20 19 9 8 36 64
0.36/.~ 0.36 0.36 0.36 0.36 0.36 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52
4.88 4.09 4.04 6.11 6.25 6.28 4.89 4.09 4.00 0.94 3.76 5.77 5.93 6.11 6.26 6.25 6.36
+8 +4 $6 -0 +2 $16 +20 +24 +17 +17 +23
2 5 3 2 3 3 5 5 5 3 6
28.22 5.15 1.55 12.43
Cl54 Cl54 H224 Cl54
33 61 66 101
0.83 0.83 0.83 0.83
5.77 6.10 6.30 6.39
+21 +31 +32 +25
5 9 9 6
0 Total integration time * Difference in measured
h
Polarization (x104)
Tel.
in minutes. polarization
between
(x104)
+4*5 -8 12 -16 11 t-8 6 +13 11
+8
two
1 2 0
6 1 0 1 2 1 1
1
3
unpolarized
stars. *
I
PX
0.3 -
+
+AA
4
0.2 -
+
c.1-+
+
+
+ +
0
l 0.0 -
4.1
0
0
l
Aot
l & 0 00
0
0
-
0 I I
t 2
3
4
5
6
SOLAR PHASE ANOLE FIG. 1. Polarization observations of Titan. Open light (0.36pm), filled circles in the green (0.52 pm), Crosses represent white-light observations by Veverka
circles indicate observations in and triangles in the near infrared (1970).
ultraviolet (0.83p.m).
POLARIZATION
I
/ N
663
OF TITAN
0.02
/
/
I
0.06
al0
I
Q20
0.40
0.60
GEGMETRIC ALBE&
FIG.
pure global limited unity
2. Theoretical contours of constant polarization (per cent) at 6” phase angle, for a Rayleigh atmosphere of optical depth r above a Lambert sphere. The abscissa is the geometric albedo of the atmosphere plus the underlying surface. The range of models is by the condition that the ground reflectivity must fall between zero (top heavy line) and (heavy line on the right).
of the optical depth of the atmosphere and the global geometric albedo of the atmosphere plus the ground. The calculations were made for a spherical planet by means of a computer program originally written by J. Dave and modified by D. L. Coffeen. The low surface brightness of Titan puts upper limits on the optical depth of a conservative Rayleigh atmosphere : The arrows in Fig. 2 correspond to geometric albedos of 0.07 and 6.22 in the U and V bandpasses, respectively, obtained from the diameter of 0.70 arcseconds according to Dollfus (1970), an absolute V magnitude of -1.17 (Harris 1961; Veverka 1970), and the spectrophotometry of McCord et al. (197 1). The limit of zero ground reflectivity then requires that rn < 0.12 and 7v < 0.45. The observed polarization of +O.lS% at phase angle Co in the green filter (albedo
0.19) corresponds in Fig. 2 to ho = 0.11. However this model fails to explain the observations in several respects: (1) The Rayleigh optical depth would violate the albedo constraint at 0.36pm, and at 0.83El.m it would be much too low to account for the observed +0.25% polarization at that wavelength. (2) The observed wavelength dependence of polarization is in the wrong direction for Rayleigh scattering. In particular, negative polarizations are never produced, even if suspended uv absorbers are introduced to reduce the effective scattering optical depth at short wavelengths, (3) At optical depth 0.11 in green light, the negatively polarizing ground still contributes more than two-thirds of the total brightness. Hence an opaque cloud layer, which can itself introduce strong polarization of
664
ZELLXER
either sign, must be invoked. (4) The steep drop in green-light polarization from 6” to 4’ phase is not reproduced in the Rayleigh models. Thus, the polarimetry confirms an already-suspected conclusion, that the atmosphere of Titan cannot be described by a clear gas above a diffusely reflecting surface. In fact, there is nothing in the polarization data which suggests molecular scattering. As yet we have no real understanding of Titan’s atmosphere. More extensive polarization observations, together with models which take into account the polarizing effects of a cloudtop, are needed. ACKNOWLEDGMENT The National Science Steward, Catalina, and tories provided essential
Foundation and the Mauna Kea Observasupport of this work.
