Search for the 22 GHz water maser emission in selected comets

C.B. Cosmovici a,n, S. Pluchino b, S. Montebugnoli b, S. Pogrebenko c

a IAPS – INAF, Fosso del Cavaliere, 00133 Roma, Italyb IRA – INAF, Bologna, Italyc JIVE – Dwingeloo, The Netherlands

a r t i c l e i n f o

Article history:Received 3 April 2013Received in revised form24 February 2014Accepted 6 March 2014Available online 21 March 2014

Keywords:CometsWater maser emissionRadiospectrometry

a b s t r a c t

Following the first evidence of planetary water maser emission induced by the collision of cometShoemaker/Levy with Jupiter and the puzzling detection of the 22 GHz water emission line in CometHyakutake we started in the period 2002–2008 systematic observations of selected comets at 22 GHz(1.35 cm) with the aim of clarifying the unusual behavior of the maser line in the cometary “scenario”.Using a fast multichannel spectrometer coupled to the 32 m dish of the Medicina (Bologna, Italy) RadioTelescope we investigated 6 bright or sungrazing comets down to a heliocentric distance of 0.11 AU: 96P/Machholz, 153P/ Ikeya–Zhang, C/2002 V1 (NEAT), C/2002 X5 (Kudo–Fujikawa), C/2002 T7 (Linear),and 73P/Schwassmann–Wachmann 3. All of them, similarly to Comet Hyakutake, demonstrate spectralfeatures that, if real and due to the 1.35 cm water vapor transition, are strongly (up to tens of km/s)shifted relative to the radial velocity of the nucleus and, at least sometimes, seem to be present as twoseparate peaks. If our interpretation of these spectral peaks is correct, there must be some mechanism ofacceleration of neutral water molecules up to the velocities of ions. We discuss here the results achievedand the possible explanation of the chemo-physical constraints.

& 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Water was detected already often in comets at infrared wave-lengths starting with the detection in comet Halley (Mumma et al.,1986; Crovisier and Schloerb, 1991) and also jets of H2Oþ could bevisualized in the same comet by a special image processingtechnique (Cosmovici et al., 1995).

The detection of the 22 GHz emission line has been claimed inthe past for comet Bradfield 1974 III (Jackson et al., 1976) andcomet IRAS–Araki–Alcock 1983 VII (Altenhoff et al., 1983). Becauseof the low signal to noise ratio and because the 1.35 cm transitionhas been searched for but not confirmed in other bright cometslike Hale–Bopp (Bird et al., 1997; Graham et al., 2000), thesedetections have been questioned.

The water maser detection during the comet Shoemaker–Levy9/Jupiter event (Cosmovici et al., 1996) was induced by a cata-strophic impact in a planetary atmosphere, thus it cannot beconsidered a cometary emission.

The detection in comet Hyakutake (Cosmovici et al., 1998) wasthe first reliable detection despite the puzzling velocity offset andline splitting but it occurred under very special conditions: at adistance of only 0.23 AU from the Sun and during a strong coronal

activity observed by the LASCO coronograph C3 on the SOHOspacecraft.

The 22 GHz transition occurs between the rotational levels616–523 at 643 K above the zero point energy and may be invertedin some regions of the cometary coma under particular conditions.A very detailed investigation on the probabilities of detecting the22 GHz line in comets was published by Graham et al. (2000).They concluded from their calculations that the line could bedetected only if it were masing in a region observed along a jetagainst the radio continuum background of the nucleus.

As our detection in comet Hyakutake needed more investiga-tion in sungrazing comets because of the anomalous splitting ofthe line and its Doppler shift by a variable amount and with avariable separation we decided to observe, starting in 2002, allpossible sungrazing and very bright comets in order to check if theSun distance and the solar activity could explain the puzzlingbehavior of the water maser line.

2. The observations

The observations were carried out with the 32 m Medicina(Bologna, Italy) VLBI antenna equipped with a 22 GHz receiver. Thevery high frequency and time resolution, direct Fourier transformspectrometer (MSPEC0) was used (Montebugnoli et al., 1996) after

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/pss

Planetary and Space Science

http://dx.doi.org/10.1016/j.pss.2014.03.0040032-0633/& 2014 Elsevier Ltd. All rights reserved.

n Corresponding author. Tel.: +39 06 4993 4388; fax: +39 06 4548 8789.E-mail address: cosmo@iaps.inaf.it (C.B. Cosmovici).

