Characteristics and large bulk density of the c-type main-belt triple asteroid



Yüklə 1,93 Mb.
Pdf görüntüsü
səhifə1/2
tarix12.08.2017
ölçüsü1,93 Mb.
  1   2

Characteristics and large bulk density of the C-type main-belt triple asteroid 

(93) Minerva 

F. Marchis 

a

,



b

,



, F. Vachier 

b

, J. D



ˇ urech 

c

, J.E. Enriquez 



a

, A.W. Harris 

d

, P.A. Dalba 



a

,

e



, J. Berthier 

b

, J.P. Emery 



f

,

H. Bouy 



g

, J. Melbourne 

h

, A. Stockton 



i

, C.D. Fassnacht 

j

, T.J. Dupuy 



i

, J. Strajnic 

k

a

Carl Sagan Center at the SETI Institute, 189 Bernardo Av., Mountain View, CA 94043, USA 



b

Institut de Mécanique Céleste et de Calcul des Éphémérides, Observatoire de Paris, UMR 8028 CNRS, 77 av. Denfert-Rochereau, 75014 Paris, France 

c

Astronomical Institute, Faculty of Mathematics and Physics, Charles University in Prague, V Holesovickach 2, 18000 Prague, Czech Republic 



d

DLR Institute of Planetary Research, Rutherfordstrasse 2, 12489 Berlin, Germany 

e

University of California at Berkeley, Department of Astronomy, B-20 Hearst Field Annex #3411, Berkeley, CA 94720, USA 



f

Earth and Planetary Sciences, University of Tennessee, 306 Earth and Planetary Sciences Building, Knoxville, TN 37996-1410, USA 

g

Centro de Astrobiología, INTA-CSIC, PO Box 78, 28691 Villanueva de la Cañada, Madrid, Spain 



h

California Institute of Technology, MS 301-17, Pasadena, CA 91125, USA 

i

Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822, USA 



j

Department of Physics, University of California Davis, 1 Shields Avenue, Davis, CA 95616, USA 

k

Direction de l’enseignement supérieur et de la recherche, Rectorat d’Aix-Marseille, Place Lucien Paye, 13100 Aix-en-Provence, France 



a r t i c l e

i n f o


Article history: 

Received 18 December 2012 

Revised 11 February 2013 

Accepted 11 February 2013 

Available online 5 March 2013 

Keywords:

Asteroids

Satellites of asteroids 

Adaptive optics 

Orbit determination 

Interiors

a b s t r a c t

From a set of adaptive optics (AO) observ ations collected with the W.M. Keck telescope between August 

and September 2009, we derived the orbital parameters of the most recently discovered satellites of the 

large C-type asteroid (93) Minerva. The satellites of Minerva, which are approximately 3 and 4 km in 

diamet er,  orbit  very  close  to  the  primary  ($5 and $8 Â R

p

and $1% and $2% Â R



Hill

) in a circular manner, 

sharing common characteristics with most of the triple asteroid systems in the main-belt. Combining 

these AO observations with lightcurve data collected since 1980 and two stellar occultatio ns in 2010 

and 2011, we removed the ambiguity of the pole solution of Minerva’ s primary and showed that it has 

an almost regular shape with an equivalent diameter D

eq

= 154 ± 6 km in agreement with IRAS observ a- 



tions. The surprisingly high bulk density of 1.75 ± 0.30 g/cm 

3

for this C-type asteroid, suggests that this 



taxonomic class is composed of asteroids with different compositions, For instance, Minerva could be 

made of the same material as dry CR, CO, and CV meteorites. We discuss possible scenarios on the origin 

of the system and conclude that future observations may shine light on the nature and composition of 

this fifth known triple main-belt asteroid. 

Ó 2013 Elsevier Inc. All rights reserved. 

1. Introduction 

The serendipitous discovery of Dactyl, a companion of the 

main-belt asteroid (243) Ida, seen during the Galileo spacecraft 

voyage to Jupiter (

Chapman et al., 1995 

), gave birth to a new field

of study for small Solar System bodies. Today, using techniques 

such as high angular resolution imaging with the Hubble Space 

Telescope (HST), adaptive optics (AO) on ground-based telescopes, 

radar observati ons and photometric studies, about 200 multiple 

asteroids (asteroids with one or several moons) are known in all 

populations of small Solar System bodies, from the near-Earth 

asteroids to the Kuiper Belt Objects. The study of these multiple 

asteroids is an opportun ity for planetary astronomers to obtain in- 

sights on the asteroids such as their masses and densities and how 

these quantities possibly relate to their compositions , see e.g. 

