Observations au sol

IMAGES OF TITAN IN THE NEAR-INFRARED WITH ADAPTIVE OPTICS

A. Coustenis1, M. Combes1, E. Gendron1, O. Lai1, M. Hirtzig1,

Th. Fusco2, L. Mugnier2, P. Rannou3, L. Vapillon1, J.-P. Véran4, J. Woillez1

1LESIA, Observatoire de Meudon, Meudon 92195 Cedex, France, Athena.Coustenis@obspm.fr,

2ONERA, France, 3Serv. d'Aéronomie, Univ. Versailles-St Quentin, France

4H. I. A., National Res. Council, Canada

Introduction

Titan, Saturn's largest Moon, has aroused a strong interest in the scientific community due to its thick atmosphere (1.5 bar) dominated by molecular nitrogen, where an active photochemistry occurs producing hydrocarbons, nitriles and more complicated organics including oxygenated compounds. These organics make the stratosphere of Titan opaque in the visible spectral range even in the conditions of a close fly-by (like Voyager 1).

            Titan's surface was expected to be covered, at least partially, by liquid hydrocarbons in form of lakes or even a global ocean. In the near-infrared range, from 1 to 5 micron, methane atmospheric « windows » allow us to probe the lower atmosphere and surface of the satellite.

            We have observed Titan in adaptive optics (or associated systems) since 1991 and until this year. We have essentially recovered diffraction-limited images since 1994 which we have analyzed with three main systems (ESO/ADONIS, CFHT/PU’EO and CFHT/OASIS), as follows:

Titan observations with adaptive optics 1994 - 2001

Year

Filter

    AO

Orb.

Date

0.83

0.94

1.08

1.09

1.18

1.29

1.57

1.60

1.64

2.0

2.2

2.26

system

phase

1994

14-18 Sep.

             

x

 

x

x

 

ADONIS

ESO

GEE

1995

7-13 Oct.

             

x

 

x

x

 

ADONIS

ESO

GWE

Sc

1996

4-11 Nov.

             

x

x

x

x

 

ADONIS

ESO

Sc

GEE

Ic

1997

6-10 Nov.

             

x

x

x

x

 

ADONIS

ESO

Sc

GEE

1998

26 Oct.

       

x

x

 

x

x

     

PUEO

CFHT

GEE

2000

8-11 Dec.

             

x

x

x

x

 

ADONIS

ESO

GWE

2000

17 Nov.

 

x

                   

OASIS

CFHT

>GEE

2001

6-7 March

4 Dec.

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

 

x

x

x

PUEO

CFHT

GEE

GEE

Orientation of the Titan images

The authors of various reports on Titan's images in recent papers have used different conventions for orientation and longitude notations (sometimes changing systems within the same paper). There are two orientation conventions which are usually used in the various papers related to spatially resolved images of Titan. The astronomical convention with North up and East-sky to the left of North (the point of view of an Earth observer looking at the sky with North above the head) and the cartographic - or geographic - convention with North up and East-map to the right of North (the point of view of an observer on the planetary body, with North above the head). Both conventions are identical in the projections of the images. They differ only in wording. HST papers (Smith et al. 1996, Meier et al. 2000) and Combes et al. (1997, Icarus, 129, 482-497) use the cartographic description, whereas Gibbard et al. (1999) use the astronomical one.

Fig. 1: Orientation conventions at GEE.

Given Titan's synchronous rotation, the sub-Saturn point is fixed and the sub-Saturn meridian is used by all of the authors as the origin of longitudes (L=0°). Then the Titan leading hemisphere is always facing Earth at Greatest Eastern Elongation (GEE) and the trailing hemisphere at GWE. GEE is the maximum geocentric angle between Saturn and Titan, Titan being on the East-sky side. GWE is the opposite configuration. The morning limb is then the anti-Saturn (West-map, East-sky) hemisphere at GEE and the sub-Saturn (West-map, East-sky) at GWE. Titan's rotation around itself and around Saturn both are counter-clockwise as seen from the North (Fig. 1). In other words, in its course around Saturn, Titan presents to the Earth successively the leading, the anti-Saturn, the trailing and the sub-Saturn hemispheres.