REFERENCES COYNE, G.V.,AND
GEHRELS, T. dependence of polarization. polarization. A&m. J. DOLLFIJS, A. (1970). Diametres satellites. In “Surfaces and Planets and Satellites” (A. length stellar
(1967).
WaveX. Inter72, 887. des plan&es et Interiors of Dollfus, ed.),
Chapt. York.
2. Academic
Press,
London
and
New
DOLLFUS, A. (1971).
Physical studies of asteroids by polarization of the light. In “Physical Studies of Minor Planets” (T. Gehrels, ed.), pp. 95-115. NASA SP-267. HARRIS, D. L. (1961). Photometry and colorimetry of planets and satellites. In “Planets and Satellites” (G. P. Kuiper andB. M. Middleburst, eds.), Chapt. 8. University of Chicago Press, Chicago. LYOT, B. (1929). Research on the polarization of light from planets and from some terrestrial substances. Ann. Z’Obs. Paris VIII, No. 1.
NASA TT F-187. MCCORD,T.B.,JOHNSON, (1971). Saturn speetrophotometry
and
T.V., ANDELIAS,J.H. its satellites : Narrow-band (0.3-1.1 pm). Astrophys. J.
165, 413. MORRISON,DAVID,CRUIKSHANK,DALE P., AND MWPHY, ROBERT E. (1972). Temperatures of Titan and the Galilean satellites at 20 microns. A&o&/s. J. (L&t.) 173, L143. VEVERKA, J. (1970). “Photometric and Polarimetric Studies of Minor Planets lites.” Ph.D. thesis, Harvard Cambridge, Massachusetts. VEVERKA, J. ( 1973). Titan : polarimetrio for an optically thick atmosphere?
657-660. ZELLNER, B. (1972). Aetrophys.
J. (Lett.)
On
the
nature
174, L107.
and SatelUniversity, evidence Icarus of Iapetus.
18,
18,
661-664
(1973)
The
Polarization
of Titan
BEN ZELLNER Lunar
and
Planetary
Laboratory, Received
The
University September
of Arizona,
Tucson,
Arizona
85721
16, 1972
New polarization observations of Titan in three spectral regions are presented. The results are not consistent with scattering from either an ordinary planetary surface or a pure molecular atmosphere. Apparently an opaque cloud layer with strongly uv-absorbing constituent is needed.
Titan, the largest satellite of the solar unique in system, is photometrically several respects. These include limb darksmall variations of ening, vanishingly brightness and color with orbital revolution and solar phase angle (Harris 1961; Veverka 1970), a steep linear drop in albedo from 0.64 to 0.38p.m (McCord, Johnson and Elias, 1971) and suppressed thermal emission at 20p.m (Morrison, Cruikshank and Murphy 1972). While these peculiarities are presumably produced by Titan’s atmosphere, none of them have been adequately explained. Veverka (1970, 1973) has demonstrated that the linear polarization of Titan is positive, reaching +0.2% to +0.3% at phase angle 6” in his white-light observations. All other small solar-system bodies, by contrast, show negative polarizations ranging from -0.2% to -1.7% at small phase angles (e.g., Dollfus 1971). The absence of a negative polarization branch for Titan might conceivably be due to a layer of pure snow (Lyot 1929) or to a microscopically smooth surface of bare rock such as a glassy lava. However the albedo of Titan is too low for snows, and dust-free surfaces are otherwise unknown in the solar system. Thus we are led to attribute Titan’s polarimetric behavior to scattering in its atmosphere. A new polarimetric survey of the satellites of Saturn was begun in 197 1. The twochannel photoelectric polarimeter described by Coyne and Gehrels (1967) was used at the 154-cm Catalina reflector of Copyright 0 1973 by Academic Press, Inc. All rights of reproduction in any form reserved.