Mar-


chis et al. (2012a)

. Their existence and the understand ing of their 

formatio n and evolution also provide a direct window to the his- 

tory of our Solar System. 

In this work, we describe a study of one of these recently dis- 

covered triple asteroid systems. (93) Minerva is a large (H

v

= 7.7, 


P

spin


= 5.97909 h in  Tungalag  (2002)) asteroid  discovered  by  J.C. 

Watson at Ann Arbor, MI, USA in 1867. Located in the middle of 

the main-belt (a

Minerva


= 2.75 AU,  e

Minerva


= 0.14,  i

Minerva


= 9°), it 

was initially classified as a member of an old collisiona l family 

(1.2 ± 0.4 Gyr) named  ‘‘Gefion’’ consisting  of  973  members  based 

on

Nesvorny 



´ et al. (2005)

. Because most of the members of this 

family are identified as S-type, and (93) Minerva is known to be 

a C


b

-type (


Lazzaro et al., 2004 

) or C-type (

deMeo et al., 2009 

),

the large Minerva is likely to be an interloper in the Gefion family. 



0019-1035/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. 

http://dx.doi.org/10.1016/j.icarus.2013.02.018

Corresponding author at: Carl Sagan Center at the SETI Institute, 189 Bernardo 



Av., Mountain View, CA 94043, USA. 

E-mail address: 

fmarchis@seti.org

(F. Marchis).

Icarus 224 (2013) 178–191

Contents lists available at 

SciVerse ScienceDi rect 

Icarus


j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / i c a r u s

The size and geometri cal albedo of Minerva were reported in 

several mid-infrared surveys such as IRAS (D = 142 ± 4 km, 

p

v

= 0.073 ± 0.004  in 



Tedesco et al. (2002)

), MSX (D = 157 ± 3 km, 

p

v

= 0.060 ± 0.002  in 



Price et al. (2001)

), AKARI (D = 147 ± 2 km, 

p

v

= 0.068 ± 0.003 in 



Usui et al. (2011)

) and in a reanalysis of IRAS 

and MSX data (D = 165 ± 8 km,  p

v

= 0.0597 ± 0.0021  in 



Ryan and 

Woodward (2010)

). There is a clear discrepancy in the sizes of 

Minerva derived from mid-infrared catalogs (up to $23 km), indic- 

ative of different viewing geometries of the asteroid at the time of 

the observations, but also linked to a slightly different modeling in 

these works. The disparate diameters could be due to the use of 

different absolute magnitudes (H

v

) for Minerva, assumptions on 



the thermal model and distinctive wavelength bands of observa- 

tions between these surveys. 

In the next section of this work, we describe observati ons of 

(93) Minerva from different sources. The key finding of this work 

is the discovery of two satellites around the large asteroid, which 

was made possible by W.M. Keck AO observations . We also com- 

bined our analysis with a large set of lightcurve data and stellar 

occultation data already published in the literature. Section 

3

de-


scribes how a coherent analysis of these disparate sets of data al- 

lowed us to constrain the pole orientati on, size and shape of the 

primary of the system. The mutual orbits of the satellites are deter- 

mined in Section 

4

and compared to other known triple asteroid 



systems. We discuss the composition and the origin of the system 

based on the first direct measure ment of its density in Section 

5

. In 


Section

6

we discuss the origin of the system based on the orbital 



parameters and contrast our study with other multiple asteroids. A

summary of our findings and future work is discussed in Section 

7

.

2. Observations and data processi ng 



2.1. Adaptive optics observatio ns 

Observation s of (93) Minerva were collected on August 16, 2009 

using the W.M. Keck II telescope located atop Mauna Kea on the Big 

Island of Hawaii. The Nasmyth platform of this 10 m telescope has 

been host to the NIRC2 infrared camera equipped with an AO sys- 

tem since 2001 (

Wizinowich et al., 2000; van Dam et al., 2004 

).