In articles dealing with Titan spectra, the Longitude of Central Meridian (LCM) convention is always used and hence there is no confusion. The reports of the authors dealing with Titan images differ mainly in longitude notations. HST (Smith et al. 1996, Meier et al. 2000) used geographical longitudes, with their maps centered at L=180° and East-map longitudes increasing to the right of the map and West-map longitudes increasing to the left and noted L° (W) or -L°.

Fig. 2: Longitude notations at GEE.

On the Gibbard et al. (1999) images, longitudes increase towards East-sky (to the left of North), with the reported dark regions at longitudes of 150 or 160 ° (Gibbard, private communication, 1997), which is the same as Combes et al. (1997) did, i.e. the LCM system with longitudes increasing towards the West-map (according to the IAU cartographic convention). Similarly to the latter, in Coustenis et al. (2001, Icarus, 154, 501-515) we use the cartographic convention (North up and East-map to the right of North).

Note that the cylindrical projection maps and/or orthographic projection hemispheric images published by the four teams are very similar and hence in agreement (e.g. Fig. 7 of Combes et al. 1997), with bright and darks spots seen on the same locations and rotating at the same rate from West-map to East-map. This was confirmed between this paper's images and HST images on a map projection hereafter (Fig. 3, Lorenz, private communication, 2000).

Fig. 3 Cylindrical projection of Titan images from Smith et al (1996) and from Coustenis et al. (2001). Provided by Lorenz, Dec. 2000, private communication.

ADONIS

K band images: leading and trailing hemispheres 1994-1995.

            The first spatially resolved image of Titan has been recorded in 1991 at the 3.6- meter telescope of ESO in Chile, equipped with an adaptive optics system (COME-ON +). This system is the first prototype of adaptive optics devoted to astronomy. It has been built for ESO by the Space Research Department (LESIA, ex DESPA) of Paris Observatory. A thin deformable miror actuated at high frequency associated with a wavefront sensor compensates in real time the deformations of the incident wavefront induced by the atmospheric turbulence. This demonstrated the feasibility of obtaining spatially resolved images of Titan in spite of its angular diameter as seen from the Earth (0.8 arcsec only) and of the effects of atmospheric turbulence.

            A large number of images were obtained since then and until 2000. Hereafter we discuss the images taken in September 1994, October 1995 and November 1996. These images have been recorded mainly through a set of narrow-band filters around 2.0 and 1.6 micron where the atmosphere of Titan is transparent because of very low molecular methane absorption. We also have CVF images.

The ADONIS/ESO Titan images at 2 micron were recovered in 1994 (at Titan's GEE, leading hemisphere) and in 1995 (at Titan's GWE, trailing hemisphere). The diffraction-limited high-contrasted ADONIS images (Combes et al., 1997, Icarus 129, 482) rivaled with the HST 1994 Titan images taken near 1 micron (Smith et al., 1996).

Fig. 4: Titan's leading side surface features at 2 micron (left) and near 1 micron (right) from ADONIS (Combes et al., 1997) and HST (Smith et al., 1996) images respectively.

            The surface images show a well-defined equatorial bright spot associated with smaller and less contrasted features near the poles, all rotating over 6 consecutive nights at the expected rotation rate of Titan's solid body. They are attributed to ground features at the surface. Our images are in agreement with HST images (Smith et al. 1996), both on the location and on the shape of the bright features. As expected, the contrast of the surface marks is higher in our images (about 30%), which are less affected by scattering extinction in Titan's atmosphere since they are recorded in the infrared range. The spatial resolution is at the limit of diffraction (0.13 arc sec), very similar to the resolution of HST images.

The high-contrasted ADONIS images in particular showed a) a well-defined bright spot on the leading hemisphere, with 2 or possibly 3 peaks (centered at around 110 degrees LCM and 10 degrees Southern latitude). Smaller and less contrasted features were found near the poles (Fig. 5); and for the first time b) high-latitude bright zones on the trailing hemisphere and possibly all over Titan’s disk (sort of polar caps).

Fig. 5: Titan ADONIS images taken in 1994 (leading side, right) and in 1995 (trailing side, left) at 2 micron. The images show surface features after deconvolution and subtraction of the atmospheric component have been applied. Note the large bright equatorial area on the leading side and the bright zones appearing on the trailing side. From Combes et al. (1997).