a
the Lunar and Planetary Laboratory, at the 229-cm reflector of Steward Observatory, and at the 61- and 224-cm instruments of the Institute for Astronomy, University of Hawaii. Observations of Iapetus have been reported elsewhere (Zellner 1972); preliminary results for Titan are given in this paper. Table I gives polarization observations of Titan in three wavelength bands. Since the polarizations are generally only a few times larger than the random probable errors, careful control of systematic errors is essential. The Wollaston prism was oriented at only the two precomputed position angles corresponding to polarization perpendicular to the Sun-TitanEarth plane (positive) and parallel to that plane (negative). Usually two unpolarized standard stars were observed at the same Wollaston angles, immediately before and after the Titan observations; the instrumental polarization was typically found to be less than 0.02%. Halation from Saturn and its rings presented a minor problem only in the ultraviolet filter. As illustrated in Fig. 1, the new measurements are in qualitative agreement with Veverka’s results; however the rather strong wavelength dependence, with negative polarization in the ultraviolet near 4” phase, was previously unsuspected. Figure 2 gives theoretical results from the Rayleigh-Chandrasekhar theory for the polarization produced by a pure molecular atmosphere above a (nonpolarizing) Lambert surface, as a function 661
662
ZELLNER TABLE POLARIZATION
Date
UT
71 71 72 72 72 72 71 71 71 71 71 72 72 72 72 72 72
Oct. Oct. Jan. Feb. Feb. Aug. Oct. Oct. Oct. Nov. Dec. Jan. Feb. Feb. Feb. Feb. Sep.
72 72 72 72
Jan. Mar. Sep. Sep.
I
OBSERVATIONS
OF TITAN Solar phase angle
AIP*
T’
9.42 19.32 2.12 7.18 24.14 31.57 9.32 19.29 20.29 18.30 30.10 28.10 1.26 7.26 22.13 24.20 5.57
Cl54 5229 Cl54 Cl54 Cl54 H224 Cl54 5229 S229 Cl54 Cl54 Cl54 Cl54 Cl54 5229 Cl54 H 61
64 45 80 73 53 30 20 20 18 130 34 20 19 9 8 36 64
0.36/.~ 0.36 0.36 0.36 0.36 0.36 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52
4.88 4.09 4.04 6.11 6.25 6.28 4.89 4.09 4.00 0.94 3.76 5.77 5.93 6.11 6.26 6.25 6.36
+8 +4 $6 -0 +2 $16 +20 +24 +17 +17 +23
2 5 3 2 3 3 5 5 5 3 6
28.22 5.15 1.55 12.43
Cl54 Cl54 H224 Cl54
33 61 66 101
0.83 0.83 0.83 0.83
5.77 6.10 6.30 6.39
+21 +31 +32 +25
5 9 9 6
0 Total integration time * Difference in measured
h
Polarization (x104)
Tel.
in minutes. polarization
between
(x104)
+4*5 -8 12 -16 11 t-8 6 +13 11
+8
two
1 2 0
6 1 0 1 2 1 1
1
3
unpolarized
stars. *
I
PX
0.3 -
+
+AA
4
0.2 -
+
c.1-+
+
+
+ +
0
l 0.0 -
4.1
0
0
l
Aot
l & 0 00
0
0
-
0 I I
t 2
3
4
5
6
SOLAR PHASE ANOLE FIG. 1. Polarization observations of Titan. Open light (0.36pm), filled circles in the green (0.52 pm), Crosses represent white-light observations by Veverka
circles indicate observations in and triangles in the near infrared (1970).
ultraviolet (0.83p.m).
