Skip to main content

Full text of "NASA Technical Reports Server (NTRS) 19870012330: Research in planetary studies and operation of the Mauna Kea Observatory"

See other formats


1 


4 


APPENDIX / * 


UNIVERSITY OF HAWAII 

INSTITUTE FOR ASTRONOMY 

2680 Wood I awn Drive 
Honolulu, Hawaii 96822 


C /€. 

6>7?z.^ 

f 


NASA GRANT NGL 12-001-57 


SEMIANNUAL PROGRESS REPORTS #32 and #33 

Dale P. Cruikshank, Principal Investigator 


( NAS A -CR- 1805 13) cf eAUbA REA 

'iDDlES ASD 0fEEil ^? al progress Report, 

«*«!«£ G3/89 

J fiu» u 

92 p 


887 - 217*3 


Unclas 

43361 


For the Period 
January-December 1986 



1 


I 

TABLE OF CONTENTS 

Page 

I . PERSONNEL 1 

II. THE RESEARCH PROGRAMS 2 

A. Highlights 2 

B. The Major Planets 3 

C. Planetary Satellites and Rings 23 

D. Asteroids 50 

E. Comets 62 

F. Laboratory Studies of Dark Organic Materials 70 

G. Theoretical and Analytical Studies: Thermal Inertias 

and Thermal Conductivities of Particulate Media 72 

H. Extrasolar Planetary Material: The Search for 

Dark Companions of K and M Giants 77 

III. OPERATION OF THE 2.2-METER TELESCOPE 79 

IV. PAPERS PUBLISHED OR SUBMITTED FOR PUBLICATION IN 1986 84 

ATTACHMENT: "Albedo Maps of Comets P/Giacobini-Zinner 

and P/Halley," H amine 1 et al 86 


i 



I . PERSONNEL 


This report covers the period January-December 1986. Scientific person- 
nel engaged in planetary research who were supported fully or in part by this 
grant during the report period area as follows: 

D. P. Cruikshank D. Morrison C. B. Pilcher 

W. M. Sinton M. W. Buie D. J. Tholen 

In addition, graduate students A. D. Storrs, H. B. Hammel, J. R. 
Piscitelli, and I. Heyer received salary and/or travel support on research 
assistantships . K. Uchida and Dr. N. Lark also received salary and travel 
support while working on special research projects during the year. 


1 



II. THE RESEARCH PROGRAMS 


A. HIGHLIGHTS 

1. Determination of the physical and orbital parameters of Pluto and Charon, 
including their diameters and the mean density of the system. 

2. Acquisition of the best images of Neptune ever taken, showing cloud 
patterns and limb brightening in the light of methane, 8900 A. 

3. Determination of the rotation period of Neptune from high-resolution 
imagery, giving a result consistent with other determinations by 
photometry and imagery. 

4. Acquisition of high-resolution images of Uranus with photometric calibra- 
tion for studies of the planet's atmosphere. 

5. Continued study of the thermal properties of the Galilean satellite Io from 
infrared observations during eclipses and occultations , and by polarimetry. 

6. Discovery of major changes in the spectrum of Triton from observations 
with the cooled-grating array spectrometer in the near-infrared. 

7. Study of the infrared spectrum of Io and a determination of the isotopic 
content of the sulfur dioxide snow on the satellite's surface. 

8. Detection of a nonwater-ice volatile on Europa and its apparent time 
variability. 

9. Observations of stellar occultations by the rings of Uranus with model fitting. 

10. Observations of the rotational lightcurves of numerous peculiar asteroids, 
including Trojans, Hildas, and other peculiar objects. 

11. Continued study of infrared spectra of asteroids in an effort to trace 
the origins of the S types and the A types . 

12. Discovery of the C-H organic signature in the infrared spectrum of the wet 
C-type asteroid 130 Elektra. 

13. Photometry of 11 planet-crossing asteroids. 

14. Acquisition of an enormous body of photometric, spectroscopic, photo- 
graphic, and electronic (CCD) imagery of Comet Halley, the analysis of 
which is in progress. 


2 



* 


15. Laboratory study of a suite of dark materials, mostly organic in nature, 
in connection with our study of the dark material on solar system bodies 
(comets, planetary satellites, asteroids). 

16. Breakthrough in the modeling of thermal inertias and thermal conductiv- 
ities of solar system bodies. 

17. Study of very high-resolution spectra of several K and M giant stars in 
the search for velocity shifts indicative of planet-size bodies in orbit 
around these stars. 

18. Use of 32% of 2.2-m telescope time for solar system studies of interest to 

NASA during the calendar year 1986. 

B. MAJOR PLANETS 

Bl. Outer Planet I magi ng-Nep tune and Uranus 


Over 500 images of Neptune and nearly 150 images of Uranus were obtained 
by H. Hammel in three observing runs in 1986 using three different methane- 
band filters and three continuum filters (Figs. 1-3). Most of the images are 
good quality; seme of the images of Neptune are of exceptional quality and may 
represent the finest images of this planet to date. 

Discrete cloud features were observed on Neptune in many images. One par- 
ticularly bright feature was observed crossing the planetary disk on at least 
two nights (Fig. 2) and was used to determine an atmospheric rotation period 
(Hammel and Buie, 1987). The measurements from a single night are not suffi- 
cient to calculate a rotation period accurately because of possible confusion 
from fainter spots, ambiguity in the pole position of Neptune, and other 
problems. Fortunately, the problems are minimal when a feature transits 
(crosses the central meridian of the planet as seen from Earth). By combining 
the observations from a night, the transit time can be calculated very accu- 
rately (Fig. 4). From the difference in transit times separated by many days, 
an atmospheric rotation period of 17.86 (± 0.02) hours was obtained for lati- 
tude -38° (± 2°). This period is consistent with earlier observations of 
cloud motion on Neptune. Hammel is working on determining the rotation period 
of some of the fainter features at different latitudes in hopes of deriving 
the vertical shear. The imaging method supersedes previous photometric work 
because the precise latitudes of atmospheric activity can be isolated. 

The distribution of bright clouds on Neptune appears to have changed sig- 
nificantly since 1983. All images obtained prior to 1984 showed bright fea- 
tures of comparable magnitude in both the northern and southern mid-latitude 
regions, giving the planet the appearance of having a dark equatorial belt. 

But in 1986, no bright features were detected in the northern hemisphere 


3 


t 


(Fig. 1). The imaging covers the full rotation cycle of Neptune over many 
weeks. Activity in the southern hemisphere is confined to latitudes -30° 
through -60°, consistent with that seen earlier. But activity in the northern 
hemisphere must have been subdued, at least during the months of May and June. 
The time scale of change of features in the southern hemisphere seems to be on 
the order of weeks to months. At least one fainter feature was seen to 
brighten over a period of three weeks to the same level as the bright feature 
discussed above. 

Center-to-limb brightness variations are visible on both planets. 
Methane-band images show limb-brightening; continuum images show either limb- 
darkening or a uniform light distribution. The excellent quality of the images 
of Neptune allows different center-to-limb profiles to be extracted for vari- 
ous regions on the planet. A mean profile was obtained for the dark region by 
summing a series of radial profiles at different position angles across the 
northern hemisphere. By measuring the brightness of the brightest feature at 
a variety of positions on the disk, a center-to-limb profile was created for a 
bright region. Hammel is working with colleague K. Baines at JPL to generate 
atmospheric vertical structure models that fit these two differing regions. 

On Uranus, the peak of the brightness distribution in the continuum 
wavelengths is offset from the center of the planetary disk, consistent with 
Voyager observations indicating a polar haze. The methane-band images show 
very obvious limb-brightening (Figs. 3 and 5). The very good signal-to-noise 
in these Uranus images will allow refinement of atmospheric vertical structure 
models by comparison with the observed radial brightness variations. Data 
reduction is in progress. 

B2. Photometry of Plato 

A major effort was devoted to the observation of Pluto-Charon mutual 
events. Tholen and Buie, with assistance from Lark and Storrs, succeeding in 
observing thirteen different events that occurred during the 1986 opposition 
of Pluto. Most of these events were observed through a Johnson B filter with 
the 2.2-m telescope, while a few were observed through an ultraviolet-blue 
blocking filter with the #1 0.61-m telescope on nights that were primarily 
devoted to photometry of Comet Halley. Representative data sets are shown in 
Figures 6-8. Figure 6 shows the inferior event (Charon passing in front of 
Pluto) of 15 January 1986 UT. The ordinate is the mean opposition blue 
magnitude and the abscissa is the Plutocentric ephemeris time. Coverage began 
with Pluto rising at 4 airmasses and continued until the onset of astronomical 
twilight. With the light gathering power of the 2.2-m telescope, it was 
possible to obtain a photometric resolution of approximately 0.004 mag per 
minute. At this level of precision, small deviations from the best fit model 
lightcurve are detectable. In particular, the end of the event seems to be 


4 



ORIGINAL PAGE IS 
OF POOR QUALITY 




ORIGINAL PAGE IS 
OF POOR QUALITY 



NEPTUNE TIME SEQUENCES 


20 May 1986 


4 June 1986 











8900 h 
LOU LIGHT 


LEVELS 




8 


Universal Time (hours) 



Profiles Across Uranus 



Distance (arcsec) 



15.90 



Photometric observations of Charon passing 
in front of Pluto 











systematically high compared to the model lightcurve (solid line), which may 
well be due to a dark spot on the southern limb of Pluto (see below). 

Figure 7 shows the inferior event of 11 June 1986 UT. Again the abscissa 
is the Plutocentric ephemeris time, but this time the ordinate is a synthetic 
mean opposition blue magnitude. The data were actually obtained through the 
ultraviolet-blue blocking filter with the 0.61-m telescope. This filter 
transmits all light longward of about 5300 A. The GaAs photomultiplier tube 
used for all the observations begins to lose sensitivity rapidly longward of 
9000 A and is essentially insensitive at 9300 A, so the effective bandpass of 
the filter-PMT combination is roughly 4000 A. This wider bandpass is neces- 
sary to increase the photon count rate to a high enough level to do useful 
time-resolved photometry with the 0.61-m telescope. 

Figure 8 shows the superior event of 27 June 1986 UT observed with the 
2.2-m telescope. The ordinate and abscissa are the same as for Figure 7. All 
but three of the data points were obtained through the blue filter. The 
remaining three, shown as filled circles, were obtained with the blocking 
filter normally used at the 0.61-m telescope. These three points clearly show 
that the event depth is less in yellow light than it is in blue light, which 
means that Charon's reflectance spectrum is not as red as Pluto's reflectance 
spectrum. If the entire surface of Charon was the same color, however, an 
inferior event ought to be deeper in yeiiow light than in blue light. Figure 
7 shows yellow light data plotted along with the blue light model lightcurve, 
and the data agree with the model rather well. Apparently the Pluto-facing 
hemisphere of Charon has a fairly neutral color, while the anti-Pluto 
hemisphere has approximately the same color as Pluto. Charon is yet another 
example of a solar system object with quite different hemispheres. 

Figure 8 also shows a bump shortly after first contact, which may well be 
due to a small-scale surface albedo feature on Charon. Supporting data were 
obtained by R. Binzel at McDonald Observatory with their 2.1-m telescope. All 
data collected so far would seem to suggest that surface albedo features 
produce very subtle deviations from smooth model lightcurves, which means that 
large-aperture telescopes are required to extract such information from the 
mutual events. 

Analysis of these data has yielded the most accurate values yet obtained 
for the various orbital and physical parameters for the Pluto-Charon system. 
These parameters are summarized in Table I. The derived density is surpris- 
ingly high. Given the water-ice compositions of many of the satellites of the 
outer planets and the methane-frost surface material known to exist on Pluto, 
most people were expecting the density of Pluto to be closer to 1.0 gm/cc. A 
density of 1.8 gm/cc would imply that at least half the mass of Pluto is due 
to rocky material, although further refinement of the density is necessary to 
place tighter constraints on the rock/ice ratio. 


13 



Table I 


PLUTO-CHARON ORBITAL AND PHYSICAL PARAMETERS 



a = 

19,130 


e = 

0.0 


i = 

91.6 

± 1.6 

n = 

222.44 

± 0.15 

L = 

122.03 

± 0.22 


E = 2,446,600.5 


km (assumed) 

(assumed) 

deg~\ equator and equinox 

deg_y of 1950 

deg 


P = 6.387204 ± 0.000047 dags 


Pluto radius = 1145 ± 46 k.m 

Charon radius - 642 ± 34 k.m 

5um of Pluto and Charon radii * 1786 ± 19 k.m 


Pluto blue geometric albedo = 0.612 ± 0.017 ”\ occulted material 
Charon blue geometric albedo = 0.424 ± Q.018 _/ onlg 


Pluto reflectance spectrum: reddish 

Charon reflectance spectrum: neutral (Pluto-facing hemisphere) 


Mean density of system = 1.84 ± 0.19 gm/cc 




Although observations of Pluto-Charon mutual events are being obtained 
worldwide, approximately 40X of all data have been collected at Mauna Kea 
since the onset of the events in late 1984. 

The heavy observing schedule triggered by the appearance of Comet Halley 
provided us with a golden opportunity to monitor the rotational lightcurve of 
Pluto frequently. Altogether, Tholen and Buie collected 46 points on the 
lightcurve, which is shown in Figure 9. A detailed comparison of this light- 
curve with the one assembled by Tholen and Tedesco from data collected between 
1980 and 1983 shows that the lightcurve is continuing to evolve slowly. In 
particular, maximum light appears to have become slightly broader and 
brighter. These data have been used to develop a model for the surface albedo 
distribution of the entire surface of Pluto (see below). 

Eventually, we hope to combine all data collected at various observa- 
tories worldwide to determine definitive orbital and physical parameters for 
the Pluto-Charon system. The success of this effort will be dependent upon 
how well each observers can transform their instrumental data to a common 
photometric system. To aid in this effort, Tholen and Buie have been select- 
ing and standardizing comparison stars to be used worldwide during each oppo- 
sition. In addition, a couple of transformation stars have been selected and 
are being standardized to help observers determine their color terms to the 
necessary precision. 

Of course, to acquire Pluto-Charon mutual event data, one needs to know 
when to observe. One could simply add half-integer multiples of the orbital 
period to the time of one event to determine when future events will occur, 
but the changing geometry of the Earth, Sun, Pluto, and Charon produces time 
shifts that can lead to changes in the times of events by more than a half 
hour. Many observers require more accurate information regarding the times of 
the various contacts to plan their observations properly. Once again, the 
best available orbital and physical parameters were used by Tholen, Buie, and 
Swift to generate event circumstances for the 1987 opposition. This table of 
circumstances was published along with the selected comparison stars at the 
beginning of the 1987 opposition. Figure 10 is excerpted from that paper. 

As the 46 new measurements of the brightness of Pluto have shown, the 
lightcurve continues to evolve. Persistent monitoring of the out-of -eclipse 
brightness is crucial to making the most of the mutual event observations. 

Buie has compiled a complete list of all Pluto photometry that spans the years 
1953 to 1986. We now have a data base comprising over 340 points. A small 
excerpt from this data base is shown in Figure 11, where both the orbital 
lightcurve (formerly referred to as the secular dimming) and the rotational 
lightcurve are seen for different times since 1953. 

A model was constructed to try and reproduce the lightcurve with a fixed 
surface albedo distribution. The model parameters include two bright polar 
caps, a dark equatorial band, and two spots near the equator. The sizes and 


15 




Mag v. 

Rot at i ona 1 p h as< 






albedos of these surface features were optimized with a nonlinear least- 
squares analysis run on the CRAY XMP/48 at the San Diego Supercomputer Center. 
After more than 100 hours of execution time we have two possible surface 
albedo models that reproduce the the observed photometric behavior of Pluto to 
a high degree of accuracy. Table II summarizes the model parameters. The two 
models are very similar, with the only exception being the cause of the light- 
curve from maximum light to the shelf near 330° E longitude. The smaller spot 
(*2) is either brighter than the equatorial band and causes the maximum or is 
darker than the background and is responsible for the shelf. Both models pre- 
dict a dark equatorial band which has accurately predicted that the mutual 
events in 1987 would not be as deep as predicted by Tholen, Buie and Swift. 
With data of the quality that has been taken at the 2.2-m on previous events 
it will be possible to discriminate between the two models when the geometry 
is right sometime in 1987 or 1988. 

Figure 12 shows an example of the quality of the model fit to the data. 
The fit is in general this good for all of the data. When grouped by the year 
of observation the typical residual is about 0.01 mag with some of the better 
years as low as 0.007 mag. For the first time we have a model that can 
explain the lightcurve of Pluto with geometry and a surface albedo model. We 
have generated some synthetic images of Pluto from the surface albedo model. 

A view of the shelf model can be seen in Figure 13. These are all equatorial 
views of Pluto with each image at 90° from the previous. In all images plane- 
tary North is up; this is intended to be a view that one might see from a 
spacecraft. 


B3 . Stellar Occultatlons 


A number of predicted stellar occultatlons by Uranus, Neptune, and Pluto 
were observed by Buie, Tholen, Cruikshank, and W. K. Hartmann. 

Confirmation of a dip in the signal as due to ring material generally 
requires observations at two different telescopes. For a subset of the obser- 
vations made at the IRTF, supporting observations were made at the 2.2-m tele- 
scope through an 8600-A filter. Both Uranus and Neptune are much brighter at 
8600 A than at 2.2 Jim, so the signal-to-noise ratio (SNR) is much lower. 
Although events corresponding to full extinction of the star could have been 
detected, the SNR was inadequate to confirm any of the short-duration partial- 
extinction dips seen in the IRTF data. 