Thanks to the large aperture size of the telescope and the potential 



of AO, the angular resolution expected on these images for a V $ 12

bright target (see

Table 1

) is $40–50 milli-arcsec (mas), very close 



to the diffraction limit of the telescope in Fe II (k

c

= 1.6455 



l

m,

D



k

= 0.0256 

l

m) and Kcont (k



c

= 2.2706 

l

m,

D



k

= 0.0296 

l

m). The 


first images were taken in these broadband filters (

Table 1


) from 

13:38 to 13:45 UT using the narrow camera with a pixel scale of 

9.94 mas. A final image was obtained using our automatic pipeline 

while observing at the telescope by shift-adding 3–6 frames with 

an exposure time of 60 s (30 s Â 2 coadds). These frames were 

flat-field corrected, and we used a bad-pixel suppressing algorithm 

to improve the quality before shift-add ing them. After this basic 

data processing, the final images revealed the presence of one 

small satellite at 3 o’clock at 0.42 arcsec (see

Fig. 1


). Addition al 

observati ons taken at 13:56 UT revealed the existence of an even 

closer ($0.26 arcsec) compani on at 5 o’clock on the image. These 

existence of these satellites was corrobor ated by additional data 

taken 1 h later (after 14:40 UT), which confirmed that these com- 

panions are gravitationally linked to the primary and orbit around 

the primary in the clockwise direction. 

Addition al data were collected from September 6 to September 

28, 2009 when the asteroid was approaching its opposition (elon-

gation reached 160 ° on September 28). Because the asteroid was 

brighter (V $ 11.7) and closer (distance to Earth d $ 1.80 AU), the 

AO correction was significantly better and the two satellites were 

easily detectab le in the images after the basic processing described 

above. (


Fig. 2

). These additional observation runs confirmed the 

genuinen ess of our discovery and allowed us to derive the charac- 

teristics of this multiple asteroid system. 

(93) Minerva is the fifth triple asteroid discovered among the 

main belt asteroids (

Marchis et al., 2009 

), after 87 Sylvia (

Marchis

et al., 2005 



), 45 Eugenia (

Marchis et al., 2007 

), 3749 Balam (

Mar-


chis et al., 2008c 

) and (216) Kleopatra (

Marchis et al., 2008d 

). For 


the sake of clarity we adopt from now on the designation (93)

Minerva for the whole system and Minerva for the primary alone. 

Minerva’s compani ons are 5–6 magnitudes fainter than the pri- 

mary body and are separated by about 0.26 and 0.42 arcsec respec- 

tively (

Fig. 1


). They are the smallest and closest satellites of a large 

asteroid ever seen. This discovery illustrates the improvement in 

image quality and sensitivity achieved with the AO technolo 

gy 


over the past 15 years (

Marchis et al., 2012b 

).

The size of the satellites can be estimated by assuming that both 



the moons and the primary have the same albedo and comparing 

their flux ratios (/). In 

Table 2

, we list the effective diameters of 



satellites (D

eff


(satellite)), derived in each image with the equation 

D

eff



ðsatelliteÞ ¼ /

1=2


 D

eff


ðprimaryÞ

where D


eff

(primary) is the measured effective diamete r on the re- 

solved adaptive optics observati ons. For Minerv a I, the outer moon

and Minerva II, the inner moon, we derived the following diameters 

by averaging D

eff


(satellite) measureme nts:  D

I

= 3.6 ± 1.0 km  and 



D

II

= 3.2 ± 0.9 km. 



A single star (BD+11 229), with a brightnes 

s similar to (93)

Minerva (V = 11.06) and located at $0.4° from the asteroid, was ob- 

served shortly after the August 16, 2009 observati on of (93) Miner- 

va. This additional set of observations is useful to estimate the 

quality of the data, check for possible artefacts in the Point Spread 

Function (PSF) of the instrument and deconvolve the data during 

further analysis. This PSF image was observed at 13:50 UT using 

an Fe II filter with the narrow camera and a total exposure time 

of 25 s per frame (5 s Â 5 coadds) and was processed in a manner 

similar to the asteroid observations . We estimated the angular res- 

olution on the Minerva data ($41 mas) by measuring the full width 

at half maximum (FWHM) of the PSF. Interestingly, the FWHM 

measure ment of the Minerva observations varies from 0.10 to 

Table 1

Conditions of the Keck II AO observations of (93) Minerva collected in 2009. 