The bright high-latitude zones (we can’t really call them mid-latitude since they do not reach the equator, whereas we think they may well reach the poles) are clearly observed on the ADONIS images of Titan's trailing hemisphere in Combes et al (1997), where they were mentioned for the first time. They cover effectively latitudes between 25 and 50 ° N and 25 to 40° South (Fig. 5). They are indeed of lower contrast than the bright regions on the leading hemisphere, as remarked by Combes et al. (and also as noted by other investigators).


ADONIS images: the bright high-latitude zones. Are they real?

Since the bright regions outside the equator appear on several of the ADONIS images taken at different longitudes and at various wavelengths (corresponding to different limb brightening conditions), they can not well be « artifacts » of image subtraction remaining after careful reduction.

Furthermore, note that these bright areas:

-a- appear on the raw data before limb-brightening correction and deconvolution (Figure 6).

-b- appear on the Gibbard et al (1999) data with the same North/South contrast after a very different data reduction process.

-c- were recently confirmed also in the HST images of Meier et al (2001), again with a different processing method.

Their extension all the way to the poles cannot be assessed at the present time, in the lack of data on the scattering phase function of this area, which can result in an apparent darkening of the poles even if they are covered with some bright material.

Since two other teams have independently seen these features since our publication in 1997, we sustain our assertion that they are indeed "bright, high-latitude zones clearly observed" in the ADONIS images (and elsewhere) and we hope to bring more evidence and explanations for this phenomenon in a future study.

Fig. 6: ADONIS images of Titan's trailing hemisphere. Upper images are raw K1 and K2 images and lower image is the product after subtraction of K1-kK2. These images show that the bright areas appearing in the deconvolved trailing hemisphere 2-micron ADONIS images are not artifacts due to the deconvolution or further processes of the data, since they appear on the original images.

            ADONIS was, however, limited to long-wavelengths observations. In order to complete our dataset, we have therefore used the new adaptive optics system, PU’EO at the CFHT, in Hawaii, to observe Titan in 1997, 1998 and 2001.

PUEO

In 1998 our observing run corresponded to excellent seeing conditions (Coustenis et al., 2001). The Titan PU’EO observations of 1998 correspond to a sub-Earth latitude=15.5°, longitude = 92 °; phase angle = -0.5 °; and Titan angular diameter = 0.856 arcsec. Our high-resolution images in the J (1.3 mm) and H (1.6 mm) bands and the associated Point Spread Functions are of high quality : the Strehl ratios are of about 35-50 % and the contrast of 35-45 %, the highest reported at these wavelengths to this date. As a consequence, we have been able to apply sophisticated reduction to our data and the deconvolution processes have allowed us to recover about 10 individual resolution elements on Titan’s disk in J and in H.

            Having observed with narrow-band filters in the centers and the wings of the methane windows, we have recovered images that show only Titan’s atmosphere (J2 at 1.18 mm and H2 at 1.64 mm), and images that show the atmosphere and the surface (J1, H1). The J2 and H2 images provide information as to the optical depth of the aerosol in Titan’s stratosphere. They also allow us to subtract properly the atmospheric contribution from the others and thus recover information on the surface, with high signal-to-noise ratios.

Fig. 7 : [Raw images and PSF taken by PUEO in 1998. The seeing was 0.35 arcsec at best.

Reduction and deconvolution of PUEO images:

*    PSFs: obtained using the automatic AO reconstruction method from data accumulated by the AO system during the science exposure (Véran et al., 1997)

*    Deconvolution: different methods were used among which

Lucy Richardson and Maximum Entropy,

but also MCS (Magain et al., 1996) and

Myopic (Conan et al., 1998)

significantly reducing ringing artefacts while enhancing the contrast of the surface by smoothing small intensity fluctuations that result from noise, while preserving large intensity fluctuations corresponding to the edge of the planet.

*    Center-to-limb effects: corrected for J, H and K based on a model by McKay et al. (1989 - updated).

*    Atmospheric subtraction: using the Pa g, J2, H2, Hcont, FeII and K2 images.

Hereafter are two figures showing new deconvolution methods for J and H and also a figure showing 4 different methods applied to J1 and J2 images:

Fig. 8: Deconvolution of Titan J1 and J2 images with Lucy-Richardson, Wiener, MISTRAL and Magain methods (From Coustenis et al., 2001).

Fig. 9: Deconvolution of Titan H1 and H2 images with MISTRAL and Magain methods.