A stellar occultation by Pluto was predicted to occur over the Pacific 
region on 30 April 1986 UT, and the nominal ground- track was predicted to pass 
over Mauna Kea. We monitored the occultation star from both the IRTF and the 
#1 0.61-m telescope under marginal photometric conditions, but unfortunately 
no event was seen. Even though mutual event photometry of Pluto is yielding 
size information, a stellar occultation remains a valuable tool that provides 
a way to probe the atmosphere of Pluto, should one exist. 


19 



0) 

rH 

CO 

H 


1 

VO 

00 
Os 

1-H 

1 
vq 
cm 

2 

43 

E £ 

00 


s 


Iff 

o 

IJ3 


4) 

IP'S 


>» 

a ! m 

M* 


«> O oo N 'O in N h cn «n ^ h h \o 

oodcn^^H^osp^cir^doddt 

cocoso^oor-t^sor^vcoor"-^ cm so co 


M H (S 1H H 





It 

i s 


04 


oo 

.5 


CM ^ ^ ^ ^ 


CM ^ 


vOO'ON^NOQOhN'OHHWOOt 

Tr^ooosTrr^oosocoTrcocMTrvocNCN© 

S222S£SS2222£i[iiis 


SSSSoooopoooooppp 

OOOOOOOOOOOOOCJOOO 

'OTrTrvo^f^’mt^*vor^r^OO<NcMcn 

©dddddddddddddddd 


oooopooooooo<Nmpo^r 

doddodddddddddodd 

-H-H-H-H-H-H-H-H-H-H-H-H-H-H-H-H-H 

Otit , ; tfl ooosQ rt ooh»noo'OOV)oo 
rs\qyiyo^^^<s^^oooq»Nvt^Ncn 
OO^h HHio^n^dd'riricnt'n 

oo 


J« 


o 

© 

© 

© 

© 

SSSSS 

9 

9 




-S' 

o 

o o o 

© © 

© © © 

El 

Ed 




11 

J 

-H 

-H 

-H 

+\ 

-H 

-H 

+1 

-H 

-H 

+1 

-H -H 

-H 

M 

-H -H 

4H 

F 

w 

w 

VO 

d 

£ 2 
© - 

p 

^ co rf 'M' 

os oo oo so os 

d d d xr c4 

i t »n 
© 

co d © 

333 

00 

CO 

CM 

8 

CM 

SO 



88SS8 

© 







A 

O 

O 

O 

© 

© 

© © © 

© 

El 

El 

Ed 

El 





i 

-H 

-H 

+1 

-H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 


CO 

vs 

S 

cm 

OS 

r- 

so 

so 

$ 

00 

Os 

CM 

CO 


8 

r- 

© 

CO 

so 

8 

r- 

r- 

CO 




«-H 




cm 


so 

d 


00 

cm 

CO 

CM 

CM 

CM 

CO 

CM 



















oo 




»-* 




ro 














8. 

8 

8. 

8 

8 

8 

8. 

8. 

8. 

8. 









y 

O 

O 

© 

© 

© 

© 

© 

© 

© 

© 









D 

-H 

+1 

+1 

-H 

■H 

-H 

-H 

-H 

+1 

■H 

-H 

-H 

-H 

-H 

-H 

-H 

-H 


w 

VO 

o 


co 

so 

so 

CO 

Os 

r-* 

CO 

v-4 

© 

Os 

$ 



oo 


1 

so 

s 

S. 

co 

3 

<N 

00 

s2 

VO 

8 


© 

r-* 

r- 

§ 

8 


■-§ 

ON 

so 

M3 

s 

5 ; 

CM 

SO 


So 

<n 

cm 

3 

<m 

3 

CM 

"=r 

CO 

tt 

?3 

SO 


s 

oo 

co 

i 

oo 

co 

8 

r- 

CO 

»-h 

© 

3 

3 

CO 

CO 

© 

CO 

8 

d 

CO 

oo 

© 

© 

CM 

S 

CO 



co 

SO 

co 

SO 

co 

so 

CO 

SO 

CO 

SO 

CO 

so 

CO 

so 

CO 

so 

so 









** 

** 






H 

•-4 



4>» 

1 


w 

>-H 

04 

HH 

04 


04 

M 

04 

M 

04 


04 

HH 

04 

04 


« so in ws ^ ^ a a 00.00. r r ^ >-«o *o co 


I 


8 .! 


20 


L 




Pluto -Charon Spot Model 


10 

CO 



9pn:jm§Bj*i g uuosutfof 


21 


100 150 200 250 300 

East Longitude (degrees) 



ORIGINAL PAGE IS 
OF POOR QUALITY 


Pluto Spot Model 



/scr/tests/ rbig 
8BIT ~ CLIP 

Wed Apr 8 10:02:53 1987 



0. [ 3.90625e-03 ] 1.00000 


Figure 13 


22 






C. PLANETARY SATELLITES AND RINGS 


Cl. Eclipse Observations of Io, the Thermal Inertia of the Bright Regions, 
and Changes In the Thermal Output of the Lokl Volcano 

In the report of this grant from last year, Sinton reported that he got 
satisfactory agreement with the observed eclipse cooling and heating curves of 
Io only when a two-albedo model was used and that the best agreement, which 
was statistically significant, was found when the different albedo regions had 
different thermophysical parameters as well. It was found that the bright 
regions probably had ten times the thermal inertia of the dark regions. One 
possible conjecture for this difference is the presence of SO 2 fas in the 
interstices between particles. This possibility will be discussed later in 
this report in a section 6. 

The thermophysical parameters were determined solely from the extensive 
data set obtained in 1983. An important test of any model is whether it has 
predictive ability, i.e., if parameters are changed does it give agreement 
with observations. A parameter that does change significantly from year-to- 
year is the distance of Jupiter, and hence Io, from the Sun. In 1981 Jupiter 
was at aphelion; in 1986 it was at perihelion. This results in a distance 
change of nearly 5% and an insolation change of nearly 10%. Although this is 
a relatively small change in total flux, it will have larger consequences at 
8.7 pm, a wavelength that is considerably shorter than the blackbody maximum 
for Io's normal emission and an even smaller fraction of the wavelength of the 
radiation peak when Io is eclipsed. Thus, with the change in solar distance 
from 1983, when the model parameters were set, to the 1986 observations we 
expect a substantial increase in the predisappearance flux from Io and an even 
greater change in the decrease in flux accompanying the eclipse. 

Sinton and Charles Kaminski have continued their program of making IRTF 
observations of eclipse disappearances and reappearances. Their results for 
1983 and 1986 are shown in Figure 14 along with the model predictions. Notice 
the marked increase in the depth of the eclipse between 1983 and 1986 not only 
in the model curves but also in the observational data. The only change in 
the model was the indicated change in the solar distance, R. The model, as 
mentioned, was fit to the 1983 data, not only at 8.7 pm but at four other 

wavelengths out to 30 pm. The solar distance effect is much smaller at these 

wavelengths, as expected, and the model is in good agreement with the observa- 
tions at the longer wavelengths. 

Another striking difference between the 1983 data and curves and those is 
the "jump" in the middle. The jump is a consequence of the thermal emission 

of the volcanoes, and the change in the aspect of these between observations 

made at an eclipse disappearance at a phase angle of ~9° before opposition to 
those made at a reappearance at a similar angle after opposition. The model 


23 



5.e-06 



S.a-06 



Figure 14. Observations at 8.7 4m of eclipses of Io in 1983 and 1986 com- 
pared. The thermophysical model parameters were established from the 1983 
data (obtained when the Io-Sun distance was 5.45 AU) . They not only fit the 
1983 data, but they also provide a good fit to the 1986 data (when the Io-Sun 
distance was 4.97 AU). Note the marked change in the depth of the eclipse 
between the two epochs. The "jump" in the middle of the curves is due to the 
change in aspect of the volcanic sources between viewing them at disappear- 
ances and reappearances. The amount of the required step is a measure of the 
amount of volcanic thermal emission. The amount decreased markedly in 1986 
compared to previous years. 


24 





curves partake in this change between the two seasons because the parameters 
specifying the volcanic emission are fitted by a least-squares routine, while 
the thermophysical parameters were not changed. The observations at the other 
wavelengths and the inferred volcanic flux also changed markedly in 1986. 

Since the results from 1980 to 1984 showed that the principal source of vol- 
canic thermal emission as seen in eclipse observations arose from the Loki 
volcano, we infer that there has been a marked decrease in the flux from Loki. 
These data are summarized in Table III, which gives the model "volcanic" 
parameters for each year. 

In most cases it has been possible to solve for the equivalent circular 
size, temperature, and longitude of both a hot source and a warm source. The 
longitudes are obtainable from the fact that the aspect and the accompanying 
foreshortening factor are different for the observations at disappearance and 
reappearance. In most cases the longitudes are near to that of Loki. The 
Voyager 1 IRIS instrument found that Loki was, perhaps, the most productive of 
thermal output and that if eclipse observations had been made in 1979 Loki 
would have been responsible for 75* of the observed flux. The penultimate 
line in Table III is particularly accurate. The number depends on model fit- 
ting of the flux observed at seven different wavelengths from 2.2 to 20 pun. 

It is notable that the observed volcanic flux dropped markedly in 1986. 
Such a large drop can only have been produced by a marked decrease in the out- 
put from Loki. It is also significant that the longitude of the warm source 
has shifted to 334° W. During the Voyager observations there were several 
important sources, such as Creidne Patera, west of Loki. If there were a 
decrease in the flux from Loki, then a westward shift in the effective longi- 
tude of a single source, as in the model, would be expected. The hot source 
has also decreased in area and in power output. The longitude has not 
shifted. During the Voyager swing-by, the other sources near to Loki that 
would have been visible during an eclipse did not have a high-temperature com- 
ponent. Thus if Loki decreases its output, then no shift of effective longi- 
tude is expected for the hot source. We will see in the discussion of the 
polarization observations made in 1986 (section C3) that they, too, show a 
marked decrease in Loki activity. 

C2. Further Reduction of Occultation Observations of 1985 


Goguen, who is now at JPL, has been collaborating with Sinton in continu- 
ing the reduction of the set of observations of the 1985 occultations of Io by 
other satellites. Considerable progress has been made in the refinement of 
the relative positions of the two satellites (Io and the one that occulted 
it). Precision of location along the relative track of occultation has now 
achieved a precision of about 10 km. Thus the occultation method, in the rare 


25 



Table III 


Volcanic Fluxes, Loki and Environs, 1980-1986 


Parameter 


1980 

1981 

1982 

1983 

1984 

1986 

Size, hot source, 

(km) 

30 

10 

28 

34 

17 

7 

Temp . , " 

(°K) 

564 

522 

[600] 

541 

[600] 

555 

Longitude," " 

(deg W) 

305 

[310] 

293 

304 

[310] 

305 

Size, warm source, 

(km) 

382 

150 

202 

197 

150 

110 

Temp . , " 

( °K) 

250 

368 

297 

292 

360 

345 

Longitude," 

(deg W) 

[310] 

[310] 

312 

316 

[310] 

334 

Power , hot source 

(10 13 W) 

0.40 

0.04 

0.30 

0.44 

0.16 

0.02 

Power , warm 

(10 13 W) 

2.55 

1.83 

1.42 

1.26 

1.69 

0.76 

Total power 

(10 13 W) 

2.95 

1.86 

1.87 

1.70 

1.85 

0.79 

Std. err. of above 

(10 13 W) 

0.27 

0.06 

0.08 

0.06 

0.10 

— 


Note: Bracketed quantities are assumed, the remainder are solved for. 


26 



instances when it can be applied, produces a precision in the location of 
volcanoes that is nearly an order of magnitude greater than the best Voyager 
IRIS resolution on Io. The final results are not yet in, but an example of a 
well-observed occultation is shown in Figure 15. This event (occultation of 
Io by Callisto that occurred on 10 July 1985) was observed at four telescopes 
on Mauna Kea at four wavelengths. A new volcano was found, and its position is 
85° W and 25° S in a region that was poorly observed during the Voyager 
encounter. The volcano has now been located in the 8.7 jim curve, as well as 
in the 3.8 and 4.8 pm curves that were shown in the report last year. A high 
time resolution of the disappearance of the volcano behind Callisto 's limb is 
shown as Figure 16. It is clear that the hottest part of the volcano is about 
10 km in diameter. From the combination of the 3.8 pm data with the 4.8 pm 
data a color temperature of about 700 K and effective diameter of 10 km are 
obtained. 

C3. Observations of the Polarization of Io at 3.8. 4.8. and 10 urn. 

The light emitted by a dielectric (lava, for example) is polarized as a 
consequence of the Fresnel reflection laws. As a consequence the emissivity, 
which is one minus the reflectivity, is different for the viewing plane that 
contains the surface normal from that of the orthogonal plane. Goguen and 
Sinton previously used this fact to locate several volcanic sources on Io 
(1985, Science 230, 65-69). Location of sources is possible because as Io 
rotates on its axis, the plane of polarization will also rotate but in a 
manner that is uniquely related to the latitude and longitude of the source. 
Sinton and Goguen, in collaboration with B. Ellis (University of Texas), 
attempted to extend these observations on the IRTF this past year using the 
Ames Laboratory polarimeter, and, in collaboration with T. Nagata (formerly 
with Kyoto University; now at the IFA) , with the Kyoto polarimeter on the 
UKIRT. Polarization of Io was measured at both times. The observed values, 
however, were much smaller than obtained earlier, and the position angle of 
the plane of polarization did not change with time throughout the night as 
expected. It now appears that the polarization that was observed this year 
was largely not due to emission from volcanoes. Rather, it was due to resid- 
ual polarization of the reflected sunlight. This was observed, even though 
the time of making the observations was selected so that the solar phase angle 
was only about l°-2°. Previously, we had felt that this requirement would 
ensure reduction of polarization of the reflected background light to a negli- 
gible amount. At very small phase angles, the plane of polarization is 
expected to be parallel to the Earth-Io-Sun plane. Figure 16 shows the posi- 
tion angle of this plane and the observed direction of Io polarization at 3.8 
fjun as a function of time. It is clear that the observed direction tracks the 
changing direction of this plane, which swings through nearly 180° as Io 
passes through opposition. 


27 




o 


0*1 pa 2 TTT«®*ON xnLi oisjjtvo pm? oi jo -[*301 


oo c • o 

x; 03 o -M 
«P TJ +J 
M'S CIJ O 
C 3 -M 
03 Q0 ^ CO 

i— ( <— I ”H 

03 +■> CO —I 

> *2 O *2 

CO X3 (0 

* -2 <W U 

x: o 

U X T3 
S 03 & C 
O ■ CO 
<w CO •«-» 

x: o 

•M 4-> HH 

CO bJ) 03 

c x: 

TJ 03 V O 


O 03 H 4J {h 33 

o ^ o) 0) a > 

> *0 s* 

03 C0 03 03 03 

x c h d j: to 

+J 3 X V X 

(C H P o 

0 2(0 S 

•m c co o e 

C bO 5 U 03 

O fi O fe D 

CO ^ jQ 

CO *-i 

03 «-i CO • 4-> 


28 




ORIGINAL PAGE 13 
OF POOR QUALITY 



i i 

xntJ paztteujJON 


O Q< CO *H 

• O OJ ^ » 

<o m g x •v 

H CO D (O *0 

0) ^ c a; 

0) 0) £i *H C 

t-i -G pH •*-» 

3 4-> 0) 0J CO <0 

W >>»+■> 

•h cc 0} 0) -Q 

Ct4 O .fl ^H PH O 


29 




The amount of polarization observed at either 3.8 or 4.8 pun was never 
more than 0.4*. In the 1984 observations the polarization produced by Loki 
was observed in amounts up to 1.6*, and the maximum polarization, which was at 
rotational longitudes of Io that were not observed, would have been about 
2.5*. In the new observations we did acquire data when the polarization due 
to Loki would have reached the maximum value, and yet we never observed more 
than the above mentioned 0.4*. We conclude from these data that the amount of 
near-IR flux from Loki in 1986 was small enough that its contribution to the 
overall polarization of Io was negligible and that the thermal emission from 
Loki was at a low ebb in 1986. This conclusion is identical to that found 
from the eclipse observations discussed in an earlier section of this report. 
In view of the measurement of the residual background polarization, possible 
only because of the ebb of volcanic activity, we can now go back and correct 
our 1984 data for this polarization. We have not done this as yet. It is 
clear, however, that the correction will not change the major results of the 
paper. The correction will probably make invalid our determination of the 
location of a second source on the hemisphere that includes Loki. 

Sinton has collaborated with David Aitken of the University of New South 
Wales, Australia, in interpreting observations that the latter made of the 
polarization of Io at 10 Jim. Sinton had earlier advised Aitken exactly when 
to observe to have the maximum polarization from Io, which Sinton chose as the 
above mentioned rotational longitude for the maximum due to the Loki volcano. 
Aitken was able to make spectropolarimetric observations at that time and 
measured Io polarization between 8 and 12 Jim. The observed broad-band polari- 
zation was about 0.5*. The spectrum of Io can be resolved into two com- 
ponents, one due to background thermal emission at 138° K and one at 336° K. 
The latter is presumably due to volcanic emission. The position angle of the 
polarization is within 2° of the polarization expected from Loki. Note that 
in this spectral region, there will be no contribution of reflected sunlight 
and so no polarization from that source is expected. However, there might be 
a residual polarization of the background thermal flux. If there were such 
polarization, the expected position angle would be 45° from that observed. 