Date (UT)

Time (UT)

Filter 

Airmass 


Predicted (V)

Elongation (°)

a

Phase (°)



D distance to Earth (AU)

16-August-09 13:38:24 

Fe 

II 


1.03 12.47 

114.69 


19.95  2.110850 

16-August-09 13:45:02 

Kc 1.03 12.47 

114.70 


19.96  2.110799 

16-August-09 13:56:12 

Fe 

II 


1.02 12.47 

114.71 


19.96  2.110719 

16-August-09 14:25:02 

Fe 

II 


1.01 12.47 

114.72 


19.95  2.110509 

16-August-09 14:40:54 

Fe 

II 


1.01 12.47 

114.73 


19.95  2.110401 

16-August-09 15:29:45 

Fe 

II 


1.04 12.47 

114.77 


19.94  2.110048 

06-September-09 15:30:58  Fe 

II  1.23 

12.13 


135.32 

15.07 


1.916779 

12-September-09 15:10:47  Fe 

II  1.26 

12.02 


141.72 

13.17 


1.874003 

15-September-09 15:11:48  Fe 

II  1.32 

11.96 


145.02 

12.15 


1.855197 

28-September-09 13:34:08  Fe 

II  1.18 

11.71 


159.74 

7.22 


1.797652 

F. Marchis et al. / Icarus 224 (2013) 178–191

179


0.12 arcsec,  implying  that  the  primary  is  resolved  with  2–3 ele- 

ments of resolution. 

Because the final images have a high signal-to-noi se ratio vary- 

ing from 1400 to 2800, and because it is known that the Keck AO 

correction is relatively stable for bright V $ 11–12 targets, we 

can apply the AIDA myopic deconvolution algorithm (

Hom et al., 

2007


) to the observati ons using the PSF frames recorded on August 

16 as an input estimate of the real PSF. The deconvolved frames 

shown in 

Figs. 1 and 2

suggest that the projection of Minerva’s pri- 

mary is almost spherical with a maximum ratio 2a

0

/2b


0

(ratio of the 

projected major (a

0

) and minor (b



0

) axes derived by fitting an ellipse 

to the resolved primary image) of 1.2 (2a

0

/2b



0

= 1.1 in average).

From these deconvolved images, we extracted the silhouett e of 

the Minerva primary, which will be used to derive the shape of 

Minerva in Section 

4

. From simulations of asteroid observati ons, 



Marchis et al. (2012b)

showed that the typical error on the 

major-axis estimate of an asteroid resolved with $3 elements of 

resolution (SNR $ 2000, FWHM(PSF) $ 40 mas) is  3%,  which 

corresponds to 4 mas or 6 km at the distance of Minerva for the 

epochs of observations .

2.2. Lightcurve data 

We gathered photometric lightcurve observations of (93)

Minerva from various sources listed in 

Table 3


from 1980 to 

2012. The first set of lightcurv es taken from 1980 to 2009 were 

published in 

Harris and Young (1989)

,

Debehogne et al. (1982),



Denchev (2000)

and


Torppa et al. (2008)

. We retrieved them from 

the Asteroid Photometric Catalog, 

1

a repository web site for asteroid 



lightcurves . Starting in 2011, the robotic 0.6 m super-LO 

TIS tele- 

scope located at Kitt Peak was also used to record R band observa- 

tions. The data were reduced using a classic relative photomet ry 

method consisting in comparing the flux of the asteroid with the flux

of nearby stars over the course of the night. The last night of the 

super-L OTIS data (February 18, 2012) was affected by poor weather 

condit ions. On February 4, 2011, the 0.8 m telescope of the Observa- 

toire de Haute-Prov ence collected photometri c observations of over 

two-thi rds of the spin period. 

The lightcurves in 

Fig. 3


display relative intensity versus time. 

They have a remarkably low amplitude (less than 0.25 mag) even 

when the asteroid was observed edge-on in February and March 

2011, suggestin g that the asteroid primary, which dominates the 

flux variation in the lightcurves, is almost spherical . Based on the 

flux ratio between the moons (D

s

$ 3 km) and the primary 



(D

p

$ 141 km  by 



Tedesco et al., 2002 

) we expect a mutual event 

to generate a small attenuation ($0.001 mag) on  the  lightcurv e. 

Because this signature of mutual event is significantly smaller than 

the accuracy on our lightcurv e, none were reported in the light- 

curve sample. 

