Thermal physics of the inner coma: ALMA studies of the methanol distribution and excitation in comet C/2012 K1 (PanSTARRS)

1 MB·

by  M. A. Cordiner

Accepted for publication in ApJ, January 2017
PreprinttypesetusingLATEXstyleemulateapjv.5/2/11
THERMAL PHYSICS OF THE INNER COMA: ALMA STUDIES OF THE METHANOL DISTRIBUTION AND
EXCITATION IN COMET C/2012 K1 (PANSTARRS)
M. A. Cordiner1,2, N. Biver3, J. Crovisier3, D. Bockele´e-Morvan3, M. J. Mumma1, S. B. Charnley1, G.
Villanueva1, L. Paganini1,2, D. C. Lis4, S. N. Milam1, A. J. Remijan5, I. M. Coulson6, Y.-J. Kuan7,8, J. Boissier9
Accepted for publication in ApJ, January 2017
ABSTRACT
We present spatially and spectrally-resolved observations of CH OH emission from comet C/2012
3
K1 (PanSTARRS) using The Atacama Large Millimeter/submillimeter Array (ALMA) on 2014 June
7
28-29. Two-dimensional maps of the line-of-sight average rotational temperature (T ) were derived,
1 rot
covering spatial scales 0.3(cid:48)(cid:48)−1.8(cid:48)(cid:48) (corresponding to sky-projected distances ρ ∼ 500-2500 km). The
0
CH OH column density distributions are consistent with isotropic, uniform outflow from the nucleus,
2 3
with no evidence for extended sources of CH OH in the coma. The T (ρ) radial profiles show a
3 rot
n significant drop within a few thousand kilometers of the nucleus, falling from about 60 K to 20 K
a betweenρ=0and2500kmonJune28,whereasonJune29,T fellfromabout120Kto40Kbetween
J rot
ρ= 0 km and 1000 km. The observed T behavior is interpreted primarily as a result of variations
rot
8 in the coma kinetic temperature due to adiabatic cooling of the outflowing gas, as well as radiative
2 cooling of the CH OH rotational levels. Our excitation model shows that radiative cooling is more
3
importantfortheJ =7−6transitions(at338GHz)thanfortheK =3−2transitions(at252GHz),
] resulting in a strongly sub-thermal distribution of levels in the J = 7−6 band at ρ (cid:38) 1000 km. For
P
both bands, the observed temperature drop with distance is less steep than predicted by standard
E
coma theoretical models, which suggests the presence of a significant source of heating in addition to
.
h the photolytic heat sources usually considered.
p Subject headings: Comets: individual (C/2012 K1 (PanSTARRS)), molecular processes, techniques:
- imaging spectroscopy, techniques: interferometric
o
r
t
s 1. INTRODUCTION coma, at distances less than a few thousand kilometres
a
from the comet’s surface.
[ CometsareconsideredfossilsoftheearlySolarSystem
With the advent of the Atacama Large Millime-
— frozen relics containing ice, dust and debris from the
1 ter/submillimeter Array, high-sensitivity, high angular-
protoplanetary accretion disk. Having existed in a rela-
v resolution millimeter-wave interferometry of typical,
tively quiescent state since their formation (e.g. Davids-
8 moderatelybrightcometshasbecomepossible. ALMA’s
son et al. 2016), cometary compositions can provide
5 unique capabilities allow us to probe the physical and
unique information on the thermal and chemical charac-
2 chemical structure of the innermost regions of the coma
teristicsoftheearlySolarSystem. Mostofourknowledge
8
in unprecedented detail, leading to new insights into
0 on cometary compositions comes from remote (ground-
the properties of the coma and the nucleus. The first
. based) observations of their gaseous atmospheres/comae
1 cometary observations using ALMA were reported by
(Cochran et al. 2015), for which the (typically relatively
0 Cordineretal.(2014),whomeasuredthedistributionsof
low) angular resolution and incomplete spatial coverage
7 HCN, HNC and H CO in comets C/2012 F6 (Lemmon)
limitstheamountofinformationthatcanbeobtained. A 2
1 and C/2012 S1 (ISON) and demonstrated unequivocally
particularproblemisthelackofunderstandingregarding
: thatHNCandH COarereleasedinthecoma(asaresult
v the physical and chemical structure of the near-nucleus 2
of photolytic and/or thermal processes), whereas HCN
i
X originates from (or very near to) the nucleus.
[email protected]
r 1Goddard Center for Astrobiology, NASA Goddard Space For simplicity, a constant coma kinetic temperature
a Flight Center, 8800 Greenbelt Road, Greenbelt, MD 20771, is commonly assumed during analysis of microwave and
USA. sub-mm cometary observations. However, theoretical
2Department of Physics, Catholic University of America, andobservationalstudiesincreasinglyshowthattemper-
Washington,DC20064,USA.
3LEISA, Observatoire de Paris, CNRS, UPMC, Universit´e atures can vary substantially over short distance scales
Paris-Diderot,5placeJulesJanssen,92195Meudon,France. in the coma. Strong variations in temperature within
4LERMA, Observatoire de Paris, PSL Research University, a few hundred kilometers of the nucleus are predicted
CNRS, Sorbonne Universit´es, UPMC Univ. Paris 06, F-75014, by coma hydrodynamic/Monte Carlo simulations (e.g.
Paris,France.
5NationalRadioAstronomyObservatory,Charlottesville,VA K¨or¨osmezey et al. 1987; Combi et al. 1999a,b; Rodgers
22903,USA. & Charnley 2002; Tenishev et al. 2008), but these mod-
6JointAstronomyCentre,Hilo,HI96720,USA. els are largely untested due to a lack of comparative ob-
7National Taiwan Normal University, Taipei 116, Taiwan, servational data. Early observational reports of inner
ROC.
8Institute of Astronomy and Astrophysics, Academia Sinica, coma temperature variations were based on long-slit in-
Taipei106,Taiwan,ROC. fraredspectroscopyofHCNandCOintheunusuallyac-
9IRAM, 300 Rue de la Piscine, 38406 Saint Martin d’Heres, tive comet C/1995 O1 (Hale-Bopp) (Magee-Sauer et al.
France.
2
1999; DiSanti et al. 2001). In a far-IR study of H2O 14
andHDOemissionfromcometC/2009P1(Garradd)by Nancay 18 cm obs.
Bockel´ee-Morvan et al. (2012), a strongly variable tem- Weighted mean
peraturelawwassuggested—reachingaminimumat(4-
12
20)×103 km from the nucleus and rising to 150 K in the
outer coma. Advances in ground-based infrared instru-
mentationanddataanalysistechniquespermittedBonev 1)−10
etditiosantals.n(icn2e0t0shc7ea,lH2e0s20O∼8,r22o00t-1a13t5i,o02n00a1kl4me)xtcionitdafoteiutoernctctosemimgenptiesfi,rcawatnuhtricevhaorvlieeadr- 28H) ( s10
tothefirstdetailed(quantitative)comparisonswiththe- Q(O 8
ory for comet 73P/Schwassmann-Wachmann 3 (Fougere
et al. 2012).