PU’EO results in 1998:

Atmosphere

Besides the well-known “Titan smile” or north-south asymmetry observed usually), our J2 and H2 images show a 10-17% brighter western than eastern limb (on the left of the images, Figs 8-9). We interpret this as a first detection of possible diurnal effects in Titan’s atmosphere (something like a fog, observed on the morning limb after condensation processes have enhanced the atmospheric levels between 70 and 100 km with condensates, Coustenis et al., 2001).

Surface

            The resulting images of Titan’s surface at Greatest Eastern Elongation were obtained for the first time at 1.3 micron and show the equatorial region bright again, extending from 70° to roughly 120° Longitude of Central Meridian, and over 30 degrees in latitude around 10° South. All our images, at the three wavelengths (1.3, 1.6 and 2.0 mm) show that the same morphology applies to all three frequencies. The bright area consists of at least three individual components, or peaks, should this be a mountainous plateau, which are 15-20 % brighter than the surrounding bright regions. A bright extent is also observed in the North-West and the South-West regions. We also observe somewhat darker Eastern and lower-western inlets. There is an even darker terrain in the North-East, at longitudes of about 110 E and close to the equator. This region is 3 to 3.3 times darker than the bright area. Based on spectroscopic measurements of Titan’s geometric albedo (from the FTS at the CFHT: Coustenis et al., 2000) and there-on based surface albedo spectra of Titan, we infer that dark region must correspond to roughly surface albedos of 0.1-0.3, whereas the brighter terrain albedos could be 0.5-0.9 in the J and H bands.

Fig. 10 : Surface and atmosphere of Titan observed by PUEO in the  J1 and J2 filters (upper part). Diurnal effect (fog ?) observed on the morning limb (on the left of J2). The bright equatorial area on the surface is resolved into several brightness peaks (lower part).

Fig. 11 : Titan’s surface observed with PU’EO on the leading hémisphère. The central area is similarly bright in J and H. It could be due to the presence of methane/ethane ice on top of a mounda. Darker spots observed in the North-Eastern part of the image indicate the presence of additional surface components.

            By comparison with spectra of various ices that could exist on Titan, we find that only methane ice or, more probably, ethane frost (perhaps on top of a mountain) could account for the bright area observed at 0.94 and 1.08 mm (from HST data), as well as at 1.3, 1.6 and 2.0 mm (from adaptive optics).

Possibility for methane or ethane snow on a top of a mountain on Titan

We are currently investigating the possibility for the presence of methane/ethane ice on Titan’s surface at the equator, on top of a mountain, or at higher latitudes, where the temperature is expected to be lower than at the equator. Some considerations, taken from Coustenis et al., (2001), are offered hereafter.

Apparently, in thermodynamic equilibrium conditions in the Titan atmosphere, methane ice could not subsist at altitudes lower than 12-14 km (as previously found by other investigators). In the case where the particles would diverge from thermodynamic equilibrium (which was our previous interest in the case of methane ice), it would seem after further investigation of the fusion and melting condition in a CH4-N2 mixture, that there wouldn't be much to gain in the fusion altitude level and that the drops would liquefy at temperatures higher than 80.6 K and they would then change their composition by evaporating CH4 or by condensing N2 so as to stay in equilibrium with the atmospheric composition. Hence, it may be difficult to form methane ice (or snow) at any reasonable (for a mountain) altitudes (B. Schmitt, private communication, 2001).

On the other hand, ethane snow may be a more adequate candidate at Titan conditions, since its solidification temperature is 89.3 K, which would produce snow at or above 3.5 km, if it was pure. In the case of a mixture with N2 and CH4, little data is available. The phase diagram of N2-C2H6(-CH4) exists but at higher T and p conditions than on Titan, and so we can not evaluate the effect of this mixture on the temperature for ethane snow given hereabove. However, C2H6 is less sensitive to the presence of nitrogen, because it is less volatile. We may therefore expect this temperature to not vary by much. On the other hand, the ethane flux estimated from current models for Titan is rather small (on the order of 1 micron per year). It would be premature at this stage to say whether ethane ice is or not a good possibility for Titan's bright spots, but perhaps this is something that can be tested observationally.