2.3. Stellar occultatio n data 

Similarly to AO images, a successful observation of an occulta- 

tion constrain s the size of the asteroid, its spin axis orientation, 

and its shape (

D

ˇ urech et al., 2011 



). There are ten occultations by 

Minerva reported in 

Dunham et al. (2012)

but only three events 

have more than two positive chords measured. The events took 

place on 1982-11-22 (9 chords, max duration $11 s), 2010-12-24 

(9 chords, max duration $12 s) and  2011-01-28  (4 chords,  max 

duration $12 s). In  this  work,  we  used  only  occultations  from 

2010 and 2011, since the 1982 event was observed visually by 

all observers and 

D

ˇ urech et al. (2011)



showed that there are 

Fig. 1.  Observations of (93) Minerva recorded on August 16, 2009 with the W.M. Keck II telescope and its adaptive optics. Five of these observations were recorded with the 

Fe II filter centered at k

c

= 1.6455 



l

m. The processed frame taken at 13:45 UT was recorded using the Kcont filter centered at k

c

= 2.2706 



l

m. The horizontal arrow indicated 

the position of the outer moonlet ‘‘Minerva I’’ seen on the frame recorded at 13:38 UT. The vertical arrow corresponds to the position of the inner moonlet ‘‘Minerva II’’ 

detected on the frame taken at 13:56 UT. The position of each arrow is fixed on the images, hence it shows the motion of the moonlets over $2 and $1 h of observation for 

Minerva I and Minerva II respectively. 

1

http://asteroid.astro .helsinki.fi/apc/asteroids/93



.

180


F. Marchis et al. / Icarus 224 (2013) 178–191

systematic shifts in time, likely due to the reaction times of the 

observers.

Fig. 4

shows the observed chords of the stellar occultation. The 



solid chords were observed using video or CCD, the one dashed 

chord was observed visually and the dotted line in the first occul- 

tation is the negative observati on with no occultation detected. 

These data will be used to constrain the size and shape of Miner- 

va’s primary in Section 

4.1


.

Unfortunate ly, no observers reported the detection of secondary 

stellar occultation events due to the satellites . Such a detection 

would have allowed the direct measure the sizes of the satellites 

as

Descamps et al. (2008)



reported for Linus, companion of (22)

Fig. 2.  Additional W.M. Keck II AO observations of (93) Minerva and its two moonlets were recorded in September 2009 using an Fe II band filter. The positions of the 

moonlets are indicated with horizontal and vertical arrows for Minerva I and Minerva II respectively. The variable quality of the data provided by the AO system is visible. For 

instance, on September 6 the two satellites are barely visible since their brightness is very close to the uncorrected halo surrounding the primary. The intensity of this halo is 

reduced for the image taken on September 15, improving the contrast on the satellite signals. 

Table 2


Astrometric positions of Minerva’s moonlets extracted from W.M. Keck AO observations taken in 2009 with the NIRC-2 camera. The diamete rs  of  the  moons  are  estimated  by 

assuming the same albedo, and therefore surface composition, between the primary and the satellites. 

Date (UT)

Time (UT)

JD À2455000 Satellite 

codeObs X (arcsec)

Y (arcsec)

Diameter (km)

16-August-09 13:38:24 

60.0683333 

Minerva 

I

568  À0.404 0.034 



4.1 

16-August-09 13:45:02 


Kataloq: presentation
presentation -> Dunedin study of all children born in 1972, to age 21
presentation -> Vision, missions et structuration
presentation -> Göstərir. «Vital» klinika ‐ ən müasir tələblərə
presentation -> Есть ли необходимость всем научным журналам стремиться в глобальные индексы цитирования?
presentation -> 1. Родовой вид детского травматизма – повреждения костей и мягких тканей новорожденного в процессе родов: часто при неправильном положении, тазовом предлежании плода,
presentation -> NordriDesign
presentation -> A hybrid Approach to Answering Biographical/Definitional Questions Ana Licuanan, Scott Miller, Ralph Weischedel, Jinxi Xu
presentation -> MedlinePlus, MedlinePlus en español, nihseniorHealth & nlm’s main web site Joyce Backus

Yüklə 1,93 Mb.

Dostları ilə paylaş:
  1   2




Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©azkurs.org 2020
rəhbərliyinə müraciət

    Ana səhifə