To-date, a general lack of information concerning spa- 6
tialvariationsintheexcitationofcometarymoleculesim-
pedes the accuracy of important results in the cometary
literature, hinderingthe derivation ofaccurate cometary 4
1.3 1.4 1.5 1.6
mixing ratios, as well as introducing errors into the par-
rH (AU)
ent scale lengths of distributed sources. Thus, detailed
measurements of coma temperatures are required in or- Fig. 1.— Production rates for OH in comet C/2012 K1
(PanSTARRS)asafunctionofheliocentricdistanceobtainedusing
der to confirm and expand upon the previous observa-
the Nanc¸ay radio telescope. The values shown are pre-perihelion,
tional findings, to stimulate revision and refinement of andcovertheperiod2014-06-14to2014-07-17. Dashedhorizontal
theoretical models. line shows the mean value of 9.0×1028 s−1. Vertical dotted line
By virtue of its large abundance in comets (typically correspondstotheepochoftheALMAobservations.
on the order of a few percent with respect to H O), and
2 updated in real-time using the latest JPL Horizons or-
itsstrongrotationalbandsthroughoutthemm/sub-mm,
bitalsolution(sampledat15sintervalsandinterpolated
methanol(CH OH),isanidealmoleculeformappingthe
3 using a 4th-order polynomial). The data were flagged,
coma temperature distribution. Here we present results
calibrated and cleaned using standard CASA routines
exploting the high resolution and sensitivity of ALMA
(seeforexampleCordineretal.2014),withGanymedeas
to provide new information on the distribution and ex-
the flux calibrator. Imaging was performed using a grid
citation of CH3OH in the inner coma of the Oort-cloud size of 768×768 pixels, with 0.05(cid:48)(cid:48) pixel scale. The re-
comet C/2012 K1 (PanSTARRS), and interpret our ob-
sulting data cubes were corrected for the response of the
servations using non-LTE radiative transfer models.
ALMA primary beam and then transformed from celes-
2. ALMACH3OHOBSERVATIONS tial coordinates to sky-projected distances with respect
tothecenterofthecomet. Spectralfluxesperbeamwere
ALMAobservationsofC/2012K1(PanSTARRS)were
subsequentlyconvertedtotheRayleigh-Jeansbrightness
obtained pre-perihelion during 2014-06-28 19:07-20:05
temperature scale (T ) for further analysis.
and 2014-06-29 17:37-18:26, while the comet was at a B
heliocentric distance r = 1.42-1.43 AU and geocentric
H 3. NANC¸AYOHOBSERVATIONS
distance∆=1.96-1.97AU(thecometreachedperihelion
atrH =1.05AUon2014-08-28). TheCH3OHK =3−2 Production rates for the dominant volatile H2O are
rotationalbandnear251.9GHz(inALMAreceiverband required for the measurement of CH3OH mixing ratios
6)wasobservedonJune28andtheJ =7−6bandnear andthederivationofthecollisionalexcitationrates. The
338.5 GHz (in band 7) was observed on June 29. Atmo- strengthsofthe18cmOHlines(at1667and1665MHz)
sphericconditionswereexcellentthroughout(withzenith in comet C/2012 K1 (PanSTARRS) were monitored us-
PWV < 0.4 mm). Thirty 12-m antennae, with baseline ing the Nan¸cay radio telescope during the period 2014-
lengths 20-650 m, resulted in an angular resolution of 04-03 to 2014-09-27. The observational procedure and
0.80(cid:48)(cid:48)×0.43(cid:48)(cid:48) at 252 GHz and 0.71(cid:48)(cid:48)×0.33(cid:48)(cid:48) at 338 GHz. dataanalysisweredescribedpreviouslybyCrovisieretal.
These beam dimensions correspond to 1100 × 610 km (2002), and an expansion velocity of 0.70 kms−1 was
and 1000×470 km, respectively, at the distance of the retrieved following the analysis method of Tseng et al.
comet and the measured RMS noise levels per channel (2007). The resulting OH production rates (Q(OH),
were 1.8 mJybm−1 and 3.0 mJybm−1. The correlator for the period 2014-06-14 to 2014-07-17) are shown as a
wasconfiguredtosimultaneouslyobserveasmanystrong function of heliocentric distance (pre-perihelion) in Fig-
CH OH lines as possible (spanning different upper-state ure 1, including error bars due to thermal noise. The
3
energylevels)inasingle976kHzspectralwindow,using OH line inversion parameter was large for these obser-
a spectral resolution of 488 kHz (0.58 kms−1 in band 6 vations, therefore errors on the OH production rate due
to excitation uncertainties are expected to be negligible.
and 0.43 kms−1 in band 7).
There is no clear trend in Q(OH) vs. r around the
The observing sequence on each date consisted of an H
time of our ALMA observations, so we take the mean
interleavedseriesofscansofthesciencetargetandacon-
value of 9.0 × 1028 s−1 (with a standard deviation of
tinuumsource(quasar)forphasecalibration,alternating
1.7×1028 s−1). AssumingH OisthesoleparentofOH,
between 7 minutes integration on the comet and 30 s on 2
andadoptingabranchingratioof0.86fortheH O+hν
the phase calibrator, for a total of 46 min on-source at 2
−→OH+Hphotolysischannel(Huebneretal.1992),we
252 GHz and 38 min on-source at 338 GHz. The comet
was tracked, and the position of the array phase center obtainawaterproductionrateof(1.05±0.20)×1029 s−1.
3
baselinesshorterthan20m,thelargestangularscalede-
TABLE 1
CH OH lines detected in C/2012 K1 (PanSTARRS) tectable in our observations is ≈ 7.4(cid:48)(cid:48) (or ≈ 104 km),
3
which results in filtering out of coma structures larger
ALMA Transition Freq. Eu Wp thanthissize. BasedonasimpleHaserparentmodelfor
Band (MHz) (K) Kkms−1
K1/PanSTARRS(seeSection4.5),spatialfilteringisex-
6 103−102 A−+ 251164 177.5 0.14(0.10) pected to have a negligible impact on the peak CH OH
93−92 A−+ 251360 154.3 0.54(0.10) flux (less than a few percent), whereas 2000 km (13.4(cid:48)(cid:48))
83−82 A−+ 251517 133.4 0.71(0.10)
73−72 A−+ 251642 114.8 0.62(0.10) from the nucleus, only ∼ 75% of the predicted flux per
63−62 A−+ 251738 98.5 0.47(0.10) beam may be recovered.