C4. Europa , Ganymede, and Callisto; Photometry at 3.8 and 4.8 ia 

Last year we reported that Ganymede had a most remarkable linear phase 
effect of 0.08 mag/degree at 4.8 pm and 0.05 mag/degree at 3.8 pm. Based on 
one chance, rather bright observation at a phase angle of 5°, we conjectured 
that Europa might have an even greater phase dependence of about 0.4 mag/ 
degree, even though it had been difficult to detect in earlier IRTF observa- 
tions at phase angles near to 10°. To test this hypothesis, observations were 
scheduled with the 2.2-m telescope at a phase angle of 2° since data were com- 
pletely lacking for angles <5°. The observations on two separate nights gave 


30 



4.8-pm Magnitudes of 7.37 ± 0.25 and 7.21 ± 0.13. These data, when combined 
with IRTF data at 10° phase angle, yield an unremarkable linear phase depen- 
dence of 0.03 mag/degree that is not well determined and could be much larger 
but not so large as to have been responsible for the "bright” observation. 

What then was responsible for the earlier bright observation at 5° phase 
angle? We have been unable to attribute the observation to a misidentif ica- 
tion of satellites. The telescope positions were recorded by the software, 
and the observation was among several made of Europa at different wavelengths 
that are consistent. The telescope position was consistent with the position 
of Europa when compared to the indicated positions of the other satellites, 
which were observed at the same time. Moreover, there are no catalogued 
infrared sources in that region with which it could have been confused. The 
observation was made on 23 April 1981 at 9:03 UT and gave a M magnitude of 
6.12, which is about 1.8 mag or a factor of 5 brighter than expected from the 
phase curve that was derived from other observations. At this point we might 
presume that there was an "outburst" from Europa on that night. However, 
there is one other fact that distinguishes the 23 April observation. It was 
the only observation that was made with the IRTF bolometer rather than with 
the InSb detector. It is possible, though it has not been reported, that 
there is a "red" leak in the 4 . 8 -pm filters. The red leak would cause no 
problem when used with the InSb detector, but if the leak occurs beyond 12 pm 
it could cause a misinterpretation of results when used with a bolometer 
detector on cold solar system objects that are dark at 4.8 pm. Thus, at this 
time, we cannot say whether there was an outburst from Europa. 

The earlier phase dependences of Ganymede and Callisto were confirmed in 
the new measurements at small phase angles. An interesting result of the new 
observations is that the "opposition surges" observed in the visible on almost 
all solar system objects are absent in observations of the Galilean satellites 
at 3.8 and 4.8 pm. 

C5 . Charon 


Analysis of the Pluto-Charon speckle data collected in 1985 by R. Goody, 
J. Beletic, and Tholen was completed. Goody and Beletic (Harvard University) 
developed techniques to compensate for instrumental effects (such as phosphor 
persistence) in the PAPA camera, and final processing of the seven nights of 
Pluto-Charon speckle imaging was performed. Altogether 54 pairs of positional 
coordinates were extracted from the data, which makes it the largest, most 
homogeneous, and best calibrated speckle data yet obtained on the Pluto-Charon 
system. Tholen added these new data to the existing data set and performed 
another orbit improvement for Charon. The new data show the orbital radius to 
be slightly larger than previously determined, and the orbital inclination is 
somewhat higher. A paper describing these observations and results in 
currently in preparation. 


31 



C6. Outer Satellites of Jupiter 


Although we had hoped to obtain new data on the outer Jovian satellites, 
Jupiter's opposition fell at a time that coincided with the Heidelberg Halley 
Symposium, the Paris DPS meeting, and the preparations for these meetings. 

C7. Studies of Triton 

In a continuation of the ongoing observational studies of Neptune's sat- 
ellite Triton, Cruikshank, in collaboration with A. T. Tokunaga and R. G. 

Smith ( IFA) and R. H. Brown of JPL, made new observations with the cooled- 
grating array spectrometer (CGAS) on the IRTF in May and August 1986. The 
CGAS was used with a 32-element detector array, which was arranged to cover 
the entire spectrum between 2.0 and 2.5 pm. This is the region of the spec- 
trum that shows the strong 2 . 3-pm methane band and the 2 . 15-pm feature that we 
have attributed to the density-induced absorption in molecular nitrogen. In 
Figure 17 we show the data from these observing runs; see the captions for 
details . 

The main characteristics of the new spectra are evident in the figures. 
The methane band at 2.3 pm is now seen with at least three distinct com- 
ponents. While it is not fully resolved, the band does show features not vis- 
ible in the earlier data taken by Cruikshank and others with circular variable 
filter spectrometers. To study the new spectrum in some detail, we require 
comparison spectra of methane in various states. In Figure 18 we show the 
Triton average spectrum in comparison with laboratory spectra of liquid nitro- 
gen and with methane dissolved in liquid nitrogen. The quantity of methane 
dissolved in each sample in Figure 18 is indicated in grams/liter. With con- 
centrations of methane amounting to less than about 1.5X of the saturation 
value, the transmission spectra show the density- induced molecular nitrogen 
feature with a central wavelength of 2.15 pm, but as the concentration of 
methane is increased, the 2.3-pm methane band increases in strength and width 
so that the nitrogen feature is overrun by the time the concentration reaches 
about 1.5 mole percent. The central wavelength of the feature in liquid 
nitrogen from the spectra of R. N. Clark (USGS) et al . is 2.152 ± 0.002 pm, in 
precise agreement with the new Triton data. 

Several points emerge from an examination of the spectra in Figure 18. 
First, there is an overall agreement in the shape and strength of the methane 
band at 2.3 pm. Three specific features in the methane spectrum, the minima 
at 2.22, 2.31, and 2.36 pm, are found in the Triton data, though not in the 
same relative strengths. Second, the central wavelength of the pure liquid 
nitrogen band coincides exactly with an absorption feature in the Triton 
spectrum at 2.150 pm. In addition, a marginal feature in the Triton spectrum 
appears at 2.05 pm. 


32 


Relative Reflectance 


k J-Hv- 

y "T 


( b ) 0 


Wavelength ( jjl m) 


Figure 17. Spectra of Triton, (a) is the grand average of five nights 

data taken 17-21 May 1986, and (b) is the average of data from 27 August 1986. 
The vertical line shows the position of the absorption attributed to molecular 
nitrogen. 





Wavelength (/xm) 


Figure 18. Spectra of Triton (b) and laboratory samples. (a) is a 
transmission spectrum of a 40-cm column of liquid nitrogen. (c,d,e) are 
transmission spectra of a 39-cm column of liquid nitrogen with various 
amounts of methane dissolved in it. The concentrations of methane in the 
liquid nitrogen for each of the three spectra are (a) 4.6 g/1, (b) 9.3 g/1, 
and (c) 18.6 g/1. For concentrations greater than about 10 g/1, the methane 
absorption obscures the liquid nitrogen band. 


34 



It is also useful to compare the new Triton data with spectra of gaseous 
methane, as was done by Cruikshank and Apt (1984). Because there are no ade- 
quate laboratory spectra for methane gas at the low temperature and pressure 
pertinent to Triton, we are obliged to use synthetic spectra computed with the 
best available band models and absorption coefficients. In Pigure 19 we show 
the Triton data in comparison with a series of synthetic spectra computed by 
Dr. Jerome Apt and kindly made available for this purpose (for details of the 
band models see Cruikshank and Apt, 1984). Not only is the 2.3-pm band in the 
synthetic spectrum much wider for a given central depth than is the band on 
Triton, but the details of the structure are very different. There is no sin- 
gle example in the family of synthetic spectra shown that provides a satisfac- 
tory match to Triton. 

Synthetic spectra of gaseous methane are strongly model dependent, as 
described by Apt et al. (1981). As in the case of the synthetic spectrum used 
by Cruikshank and Apt (1984) in their Pigure 3, the spectra in Figure 19 of 
this paper were calculated with a low-temperature self-broadening coefficient 
of 0.355 cm" 1 atm" 1 . The gas abundances and pressures are those given in the 
figure caption; the temperature is T = 55 K. 

Apt et al. (1981) computed the curve of growth for the absorption band at 
2.3 pm under a variety of assumed model conditions. At values of less than 
about 100 cm" 1 the equivalent width (EW) of the band is not strongly dependent 
upon the model (mean line spacing and broadening coefficient) nor upon the 
temperature. We have measured the equivalent width of the methane band in our 
Triton spectrum between 2.179 and 2.387 pm (corresponding to the interval used 
in the Apt et al. calculations) and find EW = 107 cm" 1 . 

Ideally, we would like to derive the abundance of gaseous methane on 
Triton from a band that is unsaturated and independent of the pressure and 
then use the stronger pressure-dependent 2.3-pm band to derive the surface 
partial pressure of methane by means of the Apt et al. (1981) curve of growth. 
The 0.89-pm methane band is the weakest feature that has thus far been 
detected in the Triton spectrum. Johnson et al. (1981) did not find this band 
at all when they observed at an orbital phase angle of 290°, nor did Combes et 
al. (1981) find it at phase angle 116°. Apt et al. (1983) detected the band 
weakly at phase angle 352° , thus adding to the evidence that the strength of 
the methane bands is dependent upon the viewing aspect, as has been discussed 
by Cruikshank and Apt (1984). The apparent spatial variability of the 0.89-pm 
band, plus the fact that methane ice seen in diffuse reflectance also shows an 
absorption band at the same wavelength, makes it clear that the gas abundance 
cannot be derived from observations of this band in the spectrum of Triton at 
spectral resolutions presently achievable. 

We note that methane ice does absorb at the same wavelengths as the 
0.89-and 2.3-pm gas bands, as can be seen in the reflectance spectrum pub- 
lished by Fink and Sill (1982) and in unpublished but frequently referenced 


35 



ve Reflectance 






spectra by H. H. Kieffer. The strengths of the methane ice bands are strongly 
dependent upon the nature of the reflecting surface, as is the case with water 
ice (Clark, 1981). Very fine-grained frost yields the weakest absorption 
bands, while large crystals or a solid block of transparent or translucent ice 
results in very strong bands. 

It thus appears that nongaseous methane on Triton makes a major contribu- 
tion to the absorption spectrum and that we cannot derive the methane gas col- 
umn abundance on the basis of the existing data. The most straightforward way 
by which to estimate the column abundance and partial surface pressure of 
methane is to use the equilibrium vapor pressure. The problem with this 
method is that the surface temperature of Triton is unknown because we remain 
ignorant of its albedo. In addition, the variation of temperature from the 
subsolar point to the dark hemisphere and to the polar regions, including the 
pole presently in long-term darkness, will have a profound effect on the local 
methane gas abundance because of the extremely strong dependence of vapor 
pressure upon temperature. 

In Figure 20 we compare our Triton spectrum with reflectance spectra of 
methane ice from Fink and Sill (1982) and Kieffer (unpublished); the Fink and 
Sill data were taken at higher resolution than the others, but with a sample 
that gave overall less absorption strength. Several features in the Triton 
spectrum align well with those in the ice spectra, but the Triton band at 2.15 
pm is not matched by any band in the laboratory data for methane. 

While the details of the spectra match, the Triton spectrum has much less 
overall absorption in the 2.3-pm band than does either of the ice samples. It 
is, in fact, exceedingly difficult to prepare a laboratory sample of methane 
ice that will have a weak absorption spectrum, even if the sample is diluted 
with some neutral substance, such as charcoal. An additional laboratory prob- 
lem is the strong tendency of methane frost samples to evolve rapidly to a 
solid sheet or glaze of ice. The absorption spectra of frosts of various 
kinds is strongly dependent upon grain size, as Clark (1981) has demonstrated 
and quantified for laboratory samples of water frost. The infrared bands are 
weakest for exceedingly fine grain sizes (on the order of 10 pm), and as the 
grain size increases the bands gain in intensity. In another paper, Clark et 
al. (1983) showed that the rate of metamorphism of grain size from small 
grains to large and then to a solid ice sheet is strongly dependent upon the 
volatility of the material. This metamorphism occurs because molecules pref- 
erentially evaporate from crystal edges and redeposit on crystal faces, thus 
causing the crystals to grow in size. Methane is enormously more volatile 
than water (higher vapor pressure), and at a given temperature the grain 
growth rate of methane is very much greater than than water (Clark et al . , 
1983). Clark et al . showed, for example, that on Triton or Pluto arbitrarily 
small methane frost grains would grow to 10 mm in 100,000 years at the assumed 
temperature of 50 K and would grow even faster at slightly higher temperatures. 


37 



Wavelength {ft m) 


Figure 20 

The spectrum of Triton (a) compared with two reflectance spectra of methane 
ice/frost. (b) is the spectrum of methane frost from Fink and Sill at 
60K; (c) is a Kieffer spectrum of solid methane (unpublished). 





In this context we note that Triton has a complex and extreme seasonal 
cycle resulting from the rapid precession of its orbital plane around 
Neptune's pole (P = 637 ± 40 years) and Neptune's sidereal period of 165 
years. The subsolar latitude on Triton reaches 52° at some times, with large 
parts of the polar regions in long-term darkness (Harris, 1984; Trafton, 1984). 
Superimposed upon the seasonal cycle is Triton's diurnal cycle of 5.877 days. 
The combined effect of the diurnal and seasonal cycles and the alternating 
dark poles virtually ensures the mass migration of volatiles on Triton. It is 
not entirely unreasonable that the frequent condensation and sublimation of 
methane on a solid surface could produce a continuous supply of very fine 
frost grains consistent with the character of the 2.3-pm absorption band. The 
seasonal effects on the postulated nitrogen are difficult to estimate, since 
we are not yet confident of the physical state in which the nitrogen would be 
found (Lunine and Stevenson, 1985). Nitrogen is even more volatile than 
methane, however, and on a body where the temperature at a given location var- 
ies widely through the temperature range critical to the state of the material 
on a diurnal and seasonal basis, interesting effects might be expected. 

We demonstrated above and in Figure 18 that the strength of the 2.3-pm 
methane band as seen in 1986 is approximated by a very small quantity of meth- 
ane dissolved in liquid nitrogen. The spectral behavior of a liquid is a much 
closer approximation to that of a solid than to that of a gas because while 
some of the molecules are moving freely through the liquid, others are momen- 
tarily caged by their neighbors and are vibrating in the cages. The spectrum 
thus shows some of the characteristics of the molecules vibrating in a solid 
structure as opposed to the behavior in a gas (Castellan, 1983, p. 91). 

There is a distinct qualitative difference in the appearance of the 
2.3-pm band in the laboratory samples of liquid (Fig. 18) and ice (Fig. 19). 

For a given central depth of absorption (at 2.30 pm, for example), the width 
of the main core of the absorption band is much greater in the ice than in the 
liquid. In the limited set of laboratory data we present here, the Triton 
spectrum in the region 2.25-2.45 pm certainly resembles the liquid mixtures 
more than the ice samples. 

Reference to Figure 17 suggests that the strong feature seen in earlier 
low-resolution spectra of Triton at 2.15 pm consists in part of the 2.20-pm 
component of the methane band plus an additional narrower feature centered at 
2.15 pm. 

This feature is not seen in the methane spectra from the laboratory, and 
because its central wavelength is the same as that for the first overtone of 
the density-induced nitrogen spectrum, we feel that the tentative identifica- 
tion by Cruikshank et al. (1984) as nitrogen is strengthened by the new data. 

No alternatives to this identification have been proposed, to our knowledge, 
but this alone does not make the nitrogen identification unique. 

Cruikshank et al. (1984) showed that if there is sufficient nitrogen to 
show this band on Triton, it must occur in a condensed state, and they sug- 

39 



gested that liquid was more likely to produce the observed spectrum than would 
be nitrogen ice. They proposed a model of the Triton surface in which liquid 
nitrogen seas and expanses of solid methane occurred. Lunine and Stevenson 
(1985) concluded from the CH 4 -N 2 phase diagram that a more likely configura- 
tion is for methane and nitrogen both to occur as solids in either a micro- 
scopic mixture or as a disequilibrium assemblage with a nonuniform distribu- 
tion of the two components. The laboratory work on methane-nitrogen mixtures 
(Piscitelli et al . , 1987) demonstrates the difficulty imposed by the very high 
solubility of methane in nitrogen. If liquid nitrogen and solid methane were 
both present on Triton, the nitrogen would probably be saturated or nearly 
saturated with methane. The spectrum is not consistent with any more than a 
minute quantity of methane dissolved in the hypothetical nitrogen sea. 

It remains to be seen if the Triton spectrum is consistent with a solid 
mixture of the type suggested by Lunine and Stevenson (1985); laboratory data 
on solid nitrogen are difficult to obtain, and we know of no near-infrared 
spectra that relate directly to the problem at hand. We can say, however, 
that on the basis of the laboratory spectra in Piscitelli et al. (1987) and 
those given here, that the Triton spectrum in the 2.3-|im band is reasonably 
well matched feature-by-feature and in overall shape and strength by liquid 
mixtures of small quantities of methane dissolved in nitrogen. 