The strong methane ice absorption bands are shown in Fig. 17 of Schmitt et al. (1998, in Solar System Ices, Schmitt, de Bergh, Festou Eds., Kluwer ASSL Series 227, p. 227) where the absorption coefficient is high. The water ice spectrum is also shown in this figure. The other possible candidate's spectrum, that of ethane ice, can be found in Quirico and Schmitt (1997, Icarus 127, 354).

New PU’EO images (2001)

Dates : March 6-7 & December 4, 2001

Titan observed at or very close to GEE:(SSP: long=80-100°, lat=-23.4° in March 2001)

Wavelengths covered:

            Broad-band filters:                                         I (0.834±0.194 mm)

Narrow-band filters:                                       HeI (1.083 ± 0.009 mm)

PaGamma (1.094 ± 0.011 mm)

J1 (1.293 ± 0.139 mm)

                                                                                   J2 (1.181 ± 0.128 mm)

                                                                                   Jcontinuum (1.207 ± 0.015 mm)

                                                                                   Hcontinuum (1.570 ± 0.020 mm)

                                                                                   H1 (1.600 ± 0.160 mm)

                                                                                   H2 (1.640 ± 0.100 mm)

                                                                                   FeII (1.644 ± 0.015 mm)

                                                                                   Kcontinuum (2.260 ± 0.060 mm)

We have observed Titan more recently in 2001, on 7-8 March and 4 December. We have used several filters, included the specific ones hereabove, as well as other narrow-band filters particularly designed for our study (J1, J2, H1, H2). We have verified that similarly to the 1998 observations, in March and December of 2001 the phase effect should be on the east limb (evening side). In March, this effect is important (around 6°), whereas in December, it is minimal, thus perhaps allowing us to observe the morning fog effect on the west limb.

We have reduced our images as usual and applied deconvolution and center-to-limb corrections. The results are shown in the following figures.

Fig. 12 : Titan observed with PUEO in GEE on  7 March 2001. Note the phase effect (~6°) visible on the East limb (to the right) of the atmospheric images and the North-South asymmetry (with the South pole brighter on most of our images sounding the lower levels of the atmosphere).

With our December 2001 images we cover even more wavelengths than previously and we have been able to confirm that the phase effect is indeed minimum (not visible on the evening limb), whereas a suggestion of perhaps a morning fog may be visible on the Western limb on the stratospheric images FeII, J2 et Jcont. The North-South asymmetry appears as previously (since the 1992 first HST images of Titan, with the South brighter than the North) on the images sounding the lower atmosphere, but there is a suggestion of reversal (predicted by seasonal effect models) on the images sounding higher atmospheric levels (such as FeII and Kcont). After subtracting the atmosphere from the images sounding the CH4 windows, we recover information on the surface of Titan. We find the equatorial spot to be bright again in all the investigated filters from 1 –2 micron..

Fig. 13 : Titan observed with PUEO on 4 December 2001. Note the possible presence of the morning fog signature and of the inversion of the North-South asymmetry in filter FeII. The bright area is located near the upper-right edge of Titan’s images (we are at GEE+1 day).

Spectro-imaging of Titan with CFHT/OASIS (2000) 

The methane window at 0.94 micron is particularly important in our study because it may contain a signature of the tholin component (the tholin spectrum shows a strong decrease of the albedo at wavelengths shorter than 1 micron). We were able to observe Titan between 0.865 and 1.037 micron (hence covering part of the 0.83 micron and all the 0.94 micron windows). We have worked in mode LR2 with R=1000 and with a pixel size of 0.04 arcsec. The seeing was about 0.9 arcsec. After reduction of our data with XOasis, we have recovered about 10 independent resolution elements on the Titan disk (we are hoping to increase this number after deconvolution). Each element provides a spectrum associated with a region on Titan’s disk.

Fig. 14 : Titan observed with OASIS in November 2001.

            Photometry processes and more data are required to complete the analysis of Titan’s surface spectrum and to infer information on the morphology and the constituents on the ground. Understanding the nature of Titan's surface will provide major clues for understanding the prebiotic chemistry processes in a primitive planetary atmosphere. Our ground-based observations give us an efficient tool to optimize the observing program of the infrared instruments VIMS and DISR of the Cassini-Huygens ESA-NASA mission.

References: See Combes  et al. (1997) Icarus, 129, 482-497 and Coustenis et al. (2001) Icarus 154, 501-515 (and references therein).