53−52 A−+ 251812 84.6 1.07(0.10) 4.2. Rotational diagrams
43−42 A−+ 251867 73.0 0.60(0.10)
53−52 A+− 251891 84.6 0.78(0.10) Detailed spectral modeling is required to interpret the
63−62 A+− 251896 98.5 1.04(0.10) observed multi-line CH3OH data. We begin by using
43−42 A−+ 251900 73.0 0.56(0.10) themethodoutlinedbyBockel´ee-Morvanetal.(1994)to
33−32 A−+ 251906 63.7 0.44(0.10) constructrotationalexcitationdiagramstoobtainthein-
33−32 A+− 251917 63.7 0.38(0.10) ternal(rotational)temperatureoftheCH OHmolecules
73−72 A+− 251924 114.8 0.84(0.10) (T ), averaged along the line of sight. F3igure 4 shows
83−82 A+− 251985 133.4 0.68(0.10) rotraottionaldiagramsforJune28(left)andJune29(right),
7 70−60 E 338124 78.1 0.65(0.11) obtained from spectra extracted at the CH OH column
7−1−6−1 E 338345 70.6 1.08(0.11) density peak. The integrated line intensities3(along with
70−60 A+ 338409 65.0 1.22(0.11)
their statistical 1σ errors) for this position are given in
7−4−6−4 E 338504 152.9 0.50(0.11)
74−64 A− 338513 145.3 1.46(0.11) Table 1. The gradient of the rotational diagram is equal
74−64 A+ 338513 145.3 B to −1/Trot and the intercept is equal to ln(N/Q) where
72−62 A− 338513 102.7 B Q is the partition function and N is the column den-
74−64 E 338530 161.0 0.32(0.11) sity. Errors on Trot and N were obtained from the er-
73−63 A+ 338541 114.8 1.28(0.14) rors on the gradient and intercept, resulting in values of
73−63 A− 338543 114.8 B Trot =63.0±6.8K,N =(2.8±0.7)×1014 cm−2 onJune
7−3−6−3 E 338560 127.7 0.39(0.12) 28 and T =119±23 K, N =(2.7±1.0)×1014 cm−2
73−63 E 338583 112.7 0.30(0.12) on Juner2ot9. No significant differences were detected
71−61 E 338615 86.1 0.83(0.12) between the abundances of the A and E nuclear spin
72−62 A+ 338640 102.7 0.64(0.12)
states of CH OH, consistent with an equilibrium (high-
72−62 E 338722 87.3 1.63(0.14) 3
7−2−6−2 E 338723 90.9 B temperature) spin distribution as observed in comet
Hale-Bopp by Pardanaud et al. (2007). Thus, we as-
Note — Integrated line intensities Wp are given for the
CH3OH column density peak, with 1σ statistical errors in sumedequalabundancesofAandE CH3OHinouranal-
parentheses. B denotes transitions that are blended with ysis from here on.
thelineabove. Unfortunately, the rotational diagram method suffers
from several shortcomings: blended lines with differing
upper-state energies (E ), and lines with signal-to-noise
u
< 1 cannot be included in the diagram, which limits
4. RESULTS
the accuracy of the results. Furthermore, the full fre-
4.1. Spectra and maps
quency extent of each line is not known a-priori, which
Forteen individual emission lines of the K = 3 − 2 hinderstheaccuracyoftheindividuallinemeasurements
band were detected on 28 June and sixteen lines of the — in determining the integration widths for weak lines,
J =7−6 band were detected on 29 June. The complete a balance has to be found between including the full line
listofdetectedCH OHtransitionsandtheirupper-state flux and excluding noisy data in the line wings, which
3
energies is given in Table 1, and the observed spectra inevitablyresultsinsomereductionintheoverallsignal-
are shown in Figure 2. These spectra were extracted to-noise for each line (for our data, we used integration
by averaging the data over a circular aperture 1.5(cid:48)(cid:48) in limitsof±1.5kms−1 foralllines). Inordertoobtainthe
diameter, centered on the peak of the CH3OH emis- most reliable temperatures and column densities, spec-
sion. Spectrally-integrated maps of the CH3OH emis- tral line modeling is employed to maximise the available
sion (shown in Figure 3) were obtained by integrating information in our ALMA data.
over the spectral ranges covered by the lines in Table
4.3. Line modeling
1. The CH OH emission peaks were approximately spa-
3
tiallycoincidentonbothdates, offsetby1.2(cid:48)(cid:48) south-west Using a modified version of the spectral line model-
fromtheALMAphasecenter,whichiswithinthetypical ingalgorithmpreviouslyappliedtointerstellarrotational
rangeofuncertaintyforthepositionofthenucleususing emission lines by Cordiner et al. (2013), the spectra at
optically-derived cometary ephemerides. No continuum eachpointinourCH OHmapswereextractedandfitted
3
emission was detected and a 3σ upper limit of 0.3 mJy (using a nonlinear least-squares algorithm) to determine
wasobtainedfortheaveragecontinuumflux(inband7). T and N as a function of the spatial coordinate. The
rot
TheCH OHmapsshowacompactdistributionwitha techniqueworksbyfittingGaussianopticaldepthcurves
3
strong central peak. The flux falls rapidly with distance toeachemissionline(asafunctionoffrequency),param-
from the center, which is consistent with the combined eterised by T , N and the line FWHM. The observed
rot
effectsof(isotropic)outflowexpansionandphotodissoci- CH OH line intensity ratios encode the rotational tem-
3
ationduetoSolarirradiation. DuetothelackofALMA perature,andtheabsoluteintensity(T )scalingencodes
B
4
0.4 0.7
K=3 2 0.6 J=7 6
0.3 − −
0.5
0.2 0.4
K) K) 0.3
(TB 0.1 (TB
0.2
0.0 0.1
0.0
0.1
0.1
251.2 251.4 251.6 251.8 252.0 338.0 338.2 338.4 338.6 338.8
Frequency (GHz) Frequency (GHz)
Fig. 2.—CH3OHspectraobservedon2014-06-28oftheK =3−2band(left)andon2014-06-29oftheJ =7−6band(right). Ticks
indicatethedetectedCH3OHlines. Thesespectrawereaveragedovera1.5(cid:48)(cid:48)-diametercircle(≈2100km)centeredontheemissionpeak.
K1/PanSTARRS 2014-06-28 CHOH K=3-2 252 GHz flux K1/PanSTARRS 2014-06-29 CHOH J=7-6 338 GHz flux
3000 3 3000 3 200
150
2000 2000
150
m) 1000 m) m) 1000 m)
North (k 0 100m / s / bea North (k 0 100m / s / bea
Distance -1000 50 Flux (mJy k Distance -1000 Flux (mJy k
50
-2000 T -2000 T
S S
-3000 0 -3000 0
-3000 -2000 -1000 0 1000 2000 3000 -3000 -2000 -1000 0 1000 2000 3000
Distance West (km) Distance West (km)
Fig. 3.— Spectrally-integrated ALMA flux maps of CH3OH in comet C/2012 K1 (PanSTARRS) observed on 2014-06-28 at 252 GHz
(left) and 2014-06-29 at 338 GHz (right), integrated over all detected transitions. Coordinate axes are aligned with the RA-dec. grid,
withcelestialnorthtowardsthetop. Whitecrossesindicatetheemissionpeaks, whichareemployedastheoriginofthecoordinateaxes.