Cruikshank and Apt (1984) noted the variability in strength of the meth- 
ane band in Triton's spectrum throughout the range 0.8-2. 5 pm, and found that 
the evidence was in favor of maximum absorption in the sector approximately 
defined by position angles 180° to 280° , corresponding roughly to the hemi- 
sphere of the satellite facing Neptune. 

In preparing the coadded spectrum for May 1986 given in Figure 17, we 
have ignored the orbital variability of Triton's spectrum, because we could 
not see significant differences in the single-night data sets and because we 
wanted to maximize the signal precision of the entire data set. This proce- 
dure masks any orbital variability, of course, but in examining the individual 
nightly averages, we did not discern any strong tendencies toward variability 
of the methane band at a level above the intrinsic noise in each data set. 

This lack of variability was surprising because we were expecting to find the 
same trends reported in the earlier work. 

We were further surprised by the overall weakness of the 2.3-pm methane 
band in the 1986 data in comparison to the spectra obtained in previous years 
and reported in our earlier papers (Cruikshank and Apt, 1984; Cruikshank et 
al . , 1984). This startling difference is shown in Figure 21, in which the May 
1986 grand average spectrum is compared with that of 2.4 June 1980. Though 
there may be some ambiguity as to the location of the continuum in each 
spectrum, the 1986 data have a maximum absorption depth of about 40% of the 
continuum, while those of 1980 show some 70% absorption (measured at the mini- 


40 



Relative Reflectance 



Wavelength (/im) 


Figure 21 

This figure compares portions of the Triton spectrum from May 1986, and from 
June 1980. The overall strength of the methane-nitrogen complex is much less 
in the 1986 data. 


41 



mum in each trace). The fact that the spectral resolution of the 1986 data is 
about 120, compared with 60 for the 1980 data cannot account for the differ- 
ence seen in Figure 21. The band shape in 1986 is significantly narrower than 
in 1980, and the region around 2. 1-2.2 pm is much weaker, while at higher res- 
olution it should appear deeper. 

Our new spectra are consistent with those obtained in April 1984 by Rieke 
et al. (1985) at similar resolution, but over a narrow wavelength range 
(2.1-2.25 pm). The Rieke data show a distinct but narrow band at 2.15 pm, 
coincident in wavelength and in strength with the band we see, and also an 
absorption at 2.2 pm, which we attribute to methane. 

We cannot confidently attribute the difference in our spectra between 
1980 and 1986 to any instrumental or geometric effect. We do note, however, 
that when we obtained spectra at several position angles in 1985 with the CGAS 
in an earlier stage of development, we found the absorption spectrum to be 
substantially weaker in the region 2.00-2.23 pm than we had anticipated 
(Tokunaga et al., 1985). In the regions covered, the spectra of both Rieke et 
al . (1985) and Tokunaga et al. (1985) should have shown a strong decrease in 

overall intensity toward the longer wavelengths if the spectrum was comparable 
to that in 1980, but such was not the case for either data set. 

We are reluctant to conclude that Triton's spectrum changed markedly in 
the 1980s, but present evidence points in that direction. In a separate paper 
we will consider the complete data set available throughout the time period 
1978-1986 in order to try to find a solution to this apparent enigma. 

The new spectra obtained in 1986 confirm the presence of the strong meth- 
ane band at 2.3 pm, but it appears generally weaker than earlier data at some- 
what lower spectral resolution showed it six years before. Four specific 
features in the spectrum of methane, both as a frost and as a solute in liquid 
nitrogen, are seen in Triton's spectrum in the 2.3-pm region. In addition, 
all the data obtained by us in 1986 show a weak but persistent band at 2.15 pm 
coincident with the central wavelength of the 2-0 density-induced band of 
molecular nitrogen in the liquid state. We have compared the new Triton spec- 
tra to synthetic spectra of methane gas and find that a gaseous atmosphere 
alone cannot account for the relative strengths of the 0.89-pm band (Apt et 
al., 1983) and the band at 2.3 pm. The fact that methane ice shows strong 
absorptions coincident with the gas bands at both these wavelengths indicates 
that a pure gaseous spectrum cannot be observed on Triton. 

If the methane bands on Triton are caused largely by frozen methane, then 
the material must be in the form of an exceedingly fine-grained frost in order 
to produce the very narrow 2.3-pm band we observe. Such fine grains may not 
be consistent with the extremely rapid metamorphism of fine grains to coarse 
grains and then to sheet ice predicted for methane from its known volatility 
(Clark et al . , 1983) at the presumed temperature of Triton. Alternately, the 
methane bands may arise from very small quantities of methane dissolved in 


42 



liquid nitrogen; laboratory spectra by Piscitelli et al. (1987) show that the 
2.3-pm methane band in this state is comparable in shape and strength to the 
band observed on Triton. We are still unable to evaluate from a spectroscopic 
point of view the viability of Lunine and Stevenson's (1985) proposal that 
methane and nitrogen would form a solid mixture of some kind on the surface of 
the satellite. 


References 

Apt, J., N. P. Carleton, and C. D. Mackay (1983). Methane on Triton 
and Pluto: New CCD spectra. Astrophys . J . 270, 342-350. 

Apt, J., J. V. Martonchik, and L. R. Brown (1981). Comparison of band 
model and integrated line-by-line synthetic for methane in the 
2.3-jim region. J. Quant. Spectrosc. Radiat. Transfer 26, 

431-442. 

Castellan, G. W. (1983). Physical Chemistry (3d ed.). Addison-Wesley , 
Reading, MA. 943 pp. 

Clark, R. N. (1981). Water frost and ice: The near-infrared spectral 
reflectance 0.65-2.5 microns. J. Geophys. Res. 86, 3087-3096. 

Clark, R. N. , F. P. Fanale, and A. P. Zent (1983). Frost grain size 
metamorphism: Implications for remote sensing of planetary 
surfaces. Icarus 56 , 233-245. 

Combes, M. , T. Encrenaz, and J. Lecacheux (1981). Upper limit of the 
gaseous CH 4 abundance on Triton. Icarus 47, 139-141. 

Cruikshank, D. P. , and J. Apt (1984). Methane on Triton: Physical 
state and distribution. Icarus 58 , 306-311. 

Cruikshank, D. P. , R. H. Brown, and R. N. Clark (1984). Nitrogen on 
Triton. Icarus 58 , 293-305. 

Fink, U., and G. T. Sill (1982). The infrared spectral properties of 
frozen volatiles. In Comets (L. Wilkening, Ed.), pp. 164-202. 
Univ. of Arizona Press, Tucson. 

Harris, A. W. (1984). Physical properties of Neptune and Triton 
inferred from the orbit of Triton. In Uranus and Neptune 
(J. Bergstralh, Ed.), pp. 357-373. NASA Conference Publ. 2330. 

Johnson, J. R. , U. Fink, B. A. Smith, and H. J. Reitsema (1981). 
Spectrophotometry and upper limit of gaseous CH 4 for Triton. 

Icarus 46 . 288-291. 

Lunine, J. E. , and D. J. Stevenson (1985). Physical state of volatiles 
on the surface of Triton. Nature 317 , 238-240. 

Piscitelli, J. R. , D. P. Cruikshank, and J. F. Bell (1987). Laboratory 
studies of irradiated nitrogen-methane mixtures: Applications to 

Triton. Icarus (submitted). 


43 



Rieke, G. H. , L. A. Lebofsky, and M. J. Lebofsky (1985). A search 
for nitrogen on Triton. Icarus 64 , 153-155. 

Tokunaga, A. T., R. G. Smith, R. H. Brown, D. P. Cruikshank, and R. N. 

Clark (1985). The infrared spectrum of Triton at 2.0-2.25 pm. 

Bull. Am. Astron. Soc. 17, 698 (abstract). 

Trafton, L. (1984). Large seasonal variations in Triton's atmosphere. 
Icarus 58, 312-324. 

C8. Laboratory Studies Related to Triton 

In our last report, we described laboratory observations of methane- 
nitrogen mixtures made for comparison with the spectrum of Triton. The 
results of that work have been written up in a paper by Piscitelli, Cruik- 
shank, and Bell (Hawaii Institute of Geophysics, Planetary Geosciences Divi- 
sion) that has been submitted to Icarus. The Triton studies noted in the 
lengthy section immediately above have drawn considerably on the work by 
Piscitelli et al., and we do not describe the results further here. 

C9. Spectroscopy of Io 

In a collaborative project with other investigators, notably R. Howell 
(Wyoming), T. Geballe (UKIRT), and D. Nash (JPL), Cruikshank is completing the 
analysis of a set of Io spectra in the region 3. 5-4. 2 fim. The new spectra 
being studied have a resolution of about 500 and cover a range of Io longi- 
tudes and times from 1983 through 1985. Several strong bands of SO 2 are 
located in this wavelength region, and this substance is known to be a major 
component of Io's surface; these same investigators have been working on prob- 
lems of SO 2 on Io for a number of years. In the present work we have identi- 
fied several new features attributed to SO 2 , and have derived strengths of the 
33S, 34S, and 180 isotopic bands; the isotopic ratios seem to be normal. The 
data set has also been used to place new limits on the SO 2 gas abundance in 
Io's (transient) atmosphere. The upper limit, subject to various caveats, is 
0.1 cm-amagat, which would result in a very low surface pressure. 

The new data set was compared in detail with laboratory specimens of sul- 
fites and sulfates, both compounds that would be expected on Io, but neither 
class of compounds can be identified with any certainty on Io. It is estab- 
lished, though not very stringently, that the upper limits to the amount of 
surface area covered by these compounds could be as high as 30*. Clearly, the 
spectral evidence in this region covered by our data is not very restrictive 
to the quantities of these compounds permitted below the detection level . 


44 


CIO. Volatiles on Europa 


Cruikshank, in collaboration with Brown (JPL) and Tokunaga, has con- 
tinued the spectrophotometric study of icy satellites; the data on Triton are 
reported in section C7. We have also looked at Europa in the 2.3-pm region in 
search of volatile signatures other than that of water ice. New data on the 
leading and trailing hemispheres of Europa were obtained with the CGAS. 

Spectra of the trailing side (CM 300°) obtained in 1985 show two weak absorp- 
tions near 2.2 and 2.3 pm. Both of these features, as well as others, as seen 
in spectra obtained by R. N. Clark (USGS) at a similar central meridian longi- 
tude. However, with improved equipment and high-quality data, we cannot find 
these absorptions in the Europa data taken in 1986. While we are skeptical of 
any explanation that requires a major change in the spectral response charac- 
teristics of the satellite between 1985 and 1986, there are no detectable 
problems with the data sets that would suggest that the observed difference in 
the spectra result from instrumental or other systematic effects. 

Additional data will be obtained in 1987 to pursue the study of Europa in 
the context of the other icy satellites being observed in this general program. 

Cll. Oranus Ring Orrnltation Observations 


Buie and Tholen are providing observations of stellar occultations by 
Uranus seen from Hawaii as part of a collaborative effort with R. French at 
MIT. Very high signal-to-noise data were obtained from an occultation 
observed on 26 April 1986. All of the major rings were seen along with a 
grazing atmospheric occultation. Figures 22a and 22b show an overview of the 
ring events observed on Mauna Kea together with the model fits to the data. 

In all cases the quality of the fit is remarkable and stresses the importance 
of continued Earth-based monitoring of the Uranian ring system. 

Most of the modeling and analyses are being done by French. The new data 
have been added to the growing data base that constrains the orbit solutions 
for the Uranian ring system. The latest work by French has identified a number 
of resonances that are responsible for shaping and controlling the ring 
system. This pushes the accuracy of the model to ±1 km in the semi-major 
axes of the rings and ±0.01° in the direction of Uranus's pole, an accomplish- 
ment that exceeds the capabilities of Voyager alone. In Figures 23a and 23b 
we show a comparison of the stellar occultation of the e and y rings observed 
at Mauna Kea in comparison to the Voyager X-band data. 


45 





> 

cd 

CO 


u 

U 

00 


a> 

rH 

c 


0) 


•H 

CO 


u 


U 

o 

o 


•H 


o 

<u 

4-1 

cd 


-C 



H 

4J 

rH 

& 

cd 


0) 


rH 

T3 

cd 

T i 

JO 

O 

e 

a) 


6 

d 

H 

sO 

cd 

CO 

00 

• 

S 



■U 


C 

iH 

•H CO 


o 


5 rH 

• 


i— 1 

«H 

cd 

cd 

•H 

«s Q) 

CNI 

u 

U 

CO £ 

CM 

cd 

cu 

3 


na 

< 

a a) 

a> 



cd n 

J-i 


M Cd 

0 

h 

CM 


00 

o 


cr 

•H 

4J 4-1 

4-1 CO 


cd 

O 

O 



46 


Seconds after 1 3:30:41 .4032 UT Seconds after 1 3:29:45.4 UT 








in o in o in 


Y“ o o 

Ajisusiui paziiBiUJON 




<o 





Figure 22b. Mauna Kea Observatory data on stellar occultations of 26 April 

1986 by the rings of Uranus, with model fits (square wells). 


47 









Normalized Intensity Normalized Intensity 


U28 MKO Epsilon Ring Egress 



Voyager RSS X-Band Epsilon Ring Egress 



Figure 23a 

Stellar occultation of the epsilon ring of Uranus (Mauna Kea) 
in comparison with the Voyager 2 X-band occultation curve. 


48 




Normalized Intensity Normalized Intensity 


Voyager RSS X-Band Gamma Ring Egress 



Seconds after 1 :30:09.5922 UT 


26 April 1986 I RTF Gamma Ring Egress 



Figure 23b. 

Stellar occultation of the gamma ring of Uranus (Mauna Kea) 
in comparison with the Voyager 2 X-band occultation curve. 

49 




D. ASTEROIDS 


Dl. Lightcurves of Trojan and Hilda Asteroids 

Tholen and Hartmann resurrected a project to determine the rotational and 
shape characteristics of Trojan and Hilda asteroids. Because these asteroid 
populations are isolated from the main belt, one might expect them to have had 
a different collisional history. The goal of this project is to provide 
information on the collision evolution of these populations. Because the 
Trojan and Hilda asteroids are among the most distant asteroids, they are also 
among the faintest, thus requiring the 2.2-m telescope for these observations. 

In 1986, this project was advanced on two fronts. First, data acquired 
in May 1984 by J. Goguen (formerly at IFA, now at JPL) Cruikshank, and W. K. 
Hartmann (PSI, Tucson) were reduced. Second, Tholen and Hartmann acquired new 
observations of asteroids (659) Nestor, (884) Priamus, (1172) Aeneas, (1345) 
Potomac, (1404) Ajax, (1868) Thersites, (2483) Guinevere, (3451) 1984 HA1 , 
(3548) 1973 SO, and 1985 TQ, primarily in January, July, and December. Early 
results would seem to indicate that the average amplitude of a Trojan or Hilda 
asteroid lightcurve is higher than their main-belt counterparts, suggesting 
more elongated shapes, and that the average rotational period is longer. Some 
biases do exist in the data, however, and more statistics will be required to 
confirm these early conclusions. Two manuscripts describing the current state 
of the project are presently nearing completion. 

Sample lightcurves are shown in Figures 24-26. The huge amplitude of the 
lightcurve for (2483) Guinevere was a surprise (Fig. 24). The asteroid is 
apparently as least as elongated as (624) Hektor. The rotational period is 
also longer than is typical of main-belt asteroids. The other two objects, 
(1172) Aeneas (Fig. 25) and (1345) Potomac (Fig. 26), also have longer periods 
and larger amplitudes than are typical of main-belt asteroids. 

D2. Multicolor Photometry of Asteroids 

Tholen continued efforts to obtain five-color photometry of outer-belt 
asteroids, planet-crossing asteroids, members of interesting dynamical fami- 
lies, and other objects whose eight-color classifications were ambiguous or 
unusual. The observations of the planet-crossing asteroids will be discussed 
in a section D3. 

Color observations of Hilda asteroid (2483) Guinevere yielded a D-type 
reflectance spectrum, which is no surprise for that part of the asteroid belt, 
but during the course of the color observations, Guinevere brightened by 0.6 
mag! As a result, the asteroid became one of the principal targets for light- 
curve photometry during the remainder of the observing run (see section Dl). 

Asteroids that were observed include (260) Huberts, (401) Ottilia, (1404) 
Ajax, (1565) Lemaitre, (1657) Roemera, (1868) Thersites, (2741) Valdivia, 

(3106) 1981 EE, and (3451) 1984 HA1 . 


50 





51 


Rotational phase 






53 



Tholen has continued to monitor the brightness of (2060) Chiron with the 
2.2-m telescope to look for an outburst of the P/Schwassmann-Wachmann 1 
variety. Such an outburst would explain the discrepant visual and infrared 
photometry obtained in previous years. Tantalizing data on this 19th mag 
object were obtained in late December 1986. On one night, Chiron was approxi- 
mately 0.5 mag brighter than predicted, but the proximity of a field star has 
rendered the data questionable. The following night, Chiron was about 0.3 mag 
brighter than predicted, which is about a 3-a fluctuation from the mean of all 
previous observations. It seems plausible that the tail end of an outburst 
was observed, but the data are not quite sufficient to confirm such an event. 