Planetary and Space Science 96 (2014) 22–28

its successful performance during the Comet Shoemaker–Levyimpact on Jupiter (Cosmovici et al., 1996). It uses an extremelypowerful Digital Signal Processor (DSP optimized for the FastFourier Transform (FFT) computations and is able to performtransforms in a very short time: 1 k complex points need onlyabout 90 μs and 256 k complex points FFT about 20 ms. Thistremendous computation rate allows us to compute the FFT andthe square magnitude “on fly” over an input bandwidth of 8 MHz,with a duty cycle efficiency of 30–37%.With this instrument the8 MHz bandwidth was used with 8192 spectral channels whichpermitted observation of the high velocity shifted spectral linesreported in this work.

During the observations the telescope was pointed to thecenter of the comet and to the reference position (R.A. Cometþ1deg� cos elevation), thus enabling us to obtain differentialON–OFF spectra. The spectra were calibrated using as a standardnoise source generator. Secondary data reduction was done byusing the ASTRA (Pluchino, 2008) software package.

The ASTRA post-processing software performs an accurate andrigorous analysis of spectral data. This software package, after anumeric validation of the raw nodding on–off–cal data, extractseach spectrum block and header in which are stored fundamentalinformation about the related observing session (antenna andreceivers setup, meteo, etc.). ASTRA automatizes the normalizationand the calibration of a great number of spectra, performing onthem an accurate time–frequency analysis and an accumulationover the time by means of a selectable addition (stacking) ofindividual spectra. The data processing ‘core’ of this software toolis the ‘de-Doppler engine module’, which takes into account theknown Doppler shift of the target, i.e. the radial component of itsvelocity, makes the necessary correction and accumulates the finalspectrum. In the case of cometary observations, the softwareperforms a Doppler compensation taking the radial velocity as afunction of time from the JPL Horizons system.

3. The observed comets

We investigated 6 comets: two of them, Comet NEAT ( S/N: 7.20) and comet Kudo–Fujikawa( S/N: 7.13 ) present the most reliabledetection values. Four of them gave possible positive results withS/N values43–4 under acceptable weather and instrumentalconditions.The results of the observations are given in the follow-ing subheadings (I–VI), in Figs. 1–7 and in Table 1 where theordering is by distance from the Sun. It should be mentioned thatthe negative detection of water in the following 6 comets atdistances 0.13–2.0 AU from the Sun is here not analyzed as thedata were not usable because of very strong RFI, bad weatherconditions or malfunctions in the K-receiver during the short timeof observations.

The comets were M4 Swan; P1 McNaught; WM1 Linear; Q4Neat; 9P Tempel 1; and 17P Holmes. Period of observations: 2004–2007.

I) C/2002 V1 NEAT. It was the most interesting comet we couldstudy as we could carry out the Sun closest observation of acomet from the ground. Observations were made on February17,18 and 19, 2003 when the distance from the Sun wasR¼0.101–0.121 AU during the strong Coronal Mass Ejection(CME) detected by the SOHO spacecraft (Cosmovici et al., 2003)(see Fig. 1). The beam size was 20, corresponding to an observedregion at the comet of d¼80,000 km. As in the case of cometHyakutake the line is shifted toward positive velocities and itshows a puzzling double peaked feature separated by a velocitydifference of 4 km/s (see Fig. 2).

II) Comet 96P/Machholz. Also a sungrazing comet observed onJanuary 12, 2002 at R¼0.21 AU. The observed region aroundthe nucleus was d¼79,000 km (FOV¼20). The SOHO spacecrafthad a unique opportunity to observe the extended coma anddust tail structure at the extremely close perihelion transit(R¼0.1 AU). The water maser line appears shifted towardpositive velocities by 47.9 km/s. It has been calculated thatthe emitting source would cross the entire antenna beam inabout 27 min if the velocity in the plane of the sky is similar tothe observed radial velocity. This value would indicate that jetsand not the surrounding coma are the source of the wateremission. (The second peak at 30.8 km/s is not reliable asS/No3 (see Fig. 3A).)

III) C/2002 X5 Kudo–Fujikawa. Sungrazing comet observedJanuary 20, 2003 at R¼0.37 AU over a region around thenucleus d¼90,000 km (FOV¼20). The comet passed within0.2 AU from the Sun on January 29 and the UVCS images of theSOHO spacecraft could show a brightening of the comet atperihelion reflecting a fivefold increase in the comet's waterproduction rate (from Lyman alpha). In our observations thewater maser line is split in two components both shiftedtoward positive velocities, the first by 20.2 km/s and thesecond by 3.8 km/s. Velocity difference: 16.4 km/s (see Fig. 4).

IV) 153P/Ikeya–Zhang. Observations were made on March, 16,2002 at R¼0.51 AU over a region around nucleusd¼71,000 km. Despite the fact that it was not a sungrazingcomet we could detect the maser line, even if with lower SNR(3.3) with a positive velocity shift of 8 km/s. No split observed.(see Fig. 5).