Contours are plotted at 3σ intervals, where σ is the RMS noise in each map. The FWHM of the (Gaussian) restoring beams are shown
lower left. Direction of the Sun (S) and orbital trail (T) with respect to the comet are indicated lower right, along with the illumination
phaseangleof30◦.
the column density, so the best fit to each spectrum can convolved with a Gaussian broadening kernel of FWHM
be found for a unique pair of [T , N] values, as deter- 3 kms−1. This is sufficiently broad with respect to the
rot
minedbytheminimumofthesum-of-squaresoftheresid- FWHM of the observed lines (≈ 0.9−1.2 kms−1) that
uals between the observation and model (or χ2 value). it effectively results in the smoothing out of their pro-
ErrorestimateswereobtainedusingaMonteCarlonoise- files,producingalineshapepracticallyindistinguishable
resampling technique, whereby for each spectrum, 300 from Gaussian, and independent of the specific outflow
synthetic, Gaussian, random noise spectra were gener- geometry of the comet. This smoothing also has the
ated(withstandarddeviationequaltotheRMSnoiseof benefit that any small variations in the line FWHM and
theobservedspectrum), whichwerethenaddedoneata Doppler shift of the gas over the ALMA field of view
timetothe(noise-free)best-fittingmodelspectrum. The ((cid:46) ±0.25 kms−1) can be neglected, allowing these val-
same least-squares fitting procedure was then repeated uestobeheldfixedduringthefittingtofurtherimprove
for each of the 300 synthetic datasets. The 1σ errors on the accuracy of the results, which is particularly useful
T , N were obtained from the ±68% percentiles of the for the noisier data towards the edge of the field of view.
rot
resulting set of fit parameters. Examples of the line modeling results are shown in
This procedure assumes a Gaussian shape for the the lower panels of Figure 4, obtained using spectra ex-
emission lines, which can be a poor approximation for tracted at the CH OH column density peak position.
3
cometary lines observed at high spectral resolution — The quality of these fits is very good (with reduced χ2
indeed, many of our observed CH OH line profiles show values in the range 1.0-1.3), and the T and N val-
3 rot
slight asymmetries (consistent with asymmetric out- ues derived using this method (T =62.3±5.3 K, N =
rot
gassing). To avoid problems with the fitting due to (2.8±0.1)×1014 cm−2 onJune28andT =116±18K,
rot
such non-Gaussianity, each extracted spectrum was first
5
26.5 26.0
CH3OH J=7−6 data (2014-06-28) CH3OH K=3−2 data (2014-06-29)
Trot=63.0±6.8 K N=(2.8±0.7)×1014 cm−2 Trot=119.3±23.3 K N=(2.7±1.0)×1014 cm−2
26.0 25.5
25.5 25.0
/g)u /g)u
Nu Nu
n( n(
l25.0 l24.5
24.5 24.0
24.0 23.5
60 80 100 120 140 160 180 60 80 100 120 140 160 180
Eu (K) Eu (K)
00..45 CMHo3dOelH: TKro=t=36−22. 3s±pe5.c3t rKu m N (=20(21.48-±006.-12)8×)1014 cm−2 0.5 CMHo3dOelH: TJr=ot=7−161 6s±pe18c tKru m N (=2(021.74±-006.4-)2×91)014 cm−2
0.4
0.3
0.3
(K)TB 0.2 (K)TB 0.2
0.1 0.1
0.0 0.0
0.1
0.1
251.7 251.8 251.9 252.0 338.3 338.4 338.5 338.6 338.7
Frequency (GHz) Frequency (GHz)
Fig. 4.— Top: Rotational diagrams for K1/PanSTARRS CH3OH, constructed using spectra obtained at the column density peaks for
theK=3−2transitions(left)andtheJ =7−6transitions(right). Bottom: Least-squaresmodelfitstotheobserved(smoothed)spectra;
forclarity,onlyaportionoftheentirespectralregionisshownforeachdate.
N =(2.7±0.4)×1014cm−2onJune29)areingoodagree- gins at the respective column density peaks, which are
ment with those obtained using the rotational diagram marked with white crosses.
method. Due to improved utilization of the noisier data, TheCH OHcolumndensitymapsshowasingle,dom-
3
this line modeling technique results in significantly im- inant peak as expected for spherically-symmetric out-
proved accuracy of the derived parameters, and is there- gassing from a compact nucleus. The shapes of the col-
fore adopted as the preferred method of analysis from umn density peaks are strongly influenced by the ellip-
here on. The uncertainties derived from Monte Carlo tical telescope beam, which hinders interpretation of the
noise replication are also expected to be more reliable true structure of the innermost coma. On both dates,
because they implicitly account for correlations between the column density maps show a significant extension
the errors on T and N, whereas in the rotational di- towards the top left, which is in addition to the domi-
rot
agram analysis, the (partially correlated) errors on the nant outflow structure. The direction of this extension
gradient and intercept cannot be easily disentangled. coincidesapproximatelywiththedirectionofthecomet’s
orbitaltrail,perhapsimplyingsomeCH OHreleasefrom
3
4.4. Temperature and column density maps trailing material, although the presence of a directional
Spectrawereextractedpixel-by-pixelfromtheCH OH CH OH jet/vent cannot be ruled out.
3 3
datacubeswithinanareaof2(cid:48)(cid:48)×2(cid:48)(cid:48) (40×40pixels)cen- Interestingly, whereas on June 29 the T distribution
rot
teredontheintegratedemissionpeak. Usingthelinefit- shows a strong central peak about the nominal position
ting method described in Section 4.3, temperatures and of the nucleus (falling from 116±18 K to ≈ 50±10 K
column densities were obtained for each pixel to create within about 500 km), on June 28 the main source of
the maps shown in Figure 5. These maps only show the CH OH appears to be located within a temperature
3
derivedT andN valueswithuncertaintiesoflessthan trough, which forms part of a much broader region of
rot
50% — values with larger errors have been masked (and high temperatures that extends down the whole map in
are shown in white). The mean errors on these T and a roughly north-south direction. The central temper-
rot
N maps are ±14 K and ±1.8×1013 cm−2 for June 28, ature on June 28 was 62 ± 5 K and the two flanking
±15 K and ±1.7×1013 cm−2 for June 29. Naturally, a temperature maxima have Trot ≈ 85±14 K on the left
higher CH OH column density leads to a stronger spec- and ≈ 73±12 K on the right. However, the statistical
3
trum and hence lower uncertainties on T and N, so significanceofthesepeaksrelativetothecentraltemper-
rot
the regions closer to the centers of the maps are more ature is low (∼1-2σ), so a monotonically decreasing (or
reliable. The maps have been positioned with their ori- flat)temperatureasafunctiondistancefromthenucleus
6
K1/PanSTARRS CH3OH Column Density Map (2014-06-28) 10 0 0K1/PanSTARRS CH3OH Column Density Map (2014-06-29)
2.7 2.7
1000
2.4 500 2.4
500 2.12m)− 2.12m)−
c c
Distance (km) 0 11..5814n density ( 10 Distance (km) 5000 11..5814n density ( 10
m m
u 1.2u
500 1.2ol ol
C C
0.9
0.9 1000
1000 0.6
0.6
1000 500 0 500 1000 1000 500 0 500 1000
Distance (km) Distance (km)
K1/PanSTARRS CH3OH Temperature Map (2014-06-28) 1000 K1/PanSTARRS CH3OH Temperature Map (2014-06-29)
120
120
1000
105
105
500
90
500 90
Distance (km) 0 6705 emperature (K) Distance (km) 5000 6705 emperature (K)
45 T T
45
500
30
1000 30
1000 15
15
1000 500 0 500 1000 1000 500 0 500 1000
Distance (km) Distance (km)
Fig. 5.— Column density maps (top) and rotational temperature maps (bottom) for K1/PanSTARRS CH3OH, derived from pixel-by-
pixelspectrallinemodeling. CoordinateaxesarealignedasinFigure3,withcelestialnorthtowardsthetop. Leftpanelsshowtheresults
for the K = 3−2 transitions and right panels are for the J = 7−6 transitions. Values with an uncertainty of greater than 50% have
been masked (shown as white pixels). White crosses (‘+’) mark the column density peak and the hatched ellipses show the instrumental
resolution. The0.05(cid:48)(cid:48) pixelscalecorrespondsto71kmatthedistanceofthecomet.