POLARIZATION
I
/ N
663
OF TITAN
0.02
/
/
I
0.06
al0
I
Q20
0.40
0.60
GEGMETRIC ALBE&
FIG.
pure global limited unity
2. Theoretical contours of constant polarization (per cent) at 6” phase angle, for a Rayleigh atmosphere of optical depth r above a Lambert sphere. The abscissa is the geometric albedo of the atmosphere plus the underlying surface. The range of models is by the condition that the ground reflectivity must fall between zero (top heavy line) and (heavy line on the right).
of the optical depth of the atmosphere and the global geometric albedo of the atmosphere plus the ground. The calculations were made for a spherical planet by means of a computer program originally written by J. Dave and modified by D. L. Coffeen. The low surface brightness of Titan puts upper limits on the optical depth of a conservative Rayleigh atmosphere : The arrows in Fig. 2 correspond to geometric albedos of 0.07 and 6.22 in the U and V bandpasses, respectively, obtained from the diameter of 0.70 arcseconds according to Dollfus (1970), an absolute V magnitude of -1.17 (Harris 1961; Veverka 1970), and the spectrophotometry of McCord et al. (197 1). The limit of zero ground reflectivity then requires that rn < 0.12 and 7v < 0.45. The observed polarization of +O.lS% at phase angle Co in the green filter (albedo
0.19) corresponds in Fig. 2 to ho = 0.11. However this model fails to explain the observations in several respects: (1) The Rayleigh optical depth would violate the albedo constraint at 0.36pm, and at 0.83El.m it would be much too low to account for the observed +0.25% polarization at that wavelength. (2) The observed wavelength dependence of polarization is in the wrong direction for Rayleigh scattering. In particular, negative polarizations are never produced, even if suspended uv absorbers are introduced to reduce the effective scattering optical depth at short wavelengths, (3) At optical depth 0.11 in green light, the negatively polarizing ground still contributes more than two-thirds of the total brightness. Hence an opaque cloud layer, which can itself introduce strong polarization of
664
ZELLXER
either sign, must be invoked. (4) The steep drop in green-light polarization from 6” to 4’ phase is not reproduced in the Rayleigh models. Thus, the polarimetry confirms an already-suspected conclusion, that the atmosphere of Titan cannot be described by a clear gas above a diffusely reflecting surface. In fact, there is nothing in the polarization data which suggests molecular scattering. As yet we have no real understanding of Titan’s atmosphere. More extensive polarization observations, together with models which take into account the polarizing effects of a cloudtop, are needed. ACKNOWLEDGMENT The National Science Steward, Catalina, and tories provided essential
Foundation and the Mauna Kea Observasupport of this work.
REFERENCES COYNE, G.V.,AND
GEHRELS, T. dependence of polarization. polarization. A&m. J. DOLLFIJS, A. (1970). Diametres satellites. In “Surfaces and Planets and Satellites” (A. length stellar
(1967).
WaveX. Inter72, 887. des plan&es et Interiors of Dollfus, ed.),
Chapt. York.
2. Academic
Press,
London
and
New
DOLLFUS, A. (1971).
Physical studies of asteroids by polarization of the light. In “Physical Studies of Minor Planets” (T. Gehrels, ed.), pp. 95-115. NASA SP-267. HARRIS, D. L. (1961). Photometry and colorimetry of planets and satellites. In “Planets and Satellites” (G. P. Kuiper andB. M. Middleburst, eds.), Chapt. 8. University of Chicago Press, Chicago. LYOT, B. (1929). Research on the polarization of light from planets and from some terrestrial substances. Ann. Z’Obs. Paris VIII, No. 1.
NASA TT F-187. MCCORD,T.B.,JOHNSON, (1971). Saturn speetrophotometry
and
T.V., ANDELIAS,J.H. its satellites : Narrow-band (0.3-1.1 pm). Astrophys. J.
165, 413. MORRISON,DAVID,CRUIKSHANK,DALE P., AND MWPHY, ROBERT E. (1972). Temperatures of Titan and the Galilean satellites at 20 microns. A&o&/s. J. (L&t.) 173, L143. VEVERKA, J. (1970). “Photometric and Polarimetric Studies of Minor Planets lites.” Ph.D. thesis, Harvard Cambridge, Massachusetts. VEVERKA, J. ( 1973). Titan : polarimetrio for an optically thick atmosphere?
657-660. ZELLNER, B. (1972). Aetrophys.
J. (Lett.)
On
the
nature
174, L107.
and SatelUniversity, evidence Icarus of Iapetus.
18,