V) C/2002 T7 (LINEAR). Observations on May 18 and 24, 2004at R¼0.82 AU over a region around nucleus d¼22,000 km.(distance from observer 0.27 AU). The possible detection ofthe water line shows two components with a positive velocityshift of 17.8 and 15.7 km/s. Velocity difference: 2.1 km/s(see Fig. 6).

Fig. 1. Comet NEAT as seen by the LASCO Coronograph C3 on SOHO, February 17–18, 2003. Credit: NASA/ESA.

C.B. Cosmovici et al. / Planetary and Space Science 96 (2014) 22–28 23

VI) 73P/ Schwassmann–Wachmann 3 (Fragment B1). Observa-tion on May 10, 2006 at R¼1.03 AU. Closest geocentricdistance of 0.07 AU over a region around nucleus d¼5700 km(FOV¼20). It was a spectacular comet which disintegrated inmany fragments and its passage was also the closest to Earthrising the possibility of carrying out observations with highspatial resolution and high intensity. We detected the waterline in fragment B1 which was particularly active showingfrequent outbursts and fragmentations which provided aunique opportunity to investigate fresh ices in the nucleus.The observed line is broad (105 kHz) may be due to two veryclose components and has a very high negative velocity shiftof �35 km/s. This is the first negative shift in our observationsof the maser line in comets (see Fig. 7).

4. Discussion

Our observations confirmed the anomalous double peaked velo-city shift of the neutral water line with respect to the comet restvelocity (zero point on the abscissa of Figs. 2–7) observed in CometHyakutake (Cosmovici et al., 1998). Observed outflow velocities ofneutral lines normally do not exceed 1 km/s whereas ion velocitiesrange normally between 10 and 30 km/s (Spinrad et al., 1994).

It was suggested by Graham et al. (2000) that our observedlines could belong to an unknown molecular ion detectable onlyvery close to the Sun. As laboratory data are not available formolecular ions at 1.35 cm wavelength we may exclude thepresence of such a “bright” ion as water constitutes more than50% of the cometary volatiles and the oscillator strength of watermolecules for our frequency transition is much higher than forother possible molecules and ions in our observational band.

Thus we have to take into account the explanations alreadylisted for comet Hyakutake.

A possible explanation for the observed high velocity shift ofthe water line is that a considerable amount of the watermolecules is produced from charged icy grains originated in jetsor in observed big chunks split tailward from the cometarynucleus, like in comet Schwassmann–W. 3 (see Fig. 7).

Hanner (1981) calculated the grain lifetimes in relation to theirradius, refraction index and expansion velocity at 1 AU from theSun and the equilibrium temperatures of amorphous ice grainsfrom 0.5 to 2.5 AU from the Sun. From their calculations in our casewe would need either pure large icy grains which could survive atleast 1000 s at 0.2–0.3 AU or very small (10 A) particles whichwould not absorb solar UV radiation. The observed double peakcould be the Doppler shift signature caused by different expandingwater-grain jets.

Another acceleration mechanism possibility could be that a largepart of the expanding coma recondenses into tiny ice droplets withina short distance from the nucleus surface via adiabatic cooling. Thiscould help “quench” the water line emission at the zero velocity.Small icy droplets of 10 A could survive the solar radiation and willbe consequently accelerated away with the sublimation productdetected at different velocities in the observed comets.

Comet NEAT (Figs. 1 and 2) was especially interesting as itpresented the most reliable data. The LASCO C3 coronographwitnessed a large CME ( Coronal Mass Ejection ) swipe throughthe tail of the comet as it flew through perihelion on 18 February2003 (see the LASCO C3 sungrazing comets website:http://lasco-www.nrl.navy.mil/lasco.html) The high energy solar windparticles in the CME may have contributed to the ionization andacceleration mechanism, thus enabling the observed water maseremission.

Fig. 2. Spectrum of Comet NEAT observed with the Medicina Radio Telescope.

C.B. Cosmovici et al. / Planetary and Space Science 96 (2014) 22–2824

5. Summary

Six selected comets were investigated in detail by means of asophisticated software (ASTRA) in order to verify the discoverymade on comet Hyakutake, i.e. the anomalous velocity shift of theneutral water molecules in sungrazing comets.

As shown in the figures the velocity shift is always present inthe six comets with different values depending on the distancefrom the Sun, on the coronal activity and on the position of the jetsemitting icy water grains.

The ionization mechanism here proposed, necessary for theacceleration of the neutral water molecules, should be investigatedin a dedicated theoretical work which is beyond the scope of thispaper. In the absence of other observations of the 22 GHz line forthe investigated comets, our unique results could be valuableinformation for studies on grain–plasma coupling in the cometaryscenario. In any case, because of the sometimes low signal-to-noiseratio, it should be pointed out that some of our detections should beconsidered as “possible” and their interpretation as just a workinghypotesis “if the detections are real”.