in this region cannot be ruled out. Similarly, although The 1-D radial temperature profiles show a general
some of the temperature structure towards the bottom trend for falling T as a function of distance. Similar
rot
left of the central peak on June 29 could be real, the un- to the 2-D maps, on June 28 T shows a possible (1σ
rot
certainties preclude a robust interpretation of the other confidence) increase from 62±5 K to 68±6 K between
features in these maps. ρ = 0-500 km, followed by a relatively slow, steady de-
crease to 22±9 K at ρ=2500 km. Conversely, on June
29 the temperature decreased sharply from 116±18 K
4.5. Radial profiles
to42±4Kbetweenρ=0-1000km. Inadditiontostart-
Tohelpaverageoutanyclumpinessorsmall-scalecoma ing at least 30 K hotter near the nucleus, overall, the
structures and provide a clearer picture of the dominant CH OH rotational temperatures remained at least 20 K
3
physicalandchemicalprocessesasafunctionofdistance higher throughout the observed coma on the 29th than
fromthenucleus, weperformazimuthalaveragingofthe the 28th, implying the presence of a substantial addi-
ALMA data. Such azimuthally averaged data also ben- tional coma heat source on the 29th (22.5 hr later).
efit from a significantly improved signal-to-noise ratio, The possible temperature increase at distances ρ >
particularly at large radii, allowing us to probe the con- 1500 km on June 29 identified by Cordiner et al. (2016)
ditions further from the nucleus than in the 2-D maps. islessclearinournewresults;thelargeerrorbarsonour
ouTradkaintagtwheereCHbi3nOnHedcionltuomansdeerniessityofpe0a.1k(cid:48)(cid:48)a-wsitdheeacennntuelri,, Tharovtepbreoefinleaattleρas>t2p0a0rt0lykmduiendtoicaanteutnhdaetrtehstisimreastueltofmtahye
and the average flux per spectral channel was taken in errorsbyCordineretal.(2016),whousedarotationaldi-
each annulus, resulting in azimuthally-averaged spectra agram method rather than spectral line fitting. Further-
(S¯ν)asafunctionofsky-projectedradiusρ. TheseS¯ν(ρ) more, the azimuthal profiles in the present study were
were subject to the same fitting procedure outlined in taken about the CH OH column density peak, whereas
3
Section 4.3 to derive radial [Trot, N] profiles, shown in thoseofCordineretal.(2016)weretakenabouttheinte-
Figure 6.
7
140
100 CH3OH K=3-2 Rotational Temperature (2014-06-28) CH3OH J=7-6 Rotational Temperature (2014-06-29)
Temperature Model 120 Temperature Model
80
(K)Trot 60 (K)Trot10800
40
60
20 40
3.0 3.5
2.5 CH3OH Observed Column Density 3.0 CH3OH Observed Column Density
CH3OH Isotropic Outflow Model CH3OH Isotropic Outflow Model
142N ( cm)10−112...050 142N ( cm)10−1122....0505
0.5 0.5
0.0 0.0
0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500
Cometocentric distance (km) Cometocentric distance (km)
Fig. 6.— Azimuthally-averaged CH3OH rotational temperatures (upper panels) and column densities (lower panels), taken about the
column density peaks labeled in Figure 5. The K = 3−2 transitions are shown (left) and the J = 7−6 transitions (right). Simulated
(Haser-type) isotropic outflow model column density profiles are overlaid with red points, corresponding to CH3OH production rates of
1.9×1027 s−1 on June 28 and 1.4×1027 s−1 on June 29. Dashed lines in the upper panels indicate the rotational temperature profiles
obtainedfromradiativetransfermodeling,usingthekinetictemperaturedistributionsshowninFigure7(solidblacklines).
gratedemissionlinepeak;thesepositionswerecoincident inner coma. Slight discrepancies may be explained by
on June 28, but on June 29 the column density peak is the presence of clumpy structure, jets or variations in
offset north-east by 0.16(cid:48)(cid:48), leading to differences in the the outflow velocity. Some acceleration of the coma gas
radial temperature profiles. The elliptical shape of the within our ALMA field of view is plausible. Indeed, the
telescope beam introduces some additional uncertainty derived CH OH expansion velocity of 0.5 kms−1 in our
3
into our radial temperature and column density profiles, ∼ 2000 km field of interest is smaller than the value of
which would be particularly problematic for any asym- 0.7 kms−1 obtained for OH within a ∼105 km beam, as
metric coma features that happen to line up with the expected if the outflow speed increased with cometocen-
long axis of the beam. The (long-axis) spatial resolution tric distance. This possibility may be explored through
HWHM of 550 km on June 28 and 500 km on June 29 detailed analysis of the CH OH outflow dynamics in a
3
should thus be considered as upper limits for the pos- future study.
sible radial error margins on these azimuthally-averaged
5. DISCUSSION
profiles.