D3 . Planet-crossing Asteroids 

Eleven planet-crossing asteroids were successfully observed at visual 
wavelengths by Tholen in 1986, all but one with the 2.2-m telescope. The 
objects include (1566) Icarus, (2074) Shoemaker, (3103) 1982 BB, (3199) 
Nefertiti, (3361) 1982 HR, (3551) 1983 RD, (3554) 1986 EB, 1984 AB, 1985 PA, 
1986 DA, and 1986 JK. Infrared CVF observations of (3551) 1983 RD were also 
obtained by Cruikshank, Tholen, and Hartmann with the IRTF. 

In 1986, (3551) 1983 RD made its first favorable apparition since discov- 
ery. During the discovery apparition three years ago, Tholen 's five-color 
photometry yielded a reflectance spectrum that was identical to that of (4) 
Vesta, which at that time was the only asteroid known to have a reflectance 
spectrum indicative of essentially olivine-free pyroxene. New five-color 
photometry obtained in 1986 has confirmed the 1983 findings. Vesta has long 
been considered the eucrite meteorite parent body, but the delivery of eucri- 
tes from Vesta to Earth has remained a problem, one that might be solved if an 
Earth-approaching asteroid was found to be similar to Vesta. The discovery of 
1983 RD's Vesta-like spectrum likely provided the missing link. The five- 
color photometry obtained in 1983 extended only to 9000 A, however, and the 
telltale pyroxene and olivine absorption features extend out to 2 jjun. To con- 
firm the identification of 1983 RD as being Vesta-like, infrared observations 
were required. In 1986, 1983 RD became barely bright enough to observe with 
the CVF at the IRTF. Preliminary analysis of the new data clearly shows a 
strong pyroxene absorption feature, but a more careful reduction needs to be 
performed to make a more detailed comparison with the well-established infra- 
red spectrum of Vesta. 

Another exciting discovery was the A- type reflectance spectrum of (2074) 
Shoemaker. A-type asteroids are typified by a very strong olivine absorption 
feature in the near-infrared. Shoemaker is only the sixth known asteroid to 
have an A-type spectrum. Also, it lies near the V5 resonance with no known 
near neighbors. This object's orbit has probably evolved from the main belt. 
The widely dispersed A-type asteroids have unfortunately made it difficult to 
specify the likely source region for Shoemaker. 


54 


(3103) 1982 BB was found to have a very large amplitude lightcurve, mak- 
ing it one of the more elongated Earth-approaching asteroids. Additional 
observations are planned for early 1987 on the outbound leg of this object's 
orbit. Eventually, these data will be combined with similar data obtained by 
W. Wisniewski (Arizona) to provide more complete coverage of the lightcurve 
and a more precise determination of the rotational period. 

Photometry of (1866) Sisyphus obtained in December 1985 was reduced in 
1986. The lightcurve is of small amplitude (0.1 mag) and very short period 
(2.4 hr), making it one of the fastest rotating bodies in the solar system. 

The surprise is the incredible amount of structure in the lightcurve. Figure 
27 shows the data acquired on 28 December 1985 UT with the 2.2-m telescope. 

The spike was seen twice, 2.4 hr apart, and is therefore confirmed. Whether 
the spike is due to a large protrusion coming into view on the limb of the 
asteroid or due to a glint from a large relatively flat area is not known 
yet. It is important to note that without the light gathering power of the 
2.2-m telescope, the photometric resolution achieved would have not been 
possible; that is, the spike would have gone unnoticed, as well as much of the 
structure. This information may be crucial to the proper interpretation of 
some curious radar echoes obtained by S. Ostro (JPL). 

D4. Near-Infrared Spectrophotometry of Asteroids 

Cruikshank and Tholen, in collaboration with Hartmann continued their 
work on the near-infrared spectrophotometry of selected asteroids, including 
A types (olivine-rich) and their relationship to the S-type asteroids. 
Simultaneously, they obtained additional data on the spectrophotometric sig- 
natures from 0.8- to 2.5-pm of various P, C, and D types. An observing run in 
May 1986 was lost to clouds, but a run on the IRTF in October 1986 was very 
successful. During that run, excellent data were obtained for the following 
objects in the A-S and P, C, D projects: 1983-RD, 443 Photographica, 194 
Prokne, 119 Althea, 4 Vesta, 352 Gisela, 27 Euterpe, 9 Metis, 17 Thetis, 12 
Victoria, 2074 Shoemaker, 67 Asia, 1129 Atami, 14 Irene, and 87 Sylvia. 

We are now working on the reduction of the data and will not be obtaining 
new observations until this project can be completed. We believe that we now 
have enough VJHK photometry and 0.8- to 2.5-pm spectrophotometry to complete 
our project on the A-to-S transition and on the basic characteristics of the 
C, P, and D types. 

D5. Organic Molecules on Asteroid 130 Elektra 

The carbonaceous chondritic meteorites are known to contain organic 
matter. Aliphatic and aromatic polymers constitute about 6 % of the mass of 


55 



PericA, ~ ci.40 







the volatile-rich Cl chondrites, which are among the most primitive samples 
of matter in the solar system. 

While the parent bodies of the carbonaceous chondritic meteorites are 
presumed to exist among the outer asteroids, the evidence is based on their 
mutual low albedos and, in some cases, reddish colors. L. A. Lebofsky et al . 
showed in 1981 that some, but not all, low-albedo C-type asteroids have a 
broad absorption in their spectra at 3 pm that is attributable to bound water 
in the mineral lattices. This finding is consistent with the bound water in 
the Cl and CM carbonaceous chondrites. The strength of the ultraviolet 
absorption, attributed to charge-transfer transitions in iron and titanium, is 
related to the amount of water represented by the 3-pm absorption band, as has 
been shown by M. A. Feierberg et al. in 1985. 

In the most primitive carbonaceous chondrites, such as Murchison (CM), a 
complex of very shallow bands at 3.4 pm is superimposed upon the 3-pm water 
band seen in diffuse reflectance. These features are caused by the C-H 
stretching mode and are common to all organic matter, though the exact posi- 
tions and band shapes vary from compound to compound. Other carbonaceous 
meteorites, notably the less primitive samples, such as Allende (CV), do not 
show the C-H band clearly, nor do they have significant bound water evident in 
their reflectance spectra. 

Cruikshank, in collaboration with A. Tokunaga (IFA) and R. H. Brown (JPL) 
used the cooled-grating array spectrometer (CGAS) at the IRTF in August 1986 
to obtain spectra of the "wet" C-type asteroid 130 Elektra in search of the 
3.4-pm C-H bands. Figure 28 shows the asteroid spectrum in the region of the 
3.4-pm band. In the upper panel the spectrum is shown as a simple ratio to a 
solar-type star. For comparison, the laboratory reflectance spectrum of the 
insoluble organic extract from the Murchison CM meteorite is shown. The com- 
plex of bands at 3.4 pm in the meteorite sample is due to organics in the 
extract. The lower panel of the figure shows the same two spectra but with 
the continua removed computationally. Such removal of the continua aids in 
assessing the similarity and coincidence of the absorption features in the 
spectra. The natural continuum slope in the asteroid spectrum results from a 
combination of the water absorption band, the thermal emission of the asteroid 
surface, and the slight differences in the color of the Sun and that of the 
reference star in this spectral region. In the case of the laboratory 
spectrum of Murchison extract, the continuum slope results from the strong 
bound water absorption band and the details of the comparison beam of the 
spectrometer. 

We suggest that the main features of the C-H band in Murchison are 
repeated in the spectrum of 130 Elektra. The band is very weak; in concen- 
trated organic material from the meteorite it amounts to only about 5% absorp- 
tion depth. Though the band strength depends on particle size and packing, we 
would not expect it to be as strong on the asteroid as in the laboratory 


57 



Wavelength (yxm) 


Figure 28 

Reflectance spectra of asteroid 130 Elektra and an organic extract from the 
Murchison carbonaceous chondrite. (a) is the spectrum of Murchison with two- 
times vertical exaggeration to emphasize the weak features. (b) is the asteroid 
spectrum (no exaggeration) with formal error bars shown. (c) is Murchison 
with the continuum slope removed, and (d) is the Elektra spectrum similarly 
flattened (and with two-times vertical exaggeration). (e) is the standard star 
spectrum corrected for extinction. The portions of the spectra in the 
hatched zone are unreliable because of telluric extinction which cannot be 
fully corrected. 





concentrate from the meteorite. The maximum absorption depth in the asteroid 
spectrum is about 4%. 

The 3.4-pm band is diagnostic of the presence of hydrocarbons, but from 
this band alone it is not possible to determine just what molecules are 
involved. Murchison contains a vast array of hydrocarbons, any and all of 
which exhibit the C-H band we see on the asteroid. Also, the presence of the 
C-H band on an asteroid does not prove that all of the primitive carbonaceous 
chondrites originate from this or any other asteroid; comets are known to have 
this spectral feature, though it is normally seen in emission, as in the case 
of ground-based and spacecraft studies of Comet Halley in 1986. 

We shall continue our studies of the 3.4-fim spectral region on asteroids, 
comets, and planetary satellites (particularly Iapetus), in an effort to elu- 
cidate further the connections among these objects and the meteorites in ter- 
restrial collections. 


D6. Modeling of Asteroid Rotational Lightcurves 

Goguen and Research Assistant K. Uchida continued work that they had 
begun in the previous grant year on the modeling of asteroid rotational light- 
curves using Goguen' s photometric model for the surfaces of airless planetary 
bodies. They elected to model Trajan asteroid (G24) Hektor- because of its 
extreme lightcurve. Five small phase angle lightcurves were fit simultane- 
ously with a general triaxial ellipsoid model. The input data span a period 
of 11 years and sample both small and large amplitudes. The model rotates 
counterclockwise about its shortest axis, and its surface has a lunarlike pho- 
tometric function with normal reflectance of 0.022. The rotation period is 
0.288335 days and the rotational pole direction is at ecliptic longitude of 
314° and ecliptic latitude of +17®, in reasonable agreement with previous 
estimates. The semi-axes of the ellipsoid are A = 206 km, B = 93 km, and C = 
78 km. The axial ratios of the model ellipsoid are close to those expected 
for Jacobi ellipsoids, a subset of triaxial ellipsoids that satisfies the 
conditions of hydrostatic equilibrium in a rapidly spinning body of uniform 
density. If (624) Hektor is assumed to be in hydrostatic equilibrium, the 
density implied by B/A, C/A, and the rotation period noted above is 1.0 g/cm 3 . 
This model is stable with respect to binary fission, contrary to the conclu- 
sion reached by Weidenschilling in 1980 based on the maximum measured light- 
curve amplitude alone. 

Figure 29 shows the model fit (solid curve) to the five lightcurves 
studied. Note, also, that the phase is correctly fit, using the rotation 
period given above. 

D7 . Astrometry 


Tholen continued a low-level effort to acquire astrometric positions of 
asteroids using the shaft encoders of both the IRTF and the 2.2-m telescope. 


59 




Figure 29 

Observed lightcurves (points) and model for asteroid 624 Hektor. Both 
the lightcurve amplitude and phase are properly fit with the model, as 
described in the text. 


CL 


60 


Time (hrs) 





Targets basically included those objects that were in need of astrometric 
observations for purposes of orbit improvement, such as newly discovered 
Earth-approaching asteroids, asteroids whose positions deviate significantly 
from ephemerides, etc. In 1986, astrometry was obtained on (659) Nestor, 

(3124) Kansas, (3548) 1973 SO, (3551) 1983 RD, (3554) 1986 EB, 1986 DA, and 
1986 JK. All observations were published in the Minor Planet Circulars, and 
in the cases of 1973 SO, 1983 RD, and 1986 EB, they helped provide the definitive 
orbit solution that allowed them to be added to the catalog of numbered 
asteroids . 


61 



B. CONETS 


El. Photo»etry of Comet Halley 

Tholen, N. Lark (Visiting Colleague), Hammel, and Piscitelli used the #1 
0.61-m telescope nearly continuously from January to mid-June, except for the 
7-week period around perihelion. An additional 43 nights of photometry were 
obtained, making a total of 65 nights of data on Halley collected between 
October 1985 and June 1986. In particular, nine consecutive nights of data 
were obtained from March 5 through 13, which covers the time of the five 
spacecraft encounters. These data were merged with similar data collected at 
Cerro Tololo by R. Millis and D. Schleicher (Lowell) to provide a better 
description of the activity of Halley during the spacecraft encounters. At 
that time, Halley was particularly active, showing changes in gas and dust 
production of more than a factor of 2 from night to night. Figure 30 shows 
the ultraviolet continuum magnitude (open circles) versus time during the 
spacecraft encounters along with ultraviolet minus blue color (filled 
circles). The color of the comet was obviously quite stable even though the 
dust production rate was varying by a tremendous amount. Reduction of the 
remainder of the extensive data set is an ongoing process. 

E2. Monitoring of Comet Halley 

Monitoring of Comet Halley continued throughout 1986. Observations were 
made in January, March, April, and June with the 2.2-m telescope, as well as 
with the 0.61-m telescope and the 0.3-m Schmidt camera throughout the early 
part of the year. Observations were made in support of the spacecraft encoun- 
ters with P/Halley. 

The 2.2-m telescope observations were made with the Galileo/IFA CCD at 
the Cassegrain focus. Images of the near-nuclear region were made through the 
International Halley Watch (IHW) standard filters as well as with special 
filters, which isolate emissions of the OH (0-0) and (1-1) bands (around 3100 
A), in an attempt to map the velocity field of the gas in the coma. Prelim- 
inary reductions of this data were reported at the October ESLAB symposium in 
Heidelberg. Final analysis of the OH data will be published before the end of 
this year. 

A highlight of this effort is a continuous set of images in the IHW red 
continuum band taken throughout the spacecraft encounters with the comet in 
early March. Preliminary analyses of these images were uploaded to the IHW's 
computer for possible use in last-minute targeting of spacecraft to avoid 
dust jets. Further analysis will be useful in determining the state of the 
coma during the encounters and possibly in determination of the nuclear rota- 
tion rate. A list of all images submitted to the IHW near nucleus studies 
network for archiving is given in Tables IV and V. 


62 




Table IV 


Summary of CCD Images of Comet Halley, Mauna Kea Observatory, 1985-1986 



Date 

(UT) 

Time 

(UT) 

Field 

(arcsec) 

Filter 3 

22 

Oct 85 

09:23-10:07 

69 

open, IHW set 

22 

Oct 

13:31-14:12 

69 

IHW set 

23 

Oct 

10:17-12:03 

69 

open, IHW set 

23 

Oct 

12:39-12:47 

69 

CN 

24 

Oct 

11:15-12:39 

69 

IHW set 

18 

Nov 

09:08-10:11 

69 

open, IHW set, red cont. mosaic 

21 

Nov 

06:32-08:34 

69 

IHW set, H20 + mosaic, H20 + offset 

21 

Nov 

11:13-11:58 

69 

H20 + , H20 + mosaic, H20 + offset 

17 

Dec 

05:52-06:16 

69 

open, red cont. polarimetry 

20 

Dec 

07:42-08:18 

69 

IHW set , open 

04 

Jan 86 

06:03-06:25 

69 

IHW set, red cont. mosaic 

05 

Jan 

04:50-06:27 

69 

IHW set, open, red cont. mosaic 

06 

Jan 

04:57-06:10 

69 

IHW set, open, red cont. mosaic, 
red cont. offset, C2 offset and mosaic 

07 

Jan 

04:42-06:05 

69 

open, IHW set, red cont. mosaic and 
offset, C2 mosaic & offset, H2O mosaic 

08 

Jan 

05:18-06:07 

69 

open, IHW set, red cont. mosaic, red 
cont. offset 

04 

Mar 

15:29-15:35 

135 

red cont., red cont. mosaic 

05 

Mar 

15:32-15:36 

135 

red cont., red cont. mosaic 

06 

Mar 

15:13-15:41 

345 

IHW set, red cont. mosaic 

07 

Mar 

15:26-15:34 

135 

red cont. , red cont. mosaic, red cont. 
with 3° and 5° cont. 

08 

Mar 

15:14-15:40 

345 

IHW set, red cont. mosaic, red cont. 
with 3° and 5° rotation 

09 

Mar 

15:04-15:27 

135 

red cont., red cont. mosaic, red cont. 
with 5° and -5° rotation 

10 

Mar 

15:20-15:24 

360 

red cont., red cont. with rotation 

11 

Mar 

15:49-15:53 

360 

red cont., red cont. with rotation 

13 

Mar 

15:58-15:59 

210 

red cont. 

14 

Mar 

14:55-15:13 

135 

IHW set 

18 

Apr 

09:26-09:29 

135 

red cont. 

03 

Jun 

07:50-08:35 

69 

IHW set 

04 

Jun 

07:07-08:09 

69 

IHW set 


a IHW set" is the CN, C3, C0 + , mid-continuum, C2. red continuum, and H20 + 
filters of the IAU/IHW comet filter set. Images taken through the OHO, 0H1 , 
and UV continuum filters are listed in Table V. "Mosaic" means a set of four 
images with the comet nucleus located in the four corners. The "offset" 
images are centered 30, 60, 90, or 120 arcsec from the comet. 