On both dates the azimuthally-averaged column den- In the range ρ = 0-2000 km, our ALMA observa-
sities decreased smoothly with distance from the peak tionsindicate,asexpected,ageneraltrendfordecreasing
value of about 2.8 × 1014 cm−2, with a shape closely CH3OH rotational temperatures as a function of come-
consistent with uniform, spherically-symmetric outflow. tocentric distance in the coma. To interpret this result
This is confirmed by comparison of the observed N(ρ) requires an understanding of the meaning of Trot. As
profiles in Figure 6 with profiles obtained from a Haser- explained by Bockel´ee-Morvan et al. (1994, 2004), the
type spherical outflow model, in which CH3OH is as- temperature Trot obtained from the analysis of CH3OH
sumed to flow isotropically from the nucleus at a con- rotational line ratios is representative of the distribu-
stant velocity of 0.5 kms−1 (consistent with the average tion of internal rotational level populations. Unless the
CH OH line Doppler FWHM of ≈ 1.0 kms−1), and is moleculesareinlocalthermodynamicequilibrium(LTE,
3
photodissociated at a rate of 10−5 s−1 (Huebner et al. whichoccursinthedense,innercomathroughcollisional
equipartition), T candeviatesignificantlyfromtheki-
1992). To match the observational point spread func- rot
netictemperatureofthegas(T ). Therelationshipbe-
tions, the model column density maps were run through kin
tweenT andT dependsonthecompetinginfluences
the ALMA simulator in CASA with array configuration, kin rot
of microscopic collisional and radiative processes. As a
imaging and cleaning parameters matching those of our
result, the prevalence of LTE depends primarily on the
observationions. Good fits to the observed N(ρ) pro-
comadensity,thespontaneousradiativedecayrateofthe
files were obtained for CH OH production rates of 1.9×
1027 s−1 onJune28and1.34×1027 s−1 onJune29. Using gas and the flux of external (Solar) radiation. Within a
theH Oproductionrateof(1.05±0.20)×1029 s−1 (sec- fewhundredkilometersofthenucleuswheredensitiesare
2 high, the collisional rates between molecules are usually
tion 3), these results imply that CH OH was sublimated
3 sufficienttomaintainLTE,butasthemoleculesflowout
directlyfromthenucleusofcometK1/PanSTARRSwith
intothelessdense,outercomaregions(atr (cid:38)1000km),
mixing ratios of 1.1-2.2%. c
collisions become less frequent, allowing the rotational
Our model column density profiles (red dots overlaid
levels to radiatively cool so that T falls below T .
onthebluecurvesinFigure6)provideanexcellentfitto rot kin
Pumping of molecular ro-vibrational levels by Solar ra-
the observations on June 28, and a reasonably good fit
diationcanalsohaveanimportanteffectonT atlarge
on June 29. As with the study of Cordiner et al. (2014), rot
distances(beyondthoseconsideredinthepresentstudy).
this confirms that the Haser model can be usefully ap-
UsingtheradiativetransfermodelofBiveretal.(1999,
pliedtoazimuthally-averagedALMAobservationsofthe
2015), which computes the CH OH level populations
3
8
that result from collisions with H O and electrons, and as a result of adiabatic cooling. Around a cometocentric
2
pumping by solar radiation, we calculated the departure distance r ∼ 100 km, the onset of photoionisation due
c
of T from T with distance from the nucleus. Using to Solar irradiation results in the production of a pop-
rot kin
dashed line styles, Figure 7 shows T as a function of ulation of fast (hot) electrons, ions and neutrals. Col-
rot
projected distance (ρ) for the J = 7−6 and K = 3−2 lisions with these hot photoproducts heats the parent
bands,undertheassumptionofconstantkinetictemper- gases. However, the strong heating trends apparent in
atures of 60 K (left panel) and 120 K (right panel). The these conventional multi-fluid hydrodynamic models —
H O production rate was taken to be 1029 s−1 and out- which predict temperatures above a few hundred kelvins
2
flow velocity 0.5 kms−1, with a collisional cross section for cometocentric distances > 1000 km — primarily re-
between H O and CH OH of 5×10−14 cm−2. In this flects the fact that the neutral gas temperature is ob-
2 3
figure, T (K = 3−2) and T (J = 7−6) both show tained by calculating the weighted average temperature
rot rot
increasingdeparturesfromLTEwithincreasingdistance over all coma neutrals, including the fast-moving photo-
from the nucleus. However, whereas the K =3−2 band products(primarilyO,H,H2andOHfromthephotolysis
remainsrelativelyclosetoLTE(withTrot >0.75Tkin for of H2O). Because the kinetic temperatures of these pho-
ρ < 2500 km), for the J = 7−6 band, T falls much toproducts can reach thousands of kelvins in the outer
rot
rapidlywith distance, reaching0.25T by ρ=2500km coma, their presence creates a strong bias in the average
kin
in the 120 K model. This is due to the larger Einstein temperatureofneutrals,whichisthennolongerproperly
A coefficients of the J = 7−6 transitions, which result representativeofthetemperatureofparentgases. Toac-
in more rapid rotational cooling. Although this trend is count for this problem, the temperatures of the parents
qualitatively similar to the observed T pattern for the and photoproducts must be treated separately, as in the
rot
two CH OH bands in Figure 6, the observed T curves direct simulation Monte Carlo (DSMC) models of e.g.
3 rot
both fall significantly more rapidly than expected with Tenishev et al. (2008) and Fougere et al. (2012). These
a coma kinetic temperature that remains constant as a models show a clear separation in the kinetic tempera-
function of distance. Sub-thermal excitation is therefore tures of parents and photoproducts and confirm the ex-
insufficient to fully explain the observed T behaviour pectationthatparentspeciesaredominatedbyadiabatic
rot
incometK1/PanSTARRS,andavariableT (r)profile cooling for distances (cid:46)105 km.
kin
is required. However, standard DSMC models still predict a very
Avarietyofdifferentkinetictemperatureprofileswere rapid drop in the temperature of parent molecules with
testedinourradiativetransfermodel,withtheaimofre- distance that is inconsistent with our ALMA CH3OH
producing the general behaviour of the observed T (r) observations. For example, in the model for comet
rot
profiles. For the K =3−2 profile on June 28, a good fit 67P/Churyumov-GerasimenkobyTenishevetal.(2008),
totheobservedTrot(r)wasobtainedusingasingle-slope the H2O temperature falls by an order of magnitude
T (r)profile,startingat90Katthenucleusandfalling within 100 km of the nucleus, whereas on June 28,
kin
to 35 K at rc = 2500 km (shown by the solid black line Tkin(r) in K1/PanSTARRS fell by only a few percent
in the left panel of Figure 7). For the J = 7−6 data over this range. To resolve this discrepancy, an addi-
from June 29, however, it was more difficult to obtain tional source of coma heating is required in the mod-
a good fit to the measured T (r) data. Even with a els. Fougere et al. (2012) considered the sublimation of
rot
rapidlyfallingTkin(r)profileintheinnercoma, thevery (dirty) ice grains, which can significantly raise the H2O
steep initial drop in Trot(r) could not be accurately re- rotational temperature in the region rc (cid:46) 1000 km pro-
produced. This is partly due to line-of-sight averaging, vided their mass represents a significant fraction of the
becausecooler,moredistantpartsofthecomacandomi- totalgas-phaseH2Oproduction. Suchsublimationheat-
natethetemperaturecontributionfromacompact,warm ing could also raise the CH3OH rotational temperature,
inner region (combined with the need to keep the inner either through injection of a source of heated CH3OH
coma temperatures physically reasonable). Our best fit into the coma, or by collisions with the heated H2O
to the T (J = 7 − 6) profile was obtained using the molecules. Other possible heating sources that may be
rot
T (r)showninblackintherightpanelofFigure7,be- considered in future models include suprathermal elec-
kin
ginningatarelativelyhightemperatureof150K,falling trons and ions that could be produced through interac-
to 40 K at r = 1000 km, then rising back to 150 K tion of the coma with the solar wind, UV and X-rays.