64 




Table V 




Cometary 

OH Data 


Object Date (UT) 

Time (UT) 

Filter 

Comments 

G-Z 

11 Sept 

13:46-15:16 

OHO ,0H1 ,UVC 

6 images 

Halley 

22 Oct 

11:38-12:51 

OHO ,0H1 , UVC 

4 images 

G-Z 

22 Oct 

14:37-15:25 

OHO.OHl. UVC 

3 images 

Hartley-G 

23 Oct 

6:04-7:07 

OHO, OH 1, UVC 

3 images 

G-Z 

23 Oct 

13:31-15:08 

OHO ,0H1 ,UVC 

3 good images 

Hartley-G 

24 Oct 

5:33-7:10 

OHO , 0H1 , UVC 

4 images, cirrus 

Halley 

24 Oct 

12:53-15:21 

0H0.0H1.UVC 

7 images 

Thiele 

18 Nov 

6:09-8:07 

0H0.0H1 ,UVC 

3 good images 

Halley 

18 Nov 

11:17-14:49 

OHO , OH1 , UVC 

3 good images 

Halley 

21 Nov 

8:55-10:08 

OHO , 0H1 , UVC 

3 images 

Halley 

19 Dec 

7:37-8:18 

OHO , OH1 , UVC 

4 images 

Ciffreo 

19 Dec 

9:14-10:12 

0H0.0H1 ,UVC 

6 images 

Halley 

20 Dec 

5:04-7:28 

OHO , 0H1 , UVC 

3 good images 

Halley 

4 Jan 

4:39-5:56 

0H0.0H1.UVC 

3 good images 

Halley 

5 Jan 

5:29-5:56 

OHO.OHl ,UVC 

3 images 

Halley 

6 Jan 

4:57-5:21 

OHO, OH 1, UVC 

3 images 

Halley 

7 Jan 

4:45-5:10 

OHO , OH1 , UVC 

3 images 

Hailey 

8 Jan 

5:08-5:34 

OHO, OH 1, UVC 

3 images 

Halley 

4 Mar 

15:47-16:20 

OHO.OHl 

3 good images 

Halley 

5 Mar 

15:38-16:10 

OHO, OH 1 

3 , bad focus 

Halley 

7 Mar 

15:39-16:18 

OHO.OHl 

2 good images 

Halley 

9 Mar 

15:31-16:03 

OHO.OHl 

2 good images 

Halley 

18 Apr 

9:29-9:50 

OHO.OHl ,RC 

— 

Halley 

3 June 

6 : 40-7 : 34 

OHO.OHl , RC 

very weak 


Note: Images were obtained through the IHW filters on most nights. 
G-Z = Comet Giacobini-Zinner, Hartley-G = Comet Hartley-Good. 


65 




The 0.3-m Schmidt camera was used continuously during the dark phases of 
the Moon, to monitor the development of P/Halley's tail. The comet was 
observed on 31 nights during the first four months of the year, with a total 
of 72 plates being obtained. These were mostly blue-sensitive Illa-J plates, 
which showed the gas tail, but several 3a-F plates were taken with a red fil- 
ter to monitor the development of the dust tail. These have been loaned to the 
IHW large-scale studies network for archiving and analysis. A list of the 
plates submitted is given in Table VI. Figure 31 is a representative picture 
of Halley obtained with the Schmidt on 20 March 1986. 

E3. Coordinated IR and Visible Mapping of Comets 

Hammel, Cruikshank, and Storrs collaborated in 1985 and 1986 with C. 
Telesco (MSFC), H. Campins (PSI, Tucson), and others to produce simultaneous 
visible and infrared maps of comets P/Giacobini-Zinner and P/Halley. The IR 
mapping was done with a 20-element bolometric array; the Galileo/IFA CCD was 
used on the 2.2-m telescope to obtain the visible images. 

From the simultaneous visible and IR images, albedo maps were calculated. 
The morphology was similar for both comets; a region of darker albedo was 
found in the anti-sunward direction, following the direction of the tail curv- 
ature. The albedo distribution is thought to be an effect of particle size 
distribution. Details are provided in the attached paper by Hammel et al. 
(1986) . 

E4. High-Dispersion Spectroscopy of Comet Halley 

Buie and Cruikshank collaborated with T. C. Owen (Stony Brook) and B. 

Lutz (Lowell) to obtain high-dispersion spectra of the CN violet band using 
the coude spectrograph (camera number 3) at the 2.2-m telescope. The CCD cam- 
era was used for this work, in January and April 1986. 

Buie concentrated on the reductions of the spectra, and preliminary 
results were presented at the 20th ESLAB Symposium in Heidelberg. The spec- 
tra contain information about the 12 q14jj violet band system in Comet P/Halley. 
While the resolution of the data was not enough to resolve clearly any lines 
due to the less abundant 1 3 C* 4 N isotope, synthetic spectra computed by D. 
Schleicher (Lowell) enabled us to identify several 12^14^ lines that have 
excess emission in the wings of the strong lines. We have interpreted this 
excess as evidence for the presence of the 1 3 C 44 N isotope in measurable quan- 
tities. Work is continuing to determine an isotopic ratios from these data. 


66 


COMET HALLEY PLATE CATALOG Table VI 


V 


ORIGINAL PAGE IS 
OF POOR QUALITY 


W 

in 

0 

o 


• ■- 
■*« 3 
9 U 


ossssssss 3 * s : s e 


i > t I : t : : 0 3 3 3 :!^:: 


U • — 
0-^9: 
« 3 t. 

uino 


3 : : : 9 3 


O 3! 

4> 

tn 


: 3 3 : s i 


• U 

*4 o s 
D -V 


3 : : ; : t : : - 


Ul 

I- 

-I 

£ 


c c c c c c c 
o © o e o o o 
c c c c c c c 


c c c c c 

• • • • * 

4* 4* *» 4» 


c c c c 


c c c c c 


4* 4* 4*4* 4* 4» 4» V 

• »t»»««4»*»»t«»4*«ff««««*»«««4l«*»4l4*«V«»»«»»«43»««ai4>4*»»4>4*»«»*»4*»»4> 

cccccccnccccccmcccccccccccmmccmmcmccccccnmccccnmccm+cccuQmccm 

ooeoooovoooeoo».oooooooooooi.«.oo(.i.o«.ooooookkoooo^i.oo(.t.oooookooi. 

ccccccc3cceccc3ccccccccccc33cc33c3ccceec33cccc33cc33cecec2cc3 

n n mm «■ 0 nn 00 00 0 a 

tv n din nra n nro nn n« n r* 






o^*wn , »0'ONm^O'- , riin^0'<iN©^Q’-'nio«»‘0’4ir^w^o*^fj , i?’J5^^®®‘O- , o<r)'f0’«Nfflo , -o — nin , »0'ON®O“O— *r<n 


ssssp^gs$s$ss^|s|||imi||||||||||||||i|||||||||||||||||||||||iiiiiiniii 

iiiiiiSiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii^miiiif iiiii if iif f i 

B3gliSSSSSiS8HSi»ilfia8i!6»»i2»ISBS!H!S»SHS»3SSS3S3» SttSSSSSHBi! 

8322222 = :: :ig83S5S583SS88SSS22 = ::2222i2t8S8S33SSSS2-22f 2St:2t SSli- 82^^!:® 222 2 


\ S8888888882888£8888S888882?882&88i;88R88828888;;888R88«8RR88?2<‘ 8&S8888?.88?s 



IS 

2 


ssssisiiiiMiiiHiiiiiiMHHiHUiiiiiHiiiiHiiSHiHiiiinHUSHHiu 

SS2222===::388SJS8S*«5mSS22::=2!2ni2::83SS«3S6SS2::22222!:2tS5;!83222t:S222222 


< 

J 

O. 


-4 (X O * 0 -O K 


®*-2 = 222222228S88588!5888S88S88S8S5;3555J!J?t83S8S52t;R!;S333J3S5?58-2E; 


67 




ORIGINAL PAGE IS 
OF POOR QUALITY 




E5 . Other Comets 


Tholen obtained five-color photometry of Comet Wilson on one night. The 
resulting colors were rather red, which now seems to be the norm as more and 
more color statistics are accumulated on comets. Photometry through the IHW 
filters was also obtained on one other night. It showed that Wilson had not 
yet developed very strong gaseous emission features. 


F. LABORATORY STUDIES OF DARK ORGANIC MATERIALS 


In the context of our continuing studies of solar system bodies of very 
low albedo, such as regions on planetary satellites, certain asteroids, and 
cometary nuclei, we have undertaken a series of laboratory spectroscopic 
observations of materials that may be present on some solar system bodies. 

Cruikshank, with the help of Research Assistant K. Uchida, collected from 
various other investigators, a suite of stable organic compounds that were 
made under various conditions of relevance to planetary science. These sub- 
stances, described below, were prepared to a uniform series of grain sizes by 
grinding and were then observed in diffuse reflectance with the spectrometers 
at the laboratories of R. N. Clark (USGS, Denver) and J. Salisbury (USGS, 
Reston). With Clark's spectrometer, the reflectance from 0.3 to 3.0 pm was 
measured, while Salisbury's instrument was used from 2.5 to 25 Jim. Represen- 
tative spectra in both wavelength regions are shown in Figures 32 and 33. 

The materials assembled in this study include the following: 

a. Kerogen, prepared from coal tar by dissolution of the soluble frac- 
tion. This material is that proposed by Veverka and Gradie as the 
possible coloring agent in the Trojan asteroids. 

b. Lampblack 

c. Insoluble organic extract from the Murchison CM carbonaceous 
chondrite, provided by L. Alaerts. 

d. Bulk Murchison meteorite 

e. Bulk Allende CV carbonaceous chondritic meteorite 

d. Hydrogen cyanide polymers, prepared by C. Matthews 

e. Stable residue from the proton irradiation of methane ice, prepared 
by G. Strazzulla 

f. Tholins produced in the laboratory of B. Khare and C. Sagan from 
spark or UV irradiation of gaseous mixtures of relevance to planetary 
atmospheres . 

We are preparing a publication that will include an atlas of all the 
spectra. These data constitute a useful resource for those scientists inter- 
ested in the occurrence and distribution of organic matter in the solar 
system. We will continue our own comparative spectroscopic work in the search 
for the composition of the dark material that is found in so many bodies in 
the outer part of the solar system. This work takes on special relevance to 
comets in the context of the confirmation of the low albedo of Comet Halley by 
in situ spacecraft measurements and the discovery of organic emissions in the 
spectrum of the nuclear regions of the comet. 


70 


REFLECTANCE 


Karogant Haaldual lnaolubla aatarlal In coal car aftar traa tnan t 

with toluana, haxana, and acatona. Partlda slxa 90-125 un 
Praparad by Kavan Uchlda and Oala Crulkahank fro* radpa by' 
Jonathan Cradla. 



Knra g aw . 23-2. 7w. 3.3m AM MP .p ii M D2 f 313 V23Qp313gCg 


□ 

O NICOLET OX V4.se 27 Hay S8 1 1. SB. 23 



WAVELENGTH <NM> 


Figures 32 & 33. Reflectance spectra of a laboratory sample of kerogen taken in 
two different wavelength regions in the laboratories of R. N. Clark and 
J. W. Salisbury. 


71 



G. THEORETICAL AND ANALYTICAL STUDIES: 

THERMAL INERTIAS AND THERMAL CONDUCTIVITIES OF PARTICULATE MEDIA 

The remarkably high thermal inertia found for the bright regions of Io 
led Sinton to ponder whether it was a consequence of a high thermal conductiv- 
ity caused by the conduction of S0 2 gas within the interstices of the frost 
particles. Considerable work had been done on the thermal conductivity of 
particulate media, both laboratory and theoretical studies, prior to the 
Apollo lunar landings, but relatively little recent work has been done. The 
theoretical calculation of the thermal conductivity for such media is of 
interest to chemical engineers for calculating the conduction in catalytic 
beds and to the petroleum industry for locating oil bearing shales, as well as 
being of interest to planetary scientists. Sinton has found that all such 
papers in the literature are flawed by a misunderstanding of the thermal con- 
ductivity of gases at low pressures. The transfer of heat between particles 
by conduction by a gas at low pressures is a process that in first order does 
not depend on the separation of the particles. This lack of dependence on 
separation is also true for transfer by radiation. The previous treatments 
are flawed in assuming that a thermal conductivity, which implies that conduc- 
tion between particles is inversely dependent on their separation, can be 
used. The theoretical treatments in the literature do not represent very well 
the behavior of the thermal conductivity of particulates. To construct a cor- 
rect theory, Sinton has focused attention on the conductance of heat between 
individual particles. Conductance considers the amount of heat exchanged bet- 
ween two particles at their actual separation. Processes that are considered 
in this calculation are, besides radiation and gaseous conduction, conduction 
by point contacts, conduction through the particles, and gaseous conduction 
around (bypassing) the particles. After evaluating the heat conductance bet- 
ween adjacent particles, it is then possible to derive the bulk thermal con- 
ductivity of a particulate medium. 

At least seven different sets of data giving the thermal conductivity of 
particulates as a function of gas pressure ranging from high vacua to atmo- 
spheric pressure are in the literature. These have been used to test the 
theory. Model curves were fit to the data by a least-squares fitting proced- 
ure with free parameters of particle size, porosity, contact conductance, and 
a parameter from vacuum technology called the accommodation coefficient. The 
accommodation coefficient is to gaseous conduction as emissivity is to radia- 
tion. It expresses the fraction of heat that is actually transmitted from a 
molecule to a surface upon impact. 

The agreement that the model fitting gives between the known particle 
size and that determined from the fitting for the size is impressive. The 
porosities of most of the laboratory samples were not recorded. The results 
from the models are given in Table VII. The fitted parameters, rows 7 to 10, 


72 



Table VII 


Results from Fitting Model to Laboratory Data 


Sample: 

#1 

#2 

#3 

#4 

#5 

#6 

#7 

Material 

pumice 

pumice 

pumice 

basalt 

basalt 

qtz. sand 

qtz . sand 

Size (pm) 

0-44 

44-104 

104-149 

44-104 

44-104 

74-104 

590-840 

Density 

0.88 

0.80 

0.84 

1.27 

1.27 

[1 . 55] a 

[0.95] 

Porosity 

— 

— 

— 

[0.46] 

[0.46] 

0.59 

0.36 

Temp. ( °K) 

277 

296 

296 

331 

221 

[293] 

[293] 

Reference^ 

1 

1 

. 1 

1 

1 

2 

2 

a 0 

. 36±0 . 06 

0 . 54±0 . 04 

0.44±0.01 

0 . 43±0 . 03 

0.39+0.08 

0 . 32±0 . 08 

0 . 16±0 . 01 

d (pm) 

10±1 

57+5 

180+23 

94±21 

83±9 

136±8 

652±4 

e 0 

. 87±0 . 07 

1 . 00±0 . 07 

0.61±0.09 

0 . 25±0 . 07 

0 . 27±0 . 06 

0 = 37±0 . 02 

0 . 64±0 . 04 

Rc (Ks/cal) 

169±21 

54+5 

65±5 

18±4 

20±3 

2 . 2±0 . 1 

0 . 36±0 . 03 

Number 

8 

9 

6 

9 

9 

10 

13 

RMS (%) 

12.5 

11.1 

4.0 

10.5 

18.3 

5.9 

9.6 

size ratio 

0.46 

0.77 

1.43 

1.28 

1.12 

1.52 

0.91 


a Items in brackets are estimated. 

^References: 1 = Wechsler et al. (1972); 2 = Woodside and Messmer (1961). 


73 




have the meanings: a = porosity, d = particle size, e = accommodation coef- 
ficient, Rc = contact resistance (reciprocal of contact conductance). Number 
is the number of data points for each sample and RMS characterizes the accu- 
racy of the model fit to the data. What is particularly gratifying is the 
last line, which expresses the ratio of the determined particle size to the 
mean size of the sieve fraction from line 2 of the table. They are all close 
to unity. The particle sizes range from 20 to 700 pm for the different 
samples. Thus, it appears that the agreement is not fortuitous. 

Figure 34 shows the agreement between the model calculations and the data 
for the basalt samples, #4 and #5 in Table VII. Note that these two data sets 
were for the same sample at two widely different temperatures. A separate 
fitting was made combining all of the data on this sample and the determined 
parameters were close to the means of those under #4 and #5 above. Figure 34 
shows that the model is capable of fitting thermal conductivity data of 
particulate media at a wide range of environmental pressures and temperatures. 

There are a number of applications of this theory to the planets. An 
immediate application is to the extensive set of thermal inertia measurements 
made of Mars by the Viking orbiters. Briefly, it was found that there was a 
bimodal distribution of thermal inertias (Palluconi and Kieffer, 1981, Icarus 
45, 415-426). The maxima of both modes were dependent upon the altitude of 
the surface above the zero reference level. If the particle sizes and other 
parameters for either mode are not dependent upon the altitude, then the 
observed thermal inertia should depend on the altitude as a consequence of the 
variation of atmospheric pressure from site to site. From the model it is 
found that at high pressures (1 mb to 1 bar) the thermal inertia should vary 
with pressure as (A + BP) 1 / 2 , where A and B are constants and P is the 
pressure. The pressure in the Mars atmosphere was assumed to vary as 
exp(-h/H) with the scale height H = 10 km. The gas was assumed to be C0 2 . 

The resulting fitting, Figure 35, is quite good. Note that the degree of 
curvature in the model curve is dependent upon the assumed scale height. 

In some calculations the scale height was left as a free parameter. The 
obtained scale height was changed by only a few percent from that adopted 
here. The constants A and B maybe related to the porosity and particle size. 
For the high thermal inertia regions the derived particle sizes and porosities 
are 80 pm and 15*. For the low thermal inertia regions they are 45 pm and 
61*. Thus the primary difference between the high and low inertia regions is 
the porosity. For these results, an accommodation coefficient of 1.0 has been 
assumed. Changing this parameter changes only the derived particle size. The 
contact conductance is unimportant at the high pressures experienced on Mars. 

These results for the high thermal inertia regions do not agree with the 
interpretation that is given in the literature (e.g., Palluconi and Kieffer), 
where it is assumed that there is a variation of particle size with altitude. 
If such a variation did exist, then the observed thermal inertia would vary 


74 



Thermal Inertia, cal cm -2 *K " 1 sec " 1,2 Thermal Conductivity, col cm 





Figure 34. The laboratory data at two temperatures and a wide ranee of pres- 
sures are compared to the model calculations for the basalt sample having a 
size range of 44-104 pm. The model fitting yielded a size of 88 pm. 



exp(h/H) 


Figure 35. The Mars thermal inertia data from Palluconi and Kieffer (1981, 
Icarus 45 , 514-426) are compared to models based on the new theory of thermal 
conduction in particulate media. For each of the two regimes, the particulate 
media are assumed to have the same size particles and porosity. The change in 
thermal inertia with altitude on Mars is produced by the change in atmospheric 
pressure, which is here assumed to vary exponentially with the generally 
assumed scale height of 10 km. 


75 




even more with altitude than the models with the known scale height are capa- 
ble of fitting. 

It is interesting, of course, to derive properties of the lunar regolith 
from its known thermal inertia and to compare these with the "ground truth" 
obtained from the Apollo missions. For particulates in a high vacuum only a 
combination of the fundamental parameters can be obtained. We have, arbitrar- 
ily, assumed that contact conduction is negligible in the lunar case. Then we 
are able to obtain the product of one minus the porosity raised to the 2/3 
power and the particle size. The derived value for this is 52 pm. The value 
for lunar fines, using the volumetric median of the particle size, is 43 pm. 

We can use this same modeling to derive this parameter for Mercury. The 
result is essentially the same as for the Moon. This is as expected because 
it is believed that the same process is responsible for producing the regolith 
in the two cases. 

The thermal inertias have been determined for only two asteroids. The 
thermal inertia of Ceres has been only crudely obtained. It is, however, 
larger than that expected from the lunar result. The thermal inertia of Eros 
was obtained with better accuracy. It is considerably higher than expected 
from the lunar agreement. The high values may be interpreted as indicating 
a larger particle size than found on the much more massive Moon and Mercury. 
Evidently, the smaller particle sizes either are not produced in meteorite 
impacts on the asteroids or they are not retained by gravitational attraction. 

Lastly, we get back to the question of the thermal inertia in the 
bright regions of Io. Sinton found that the higher thermal inertia cannot 
be explained by interstitial gas if the particle sizes are similar to those 
in the darker regions. One alternative is that the particle sizes are 
larger. If so, then the derived particle size is 4 mm. If this be the case, 
then the SO 2 frost would be like graupel. (Anybody for skiing Io?) Another 
alternative is that contact conductance is important or that the particles are 
needle shaped and contribute important conductance of heat through the long, 
skinny needles. Yet another possibility is that there is a "solid-state green- 
house effect" as has been investigated by R. H. Brown and D. L. Matson at JPL 
for high-albedo frosts. This would produce an apparent high thermal inertia 
in observations. 


76 



V 


H. EXTRASOLAR PLANETARY MATERIAL: 

THE SEARCH FOR DARK COMPANIONS OF K AND M GIANTS 

The presence of dark companions around main-sequence and red giant stars 
can be inferred from the former's gravitational influence on the latter. By 
observing the radial velocity of the primary over an extended period of time, 
shifts in the radial velocity due to the companion can be detected. As an 
example, Jupiter causes a shift of 12 m/s in the Sun's radial velocity. 
Therefore the identification of Jupiter sized objects requires a precision in 
the radial velocity measurements of at least this much. We are aiming for a 
precision of <10 m/s. 

K and M giant stars are excellent candidates in the search for low-mass 
(<0.05 solar masses) companions because their first overtone bands of CO in 
the 2. 2-2. 5 )im spectral region are relatively unblended and quite strong, 
therefore allowing high-precision measurements. The CO line frequencies can 
be determined very accurately in the laboratory. By placing N2O cells in the 
beam we obtain the zero point calibration. Due to the brightness of our 
objects we can achieve high signal-to-noise ratios of about 200 in <4 hours of 
integration. With a spectral resolution of 2 km/s using the CO bands at 3900 
to 4500 cm -1 , we achieve the desired radial velocity precision of <10 m/s. 

With this method, possible global oscillations and large-scale convective 
effects could also be identified. 

Since 1978 Hall and collaborator K. H. Hinkle (KPN0) have been taking 
measurements on several K and M giants. Using the 4-m Mayall telescope on 
Kitt Peak with a rapid scanning Fourier Transform Spectrometer at the coud§ 
focus they have obtained high quality infrared (1 to 5 4m) spectra of several 
giants as well as of Sirius and the Sun for comparison. Each measurement con- 
sists of a pair of scans: one forward and one backward. By taking the average 
and the difference between the two, an accurate subtraction of the noise is 
possible. 

Recently Heyer retrieved the 1978-1986 data of 11 objects from KPN0 for 
data analysis at IFA. The objects are j8 And, a Boo, aCet, y Dra, 0 Gem, 

4 Gem, a Hya, a Tau, 0 UMi , Sirius, and the Sun. 

Two necessary steps have to be undertaken to facilitate the analysis of 
this data. Since KPN0 uses a different data format, the collected data needs 
to be reformatted and made readable by the devices available at IFA. Before 
attempting the actual analysis of the measurements, the radial velocities 
obtained have to be corrected to their heliocentric value. However, because 
of the high precision required, the common ephemeris correction is not suffi- 
ciently accurate. 

Step one has been accomplished. Heyer is currently investigating the pre- 
cision required of the heliocentric velocity correction. There are several 
programs available with varying degrees of accuracy. The final choice should 


77 


be sufficiently accurate but not overstrain our computer account. Concur- 
rently, Hinkle and Heyer continue observations at KPNO on a Boo, /3 UMi , and 
the Sun. 

Upon concluding the investigation concerning the heliocentric velocity 
correction, all accumulated data needs to be thus corrected and then analyzed 
for shifts in the stellar radial velocities. This will be done on the 
Institute's VAX 11/785 computer using some programs already available and some 
to be developed. At the same time further data will be collected at Kitt 
Peak, since long-term measurements are required to find possible shifts. 
Periods of 100 s to over 3 years could be expected. 


78 



s 


III. OPERATION OF THE 2.2-M TELESCOPE 

The University of Hawaii's 2.2-m telescope, which was dedicated in 1970, 
was the first major facility on Mauna Kea. This telescope and the scientific 
work accomplished with it have been largely responsible for the acceptance of 
Mauna Kea as the premier observing site in the northern hemisphere. Through- 
out the 1970s and the 1980s, observations with the 2.2-m telescope have made 
major contributions to planetary science, and this telescope remains the 
primary instrument for ground-based optical and synoptic studies of the solar 
system. 

During the past several years, we have made substantial improvements to 
the operation of the telescope. Since these improvements are not generally 
known, and since a large fraction of our proposed budget will be used for the 
operation of the telescope, we will provide a short summary here. 

A. Telescope Control System 

The status of the telescope control system on August 1, 1984, was the 
following: The position encoders were the original telescope equipment and were 

no longer being manufactured. The control computer was an LSI 11/02 with 
single-density, single-sided floppy disks. The control program was virtually 
undocumented FORTH code. This system yielded pointing with an accuracy of 19 
arcsec radius rms measured against approximately 150 stars all over the sky. 

Our plan to modernize this aspect of the telescope operation consisted of 
three parts: (1) purchase new position encoders and design and fabricate the 

interface to the control computer, (2) install an LSI 11/23 computer with 
double-density, double-sided floppy disks, and (3) completely rewrite and 
document the control software, which would be based on that used at the IRTF. 

The new encoders and the interfaces were working on the telescope by 
April 1985. This aspect of the modernization was funded by the State of 
Hawaii. The LSI 11/23 and its new disks, also funded by the State of Hawaii, 
were operating properly by early November 1985. The new software is now 
written, debugged, and partially documented. We expect to start using the new 
software during the spring of 1987. Based on the residuals to the flexure 
maps were have made, which are 1.8 arcsec radius rms, we expect to improve the 
telescope pointing by an order of magnitude. The salaries of the project sci- 
entist and part-time programmer for the software upgrade were funded by NASA. 

Our plans for the future in this area are to finish the new software and 
to modify the control system to accept inputs from an autoguider. 

B. Data Acquisition System 

The data acquisition system consists of IR and CCD subsystems. On August 
1, 1984, the IR subsystem contained and LSI 11/02 computer with single-sided, 


79 



single-density floppy disk drives, a 10 Mb hard disk, a CAMAC crate with 
devices, a 1600 bpi tape drive, and a 10-year-old lock-in amplifier. A 
version of the IRTF photometry program was implemented. The CCD subsystem was 
portable and contained an LSI 11/02 computer with 32 Kb of memory and 
single-sided, single-density floppy disk drives, a De Anza image-processing 
and display system, and a 1600 bpi tape drive. The two subsystems were 
completely independent, but both provided quick-look data reduction, e.g., IR 
magnitudes, flat fielding, and so forth. 

Our plan to modernize this aspect of the telescope was to provide a 
resident CCD on-line data acquisition and control system of greatly increased 
capability, a VAX computer for off-line analysis, a much higher capacity hard 
disk, and the sharing of all available peripherals in the acquisition system 
by the two CPUs. At present, the CCD subsystem contains a 68010-based com- 
puter from Integrated Solutions, Inc., operating at 12 MHz with 3.25 Mb of 
memory. A 190 Mb hard disk, a dual-density, dual-sided floppy disk, and a 
1600 bpi tape drive are shared between the two subsystems via an easily con- 
figurable patch panel. These upgrades were funded by the State of Hawaii with 
partial salary support from NASA. We have also purchased with State funds a 
new dual-input lock-in amplifier from Princeton Applied Research and have 
interfaced it with the IR system. We have brought up at the 2.2-m telescope a 
VAX 730 computer with 2 Mb memory, 450 Mb hard disk, and two 1600 bpi streamer 
tape drives. The same data analysis software that is used on our VAX in Hono- 
lulu is now running on the 2.2-m telescope's VAX, and we are in the process of 
implementing IRAF. 

In the future we plan to have the VAX control the IR data acquisition - 
using the program written for an identical machine at the IRTF. This scheme 
will save the 2.2-m telescope the cost of software development and improve 
maintainability. We also intend to connect the CCD data acquisition computer 
to the VAX via a high-speed link and acquire an image display for the VAX. 
Figure 36 shows our final desired configuration. 

C. Operation from the Control Room 

On August 1, 1984, the astronomer needed to be at the mirror cell in 
order to observe. Nearly every instrument was manually operated and nearly 
all guiding was done visually through an eyepiece. This situation was also 
dangerous in addition to being inefficient, since the mirror was 20 feet above 
the floor. 

Our plan to modernize this aspect of the telescope operation was to 
provide remote operation of all instruments and to build a fully remote TV 
guider. Funds for the TV guider were provided by the NSF on December 15, 

1984. We have purchased a new TV camera with a Gen II intensifier and have 


80 



81 


Fig. 36. Integrated computer systems for the 2.2-m telescope. 











built the TV guider. We commissioned this new guider during August 1986. We 
have purchased a space TV camera using State of Hawaii funds. We have also 
provided for remote operation of nearly all our instruments: the CCD camera 

with motorized filter wheel, the Cassegrain Faint Object Spectrograph, and the 
IR CCD. All future instruments are planned to be remotely controlled as well. 

D. Dome Seeing and High-Resolution Imaging 

The seeing on Mauna Kea was known to be good, but this conclusion was 
based on subjective reports of the astronomers, not on systematic, objective 
measurements. We have conducted quantitative measurements for a period of 12 
months (Aug. 1985-Aug. 1986) at the 2.2-m telescope using a seeing monitor 
permanently mounted on a folded Cassegrain focus. We find the median seeing 
at the 2.2-m telescope is 0.88 arcsec. We believe that the dome makes 
substantial contributions to this average because the seeing degrades with 
increasing temperature difference between a point near the primary mirror and 
the outside air. When this is <0, the median seeing is 0.78 arcsec. 

Even prior to these results, however, we realized that the 2.2-m dome was 
degrading the seeing. We contracted with an architectural firm that had made 
similar studies for the CFHT and IRTF to recommend changes to our dome thermal 
environment with the associated cost. This study was funded by the State of 
Hawaii and was completed in June 1985. We then identified a source of funds 
to implement the suggested improvements and sent the the project out for bid 
last spring. Unfortunately, the bids were twice as expensive as estimated by 
the architect, so we were not able to proceed. The University has given this 
item a high priority on its FY 88 capital improvement budget that is now before 
the State Legislature. 

To further modernize the facility, the Director and other scientists at 
the Institute for Astronomy have taken several new instrumentation initia- 
tives. They are noted here briefly. 

The Institute has been awarded four of the Texas Instrument CCD detector 
arrays that are from the same batch as those used in the Hubble Space 
Telescope; one has arrived so far. We are using it, chiefly on the 2.2-m 
telescope. To support this development further, we have pending with the NSF 
a major grant proposal on which Dr. Henry is the PI. 

In addition to all of the above, we have received a high-sensitivity two- 
dimensional IR array detector that we have used for several runs on the 2.2-m 
telescope. This detector shows outstanding sensitivity, and we anticipate 
future heavy use. 

A substantial fraction of the assigned time on the 2.2-m telescope has 
been given to planetary observations since the telescope was first put into 
operation. NASA, through the planetary program, has supported a concomitant 
fraction of the operating expense of the telescope and associated equipment. 


82 


Most of the NASA-supported time is used by IFA scientists, but a large frac- 
tion is used by the planetary scientists of the Planetary Geosciences Division 
of the Hawaii Institute of Geophysics. In addition, some solar system time is 
made available to outside users with programs of interest to NASA. 

The statistics on planetary use of the 2.2-m telescope in 1986 
are as follows: 

First quarter 35% 

Second quarter 46% 

Third quarter 32% 

Fourth quarter 14% 

Total for 1986 32% 


83 



IV. PAPERS PUBLISHED OR SUBMITTED FOR PUBLICATION IN 1986 


Binzel, R. P., A. L. Cochran, E. S. Barker, D. J. Tholen, A. Barucci, M. 

DiMartino, R. Greenberg, S. J. Wei dens chilling, C. R. Chapman, and D. R. 
Davis (1987). Coordinated observations of asteroids 1219 Britta and 1972 
Yi Xing. Icarus (submitted). 

Buie, M. W. , and U. Fink (1987). Methane Absorption Variations in the 
Spectrum of Pluto. Icarus (in press). 

Cruikshank, D. P. (1986). Mauna Kea: A Guide to the Upper Slopes and 
Observatories . Institute for Astronomy, Honolulu. 

Cruikshank, D. P. (1986). Dark matter in the Solar System. COSPAR (1986 
volume, in press). 

Cruikshank, D. P. (1986). Telescopic studies of the satellites of Saturn. 

In Solid Bodies of the Outer Solar System , ESA SP-242, pp. 51-59. 

(Proc. of a conference at Vulcano, Italy). 

Cruikshank, D. P. , and R. H. Brown (1986). Satellites of Uranus and Neptune, 
and the Pluto-Charon system. In Satellites (J. Burns and M. S. Matthews, 
Eds.), pp. 836-873. Univ. of Arizona Press, Tucson. 

Cruikshank, D. P. , R. H. Brown, A. T. Tokunaga, R. G. Smith, and J. R. 

Piscitelli (1986). Volatiles on Triton: The infrared spectral evidence, 

2. 0-2. 5 micrometers. Icarus (submitted) 

Goguen, J., H. B. Hammel , and R. H. Brown (1986). V Photometry of Titania, 
Oberon, and Triton. Icarus (submitted). 

Hammel, H. B. , and M. W. Buie (1987). An atmospheric rotation period of 
Neptune determined from methane-band imaging. Icarus (in press). 

Hammel, H. B. , C. M. Telesco, H. Camp ins, R. M. Decher, A. D. Storrs, and 

D. P. Cruikshank (1986). Albedo maps of comets P/Giacobini-Zinner and 

P/Halley. 20th ESLAB Symposium on the Exploration of Halley's Comet , 

ESA SP-250, Vol . I, pp. 73-77. 

Hammel, H. B., C. M. Telesco, H. Campins, R. M. Decher, A. D. Storrs, and 

D. P. Cruikshank (1987). Albedo maps of comets P/Halley and 

P/Giacobini-Zinner. Astronomy and Astrophysics (in press). 

Hanel, R. , and others, including D. Cruikshank. (1986). Infrared observations 
of the Uranian System (Voyager 2 IRIS Team report). Science 233, 70-74. 

Harris, A. W., J. W. Young, J. Goguen, H. Hammel, G. Hahn, E. F. Tedesco, and 
D. J. Tholen (1987). Photoelectric lightcurves of the asteroid 1862 
Apollo. Icarus (submitted). 

Hartmann, W. K., D. J. Tholen, and D. P. Cruikshank (1987). The relationship 
of active comets, "extinct" comets, and dark asteroids. Icarus 69, 33-50. 