c
by r = 2500 km. Difficulties in modeling the T (r) The temperature trend in K1/PanSTARRS is com-
c rot
profile on June 29 may be due to coma asymmetries or patible with the results of other observational studies.
other physical factors not included in our model. Fu- The steadily decreasing Tkin behaviour found on June
ture attempts to more robustly derive the coma kinetic 28 is comparable to the linear temperature slope de-
temperatures from these data may require modeling the duced for the inner few thousand kilometers of comet
coma structure in 3-D, as well as exploring the effects of P1/Garrad (Bockel´ee-Morvan et al. 2012). Further, on
variations in electron density profiles and temperatures, June 29 the falling kinetic temperature in the inner
and gas collisional cross sections. coma, rising back to ∼ 150 K at larger radii is quali-
Compared with previous generations of chemi- tatively similar to that study. Our observations on both
cal/hydrodynamic coma models (see Rodgers et al. dates are also broadly consistent with the general trend
2004 for a review), the temperature trends observed in for decreasing H2O rotational temperatures with dis-
K1/PanSTARRS differ from past theoretical expecta- tance over scales ρ ∼ 10−1000 km in comets C/2004
tions. In those models the neutral gas temperature falls Q2 (Machholz) and 73P-B/Schwassmann-Wachmann 3
rapidlyasittravelsoutwardfromthenucleus,startingat (Bonev et al. 2007, 2008). Comets 103P/Hartley 2 and
∼100 K and reaching a minimum of ∼10 K by 100 km C/2012 S1 (ISON) showed a more complex temperature
behavior. At r = 0.53 AU, comet ISON’s T fell
H rot
9
������ ������������������������������������������������� �������� ��������������������������������������������������
��� ����
������ ��� ������ ����
�� ��
��� ���
��� ���
��� ���
��� ���
�� ���� ����� ����� ����� ����� �� ���� ����� ����� ����� �����
��������������������������� ���������������������������
Fig. 7.— Dashed green and purple lines show simulated CH3OH rotational temperatures as a function of distance from the nucleus
(using the excitation model of Biver et al. (2015)), for the K =3−2 and J =7−6 transitions respectively, assuming a constant coma
kinetic temperature of T =60 K (left panel) and 120 K (right panel). Whereas the K =3−2 levels remain quite close to the kinetic
kin
temperature,theJ =7−6levelsquicklybecomeverysub-thermalduetofasterradiativecooling. Thesolidblacklinesshowourbest-fitting
T (r)profilesonJune28(left)andJune29(right).
kin
from ∼120 K to ∼85 K within 1000 km of the nucelus, lation between water production rate and coma heating
whereas at r =0.35 AU, evidence for a double-peaked efficiency(seeBockelee-Morvan&Crovisier1987;Combi
H
temperaturestructurewasobserved,risingtomaximaat et al. 2004). However, the relative constancy of the
cometocentric distances ∼ 500−1000 km (Bonev et al. CH OHproductionrateduringourobservationsappears
3
2014). For 103P/Hartley 2, T was observed to de- to rule out this possibility in K1/PanSTARRS, and the
rot
crease with distance on the sunward side of the nucleus, full explanation for such strong, transient variations in
whereas on the anti-sunward side (as projected on the coma heating requires further investigation.
sky plane at a phase angle of 54◦), evidence was found Severalreviewstudieshavedrawncomparisonbetween
for a significant increase in T between ρ = 0-75 km interstellar, protostellar and cometary ice abundances
rot
(Bonev et al. 2013). This T behavior is analogous to (e.g. Ehrenfreund & Charnley 2000; Mumma & Charn-
rot
thepossibledouble-peakedtemperaturestructureidenti- ley 2011; Boogert et al. 2015). To-date, every coma
fiedinK1/Panstarrson28June(Figure5). Asdiscussed species detected in the radio has also been found in in-
by Bonev et al. (2014), it may be that such unexpected terstellar clouds, and the dominant cometary ice con-
temperature peaks in the coma can arise as a result of stituents(CO,CO andCH OH)showabundanceswith
2 3
photolytic or sublimation heating, but additional theo- respect to H O that are generally within the range of
2
reticalstudieswillberequiredtoconfirmthishypothesis. values observed in protostellar environments. Bockel´ee-
Thebehaviorofthecomatemperatureasafunctionof Morvan et al. (1994) identified the similarity between
time can provide more information on the nature of the cometary and interstellar CH OH/H O ice ratios, and
3 2
coma heating processes, and is possible because of the our values of 1.1-2.2% in C/2012 K1 (PanSTARRS),
22.5 hr separation between our K =3−2 and J =7−6 compared with 1-30% in low-mass protostars (Mumma
CH OH observations. Although our excitation model & Charnley 2011), confirm this result. Recent detailed
3
shows that T (J = 7−6) may still be somewhat sub- modelsfordiskgasandicechemistryconfirmtheplausi-
rot
thermal in the inner coma, the J =7−6 and K =3−2 bility of a close chemical relationship between cometary
bands are both expected to be close to the coma kinetic and protoplanetary material — e.g. Drozdovskaya et al.
temperature at around r = 0. The implied dramatic (2016) predict CH OH/H O ∼ 1-4% in the mid-plane
c 3 2
increase in coma kinetic temperature between June 28th ices for low-mass protoplanetary disks. The CH OH
3
and 29th is therefore surprising (given the expected rel- abundance in K1/PanSTARRS is also consistent with
ativeconstancyofSolarradiationinput),andislikelyto the detection of this molecule for the first time in a pro-
have been caused by an increase in the heating rate of toplanetary disk by Walsh et al. (2016), who obtained
the inner coma, for instance, due to an increase in the a gas-phase CH OH/H O ratio ∼ 1-5% in the disk sur-
3 2
supply of hot electrons or sublimating icy grains. If the roundingthelow-massTWHyasystem(atadistanceof
electron temperature dropped below the threshold for 54 pc). The fact that CH OH appears to be depleted in
3
dissociative electron impacts with H O, this could also comets and protoplanetarydisks compared with the me-
2
result in a sudden increase in the coma heating rate, as dian protostellar abundance of 6% in the nearby Galaxy
a larger fraction of the electron energy can then be con- (Boogert et al. 2015), implies that significant process-
vertedintokineticenergy(asdiscussedbyBodewitsetal. ing of interstellar/protostellar envelope material occurs
2016). Interactions with the Solar wind could also lead during or after its passage to the accretion disk, thus
toshort-timescalevariabilityinthecomaenergybalance. confirming the importance of cometary ices as a record
LargevariationsintheCH OHrotationaltemperature for the physical and chemical processes occurring during
3
were observed over timescales of several hours in comet the formation of the Solar System.
103P/Hartley2byDrahusetal.(2012). Thesevariations
were explained as due, in part, to the theorized corre-
10
6. CONCLUSION (using CH3OH or other molecules with a high density
of rotational lines in the mm/sub-mm such as H CO
Using ALMA observations of C/2012 K1 2
or CH CN), combined with detailed theoretical model-
(PanSTARRS), the first instantaneous spatial/spectral 3
ing, is necessary to provide new constraints on the coma
maps of CH OH rotational emission have been obtained
3 physicsandfurtherelucidatethegasheatingandcooling
in a cometary coma. Through rotational excitation
mechanisms. Continued ALMA observations (including
analysis,2-DspatialmapsoftheCH OHcolumndensity
3 observationsathigherangularresolutionandhighersen-
and rotational temperatures averaged along the line
sitivity), will therefore play a crucial role in improving
of sight have been derived, revealing new information
our knowledge of coma energetics, as well as leading to
on the physics and chemistry of the coma on scales
higheraccuracyincometarymolecularproductionrates,
500-5000 km. We find that the T (J = 7 − 6) and
rot parentscalelengthsandgasoutflowvelocities. Improved
T (K = 3−2) radial profiles both exhibit a relatively
rot accuracy and statistics in measurements of cometary ice
rapid drop with distance, which cannot be explained
abundances are key requirements for ongoing studies on
purely through sub-thermal excitation and must there-
theoriginandevolutionoficymaterialsinplanetarysys-
fore be due to falling coma kinetic temperatures with
tems.
distance from the nucleus. The observed temperature
behavior is more consistent with the DSMC model of
Fougere et al. 2012 (that includes coma heating from This work was supported by NASA’s Planetary At-
sublimating dirty ice grains), than the behavior seen mospheres and Planetary Astronomy Programs and by
in standard multi-fluid models, highlighting a need for the National Science Foundation under Grant No. AST-
continued research into coma heating (and cooling) 1616306. It makes use of the following ALMA data:
mechanisms. ADS/JAO.ALMA#2013.1.01061.S. ALMA is a partner-
The CH3OH radial column density profile is in good ship of ESO (representing its member states), NSF
agreement with spherically-symmetric, uniform outflow (USA) and NINS (Japan), together with NRC (Canada)
from the nucleus. Accordingly, no evidence is found and NSC and ASIAA (Taiwan), in cooperation with the
for significant production of CH3OH in the coma, ei- RepublicofChile. TheJointALMAObservatoryisoper-
ther from icy grain sublimation or photochemistry. The atedbyESO,AUI/NRAOandNAOJ.TheNationalRa-
CH3OH mixing ratios of 1.1-2.2% in K1/PanSTARRS dio Astronomy Observatory is a facility of the National
are consistent with previous observations of comets at Science Foundation operated under cooperative agree-
infrared and radio wavelengths. Combined with the ob- ment by Associated Universities, Inc. The Nanc¸ay Ra-
servationofCH3OHoutgassingdirectlyfromthenucleus, dioObservatoryistheUnit´escientifiquedeNan¸cayofthe
ourresultsconfirmtheutilityofradiointerferometricob- Observatoire de Paris, associated as USR No. B704 to
servations as a probe for the abundances of complex or- the CNRS. The Nan¸cay Observatory also gratefully ac-
ganicmoleculesincometaryice. OurCH3OHabundance knowledges the financial support of the Conseil r´egional
adds to the evidence confirming a close chemical simi- oftheR´egionCentreinFrance. Weacknowledgethead-
laritybetweenprotostellar/protoplanetaryandcometary viceofDr. BonchoBonevonthemeasurementofspatial
ices. variability in coma rotational temperatures.
The derivation of accurate coma kinetic temperatures Facilities: Atacama Large Millimeter/submillimeter
Array
REFERENCES
Biver,N.,Bockel´ee-Morvan,D.,Crovisier,J.,etal.1999,AJ,118, Combi,M.R.,Cochran,A.L.,Cochran,W.D.,Lambert,D.L.,
1850 &Johns-Krull,C.M.1999a,ApJ,512,961
Biver,N.,Bockelee-Morvan,D.,Moreno,R.,etal.2015,Science Combi,M.R.,Harris,W.M.,&Smyth,W.H.2004,CometsII,
Advances,1,e1500863 ed.G.W.Kronk,523
Bockelee-Morvan,D.,&Crovisier,J.1987,inESASpecial Combi,M.R.,Kabin,K.,DeZeeuw,D.L.,Gombosi,T.I.,&
Publication,Vol.278,DiversityandSimilarityofComets,ed. Powell,K.G.1999b,Earth,Moon,andPlanets,79,275
E.J.Rolfe&B.Battrick Cordiner,M.A.,Buckle,J.V.,Wirstr¨om,E.S.,Olofsson,
Bockel´ee-Morvan,D.,Crovisier,J.,Colom,P.,&Despois,D. A.O.H.,&Charnley,S.B.2013,ApJ,770,48
1994,A&A,287,647 Cordiner,M.A.,Remijan,A.J.,Boissier,J.,etal.2014,ApJ,
Bockel´ee-Morvan,D.,Crovisier,J.,Mumma,M.J.,&Weaver, 792,L2
H.A.2004,CometsII,ed.G.W.Kronk,391 Cordiner,M.A.,Charnley,S.B.,Mumma,M.J.,etal.2016,
Bockel´ee-Morvan,D.,Biver,N.,Swinyard,B.,etal.2012,A&A, AstronomyinFocus,inpress,arXiv:1602.03488
544,L15 Crovisier,J.,Colom,P.,G´erard,E.,Bockel´ee-Morvan,D.,&
Bodewits,D.,Lara,L.M.,A’Hearn,M.F.,etal.2016,AJ,152, Bourgois,G.2002,A&A,393,1053
130 Davidsson,B.J.R.,Sierks,H.,Gu¨ttler,C.,etal.2016,A&A,
Bonev,B.P.,Disanti,M.A.,Villanueva,G.,etal.2014,ApJL, 592,A63
796,L6 DiSanti,M.A.,Mumma,M.J.,Russo,N.D.,&Magee-Sauer,K.
Bonev,B.P.,Mumma,M.J.,Kawakita,H.,Kobayashi,H.,& 2001,Icarus,153,361
Villanueva,G.L.2008,Icarus,196,241 Drahus,M.,Jewitt,D.,Guilbert-Lepoutre,A.,Waniak,W.,&
Bonev,B.P.,Mumma,M.J.,Villanueva,G.L.,etal.2007,ApJ, Sievers,A.2012,ApJ,756,80
661,L97 Drozdovskaya,M.N.,Walsh,C.,vanDishoeck,E.F.,etal.2016,
Bonev,B.P.,Villanueva,G.L.,Paganini,L.,etal.2013,Icarus, MNRAS,462,977
222,740 Ehrenfreund,P.,&Charnley,S.B.2000,ARA&A,38,427
Boogert,A.C.A.,Gerakines,P.A.,&Whittet,D.C.B.2015, Fougere,N.,Combi,M.R.,Tenishev,V.,etal.2012,Icarus,221,
ARA&A,53,541 174
Cochran,A.L.,Levasseur-Regourd,A.-C.,Cordiner,M.,etal. Huebner,W.F.,Keady,J.J.,&Lyon,S.P.1992,Ap&SS,195,1
2015,SpaceSci.Rev.,197,9