84 



» 


Piscitelli, J. R. , D. P. Cruikshank, and J. F. Bell. (1986). Laboratory 

studies of irradiated nitrogen-methane mixtures: Application to Triton. 
Icarus (submitted). 

Piscitelli, J. R. , D. J. Tholen, N. Lark, and H. B. Hammel (1986). 

Photoelectric photometry of comet P/Halley from Mauna Kea Observatory. 
20th ESLAB Symposium on the Exploration of Halley's Comet , ESLAB SP-250, 
Vol . Ill, pp. 499-502. 

Schleicher, D. G., R. L. Millis, D. Tholen, N. Lark, P. V. Birch, R. Martin, 
and M. F. A'Hearn (1986). The variability of Halley's comet during the 
Vega, Planet-A, and Giotto encounters. 20th ESLAB Symposium on the 
Exploration of Halley's Comet . ESA-SP 250, Vol. I, pp. 565-567. 

Sinton, W. M. (1986). Through the infrared with logbook and lantern slides. 
Pub. Astron. Soc. Pacific 98, 246-251. 

Telesco, C. M. , R. Decher, C. Baugher, H. Campins, D. Mozurkewich, H. A. 

Thronson, D. P. Cruikshank, H. B. Hammel, S. Larson, and Z. Sekanina 
(1986). Thermal-infrared and visual imaging of comet Giacobini-Zinner . 
Astrophvs. J. 310, L61-L65. 

Tholen, D. J. , M. W. Buie, and C. E. Swift (1987). Circumstances for Pluto- 
Charon mutual events in 1987. Astron. J. 93, 244-247. 

Tholen, D. J., 0. P. Cruikshank, W. K. Hartmann, N. Lark, H. B. Hammel, and J. 
R. Piscitelli (1986). A comparison of the continuum colors of P/Halley, 
other comets, and asteroids. 20th ESLAB Symposium on the Exploration of 
Halley's Comet . ESA SP-250, Vol. Ill, pp. 503-507. 

Veeder, G. J. , M. Hanner, and D. J. Tholen (1987). The nucleus of comet 
P/Arend-Rigaux. Astron. J. (in press). 


85 



ATTACHMENT 


ALBEDO MAPS OF COMETS P/GIACOB INI -Z INNER AND P /HALLE? 


H.B. HumI A.D. Storra 
D.P. Cruikahank 

Institute for Astronomy 
University of Hawaii 
Honolulu, Hawaii USA 


C.M. Talaaco R.M. Dachar 


NASA Mara hall Spaca 
Flight Can tar 
Huntavilla, Alabama USA 


H. Camplns 


Plana tary Sclanca 
Ina tltuta 

Tucson, Arizona USA 


ABSTRACT 

Near-simultaneous lnfrarad and visual naps of 
P/G lacobinl- Zinnar (P/G-Z) and P/Hallay ara com- 
blnad to craata naps of tha spatial variation of 
geometric albado. P/G-Z shows a minimum in albado 
naar 0.07 with an lacraaaa of a factor of 2 ovar 
about 30 arcane. Tha lowest albedos are offset 
from tha nucleus In the antl-suaward direction, 
coincident with a dust tall observed in tha IR. 

The P/Hallay albedos are higher than those found 
for P/G-Z and range from 0.2-0. 4, but tha trend of 
darker albedo In tha anti-sunward direction (along 
the tall) Is the sane. Ve attribute the albedo 
distribution to large, dark, fluffy grains con- 
fined to the orbital plane close to the nucleus. 
The high albedo values in P/Halley nay be due to 
enhanced flux in tha visual lange because of the 
comet's very snail phase angle. 

Keywords: Halley, Glacoblnl-Zlnner, Mapping, 

Albado distribution, Dust 


X. INTRODUCTION 

In this paper we present visual and Infrared nap- 
ping of conets P/Glacoblnl-Zlmtar and P/Hallay. 

Tha visual naps trace tha distribution of light 
reflected by dust, while the IR naps trace tha 
dust's thermal ealsslon. By coablnlng the two 
data sets, we construct naps of the spatial dis- 
tribution of albedo In the near-nuclear regions of 
these conets. Froa this information and tha infor- 
mation in the individual maps, we hope to build a 
consistent picture that explains both tha albado 
distribution in each comet and tha difference in 
albado between tha eomets. The observations and 
data reduction of P/G lacoblnl-Z Inner ara presented 
first, followed by a discussion of P/Hallay. Tha 
final section summarizes tha results. 

2. P/CIACOBINI-ZINNER 

2.1 Observations 

Near-simultaneous observations of P/Giacobini- 
Zlnnar were made from two sites on 4 August 1985. 
Additional details about the observations can be 
found in Ref. 1. 

2.1.1 Visual observations. The visual data con- 

86 


slat of a single CCD Image taken with the 
University of Hawaii 2.24-metar telescope (Mauna 
Rea Observatory). The 5-minute exposure was taken 
through an Imaging quality filter obtained froa 
the International Halley Vetch (IHV). The filter 
samples the red continuum at 6840 A. The Image 
scale of tha 500x300 array was 0.138 arcsec/plxel, 
giving a field of 69xb9 arcsec centered on the 
comet. The start time of the exposure was 12:06 UT. 

2.1.2 IR observations. The 10 . 3-ue map was made 
at the 2.3-eeter telescope at Vyoaing Infrared 
Observatory with a 20-pixel two-dimensional bolom- 
eter array. The full array covered 42x33 arcsec; 
overlapping fields were combined to produce a 
mosaic map with dimensions of approximately 80x160 
arcsec. The IR map was obtained over a period of 

5 hours centered on 10:00 UT. A contour map of 
the data is presented in Figure 1. 

2.1.3 Geometry . On 4 August 1985, the geocentric 
distance of P/Glacoblnl-Zlnner was 0.59 AU and Its 
heliocentric distance was 1.12 AU. The comet was 
at a moderate phase angle (64*) at the time of 
these observations, giving a favorable view of the 
tall structure. This was particularly evident in 
the IR thermal map, where there appears to be a 
large-grain tall superposed on the underlying coma 
and broader small-grain tail (Ref 1.). Because 
the IR observations bracket the visual obser- 
vation, time variability of the comet should not 
be a problem. 

2.2 Data Reduction 

2.2.1 Calibration . The visual data were cali- 
brated using measurements of Landolt star 144-755 
in conjunction with the IHV flux calibration for 
primary solar analogs. The star - Cep was used as 
a primary flux standard for the IR data. Details 
of the calibration for both data sets can be found 
in Ref. 1. Following absolute calibration of both 
data sets, the visual data were rescaled and 
binned to match both the beam size (7.5 arcsec 
square) and sample spacing (8.5 arcsec center-to- 
center) of the IR map. Care was taken to ensure 
that the maps were properly aligned. 

2.2.2 Albedo calculation . The average albedo was 
calculated by using the simplifying assumption 
(Refs. 2, 3) that the ratio of the total fluxes in 
the scattered solar and thermal emission energy 
distributions equals the ratio of the maximum 

ORIGINAL PAGE IS 
OF POOR QUALITY 1 



ORIGINAL PAGE IS 
OF POOR QUALITY 


values of *F\ for each distribution. The ratio of 
\F\(visible)AF\(infrared) is defined to be S. 

The alba do , g, la than equal to S/(S + X). For 
the dust temperatures Inferred from the IR map for 
Glacoblnl-Zlnner (Ref 1.), the peak thermal flux 
occurred near 10.8 Mm. We correct for the fact 
that the peak visual flux does not occur at 6840 l 
by multiplying our visual data by a factor of 
1.35. Albedos of other comets calculated with 
this method typically have values between 0.1 and 
0.3 (Ref. 3). 

2.3 Albedo nap 

2.3.1 Albedo values and distribution. The albedo 
map for P/Giacobini-Zirmer is shown" in Figure 2. 
The map was 11ml tad spatially by the size of the 
single CCD frame centered on the nucleus. The 
blanked-out region was contaminated by a field 
star. The albedos are low, ranging from 0.07 to 

0. 15. The lowest values are found offset slightly 
from the nucleus in the antl-sunward direction, 

1. e., along the tall. The values of g Increase 
radially froa the nucleus, except possibly in the 
direction of the tall, where the albedos are still 
slightly lower. 

2.3.2 Possible sources of error . A possible 
source of error in determination of the albedo map 
Is misalignment of the two data sets used to cal* 
culate the final map. To estimate the magnitude 
and morphology of such an error, tests cases were 
run with the visual image deliberately offset from 
the IR map. The intensity distributions of both 
the visual and nt naps are strongly peaked; even a 
very small error in alignment (lass than 3 arcsec) 
manifested Itself as an obvious distortion of 
albedo contours. Misalignment of the data sets 
cannot cause the observed albedo distribution. 

Another possible source of error is poor calibra- 
tion of the visual image. There was thin cirrus 
several hours before the P /Glacoblnl-Zlnner obser- 
vations were made. Simulated variation of the 
standard star flux by ±2QZ (an overestimate) of 
the observed value had no effect at all on the 
spatial distribution of albedo. The albedo values 
change by only to. 01. Therefore, even a conserva- 
tive estimate of the calibration error leads to 
material that Is dark. 

2.4 Interpretation 

The IR map indicates a tall composed of large 
(>100 um) grains (Ref. 1). The region of lowest 
albedo corresponds spatially with this large-grain 
tall (see Figures 1, 2). Because multiple internal 
scattering may cause large grains to appear dark, 
we infer we were observing a low albedo area 
caused by the presence of large dark grains* This 
is implied by both the low albedo (multiple scat- 
tering in large fluffy grains) and the location of 
low albedo relative to the nucleus and the large- 
grain tall. Large grains are expected be closer 
to the nucleus because of slower ejection veloci- 
ties. 


3. P/HALLEY 

3.1 Observations 

Observations of P/Halley were made on IS November 
1985. Additional details of the observations are 
discussed In Ref. 4. 


3.1.1 Visual observations. Five CCD images were 
obtained with the same filter, camera, and tele- 
scope described in Section 2.1.1. Two-by-two 
binning of the data was done at the telescope, 
yielding an linage scale of 0.276 arcsec/plxel. The 
field of each individual image is 69x69 arcsec. 
Four Images were taken with the nucleus of 
P/Halley shifted into each of the four corners. 

The nucleus was centered in the fifth image. The 
position of the nucleus was determined to within a 
fraction of a pixel for each image. The images 
were then shifted appropriately and added to form 
a single composite image cantered on the nucleus 
with a field of 125x125 arcsec. The nuclear region 
of the composite was carefully compared with the 
single direct CCD image to verify that no spurious 
structure was created during the mosaic pro- 
cessing. All five 2.5-mlnute exposures were taken 
between 9:50 and 10:14 UT. 

3.1.2 IR observations. The thermal IR map of 
P/Halley was made at the NASA Infrared Telescope 
Facility at Mauna Kea Observatory using the same 
bo lone trie array described in Section 2.1.2. The 
pixel size was 4. 3x4.3 arcsec, with the separation 
between pixels equal to 4.5 arcsec. The field cov- 
ered is about 80x100 arcsec. The map was created 
by combining overlapping fields obtained over a 
period of 6.7 hours centered on 11:00 UT. Figure 3 
shows a contour map of the IR data. 

3.1.3 Geometry. The geocentric distance of 
P/Halley was 0.7 AU on this data; the heliocentric 
distance was 1.7 AU. At the tine of these obser- 
vations, the comet was near opposition: the phase 
angle was 2* and changed by only about half a 
degree over the course of the observations. This 
means any significant dust tall should be along 
the line of sight. Nevertheless, the IR map shows 
the dust tall curves toward the southeast; the 
visual map Is also asymmetric, with the lowest 
brightness levels extended toward the southeastern 
corner. The IR and visual observations were not 
slmultaneous—the visual Images were obtained 

4 hours before the last sections of the IR map. 
Although the phase angle was nearly constant over 
this period, the position angle of the tall was 
changing rapidly, causing rotation of the tall by 
11* between two data sets. 

3.2 Data reduction 

The data reduction process is the same as 
described in Section 2.2 for Glaeoblnl-Zlnner. 

3.3 Albedo nap 

3.3.1 Albedo values and distribution . The 

P /Halley albedo map Is shown in Figure 4. The 
albedos are systematically higher than those In 
the P/Glacoblnl-Zlnner map. In fact, the values 
are high in general, ranging from 0.25 to 0.45. 
While there is less contrast (i.e., variation) In 
albedos in the P/Halley albedo map, the spatial 
distribution of albedo Is similar to that of 
P/Giacobini-Zinner: the region of lowest albedo 
seems to be in the direction of the tall — the 
anti-sunward direction. The albedo Increases 
radially from the nucleus, except along the tall, 
where the albedos are depressed. 

3.3.2 Possible sources of error . Misalignment of 
the ?/Halley maps Is unlikely for the same reason 
discussed above for P/Glacoblni-Zlnner: both -naps 
show a very strong central condensation. Thus the 

87 


--L 



Relative Declination foresee) Relative Declination (arceec) 


20 



Relative Right Ascension (arcsec) 

Figure 1. Map of P/Giacobini-Zinner at 10. S ym 



Relative Right Ascension (arcsec) 

Figure 2. Albedo Map of P/Giacobini-Zinner 


88 







Figure 3. Map of P/Halley at 10.8 ym 



Figure 4. Albedo Map of P/IIalley 


89 



albedo rap is very sensitive to the allgnrant; 
anil arrors rake large distortions. 

The sky wes clear and pho tone trie during the 
observations. Thera should be no problea with the 
standard star calibration. 

As discussed above, the position angle of the 
Halley's tall changed by 11* between the observa- 
tion tlaes of the visual and IR raps. However, 
the visual contours are very smooth over this 
range of angle, laplylng little or no change In 
the albedo rap if the lrage were rotated. 

3.4 Interpretation 

The overall higher albedos In P/Halley are proba- 
bly due to increased brightness in the visual rap. 
Enhanced brightness at continuum wavelengths near 
opposition has been seen in other comets (a.g., 
see Refs. 5, 6). This has been interpreted as 
enhanced backscatterlng by dust, and it ray pro- 
duce a brightness Increase as large as a factor of 
3 at near opposition phase angles (Ref. 5). If 
P /Halley is a normal comet, we would expect a 
backscatterlng peak at the phase angle observed 
here. The albedo values in the Halley rap would be 
comparable with typical cometary values if reduced 
by a factor of 2-3. 

The difference in contrast between the P/Halley 
and P/Glacoblni-Zioner raps ray indicate a smaller 
concentration of large grains, which presumably 
cause the deep depression in the P/G-Z data. 1R 
observations of Halley at moderate phase angle 
(60°-62 # ) trade in March 1986 show a dearth of 
large particles relative to Glaco b in i-Z inner (Ref. 
7). The March Halley observations and the IR 
observations of P/G-Z reported here were rade when 
the comets were at nearly identical heliocentric 
and geocentric distance. Thus, there Is a consis- 
tent picture of grain-size distribution and 
albedo. 

4. SUMMARY 

We have presented maps of the albedo distribution 
of comets P/Glacoblnl-Zlnner and P/Halley. The 
maps are derived from near-simultaneous visual and 
infrared rapping of the comets. Both comets show a 
similar distribution in albedo, with the albedo 
increasing radially froa the nucleus except along 
the antl-sunward direction, where the albedos 


remain lower. The contrast in albedo is much more 
pronounced in the P/Glacoblnl-Zinner map. We 
Interpret the albedo distribution as being pro- 
duced by large grains concentrated near the 
nucleus. The smaller contrast in the P/Halley rap 
is consistent with IR observations which show a 
dearth of large grains in P/Halley relative to 
P /G lacoblnl-Z inner • 

The high albedos in P/Halley are probably a result 
of Increased brightness in the visual data. The 
near-opposition geometry ray produce enhanced 
backscatterlng from the dust in the coma. Such 
backscatterlng peaks have been observed at low 
phase angle in other comets, typically causing an 
increase of a factor of 2-3 over the continuum 
values at moderate phase angles. 

5. ACKNOWLEDGMENTS 

This research was partly funded by NASA grant NGL 
12-001-057 to the University of Hawaii. H. 

Camp ins acknowledges NS? grant AST 34-14737 to the 
Planetary Science Institute. 


6. REFERENCES 

1. Telesco, C.M. at al. 1986, Thermal infrared 
and visual imaging of Comet Giacoblnl-Zlnner, 
Ap. J. (Letters) , in press. 

2. O'Dell, C.R. 1971, Nature of particulate 
ratter in comets as determined froa Infrared 
observations, Ap. J. 166, 675-681. 

3. Ney, E.P. 1982, Optical and Infrared observa- 
tions of bright comets in the range 0.5 sm to 
20 am, in Comets , ed. L.L. Wllkenlng, Tucson, 
University of Arizona Press, 323-340. 

4. Telesco, C.M. at al. 1986, in preparation. 

5. A' Hearn, M.F. at al. 1984, Comet Bowell 1980b, 
Astron. J. 89, 579-591. 

6. Mlllis, R.L. at al. 1982, Narrowband photom- 
etry of Comet P/S taphan-0 terms and the back- 
scattering properties of cometary grains, 
Astron. J. 37, 1310-1317. 

7. Camp ins, H. at al. 1986, Thermal IR imaging of 
Comet Halley, these proceedings. 


original 

POOR 


PAGE IS 

quality