Fig. 3. Comet Machholz as seen by the LASCO Coronograph C3 on SOHO, January 8, 2002. Credit: NASA/ESA. (A) Spectrum of Comet 96P/Machholz.

C.B. Cosmovici et al. / Planetary and Space Science 96 (2014) 22–28 25

Fig. 4. Spectrum of Comet C/2002 X5 Kudo–Fujikawa.

Fig. 5. Spectrum of Comet 153P/Ikeya–Zhang.

C.B. Cosmovici et al. / Planetary and Space Science 96 (2014) 22–2826

Fig. 6. Spectrum of Comet C/2002 T7 (Linear).

Fig. 7. Spectrum of fragment B1 in Comet 73P/Schwassman–Wachmann 3.

C.B. Cosmovici et al. / Planetary and Space Science 96 (2014) 22–28 27

Acknowledgments

Special thanks to Massimo Teodorani, Simona Righini andGiuseppe Maccaferri for the great effort in carrying out most ofthe observations with the Medicina radiotelescope operated byINAF - Istituto di Radioastronomia. This work was financed andsupported by the Italian Space Agency ASI.

References

Altenhoff, W., et al., 1983. Radio observations of comet 1983d. Astron. Astrophys.125, Ll9–L22.

Bird, M.K., et al., 1997. K-band radio observations of comet Hale–Bopp: detectionsof ammonia and (possibly) water. Earth Moon Planets 78, 21–28.

Cosmovici, C.B., Schwarz, G., Ip, W.H., Fink, U., 1995. Imaging of H2Oþ in the comaof comet Halley. Planet. Space Sci. 43, 699–708.

Cosmovici, C.B., Montebugnoli, S., Orfei, A., Pogrebenko, S., Colom, P., 1996. Firstevidence of planetary water maser emission induced by the Comet/Jupitercatastrophic impact. Planet. Space Sci. 44, 735–739.

Cosmovici, C.B., Montebugnoli, S., Orfei, A., Pogrebenko, S., Cortiglioni, S., 1998. Thepuzzling detection of the 22 GHz water emission line in comet Hyakutake atperihelion. Planet. Space Sci. 46, 467–470.

Cosmovici ,C.B., Teodorani, M., Montebugnoli, S. and Maccaferri, G., 2003, CometNEAT, IAU Circular 8094.

Crovisier, J., Schloerb, F.P., 1991. In: Newburn, R.L., et al. (Eds.), Comets in the Post-Halley Era, vol. 1. Kluver, Dordrecht, The Netherlands, pp. 149–173.

Graham, A.P., et al., 2000w. Water maser emission from comets. Astron. J. 119, 2465.Hanner, M.S., 1981. On the detectability of icy grains in the comae of comets. Icarus

47, 342–350.Jackson, W.M., Clark, T., Donn, B., 1976. Radio detection of H2O in comet Bradfield

(1974b). The study of comets SP-393, 272–280.Montebugnoli, S., et al., 1996. A new high resolution digital spectrometer for

radioastronomy applications. Rev. Sci. Instrum. 67, 365.Mumma, M.J., et al., 1986. Detection of water vapour in comet Halley. Science 232,

1523.Pluchino, S., 2008. ASTRA: un software per il post-processing di dati spettrometrici.

INAF-IRA, Bologna, Italy (Internal report 422/08)Spinrad, H., Brown, M.E., Johns, C.M., 1994. Kinematics of the ion tail of comet

P/Swift–Tuttle. Astron. J. 108, 1462–1470.

Table 1Observational data.

Comet UTC (start) date d (AU) R (AU) Ta (mK) FWHM (kHz) Int. time (h) Q (mol/s�1029)

C/2002 V1 (NEAT) 2003/02/17; 09.02 0.98 0.11 33–36 33 1.14 31.496P/Machholz 2002/01/12; 09:41 0.99 0.21 27 54 1.41 8.6C/2002 X5 (Kudo–F.) 2003/01/20; 08:44 1.11 0.37 42–34 27 1.96 2.8153P/Ikeya–Zang 2002/03/16; 09:41 0.87 0.51 40 50 2.26 1.2C/2002 T7 (LINEAR) 2004/05/18; 07:53 0.27 0.82 52–58 29 2.32 0.673P/Schwassmann–W 2006/05/10; 00:22 0.07 1.03 41 105 3.96 3.5

d¼distance from Earth.R¼distance from Sun.Ta¼antenna temperature.Q¼production rate of water in molecules/s (assuming an emitting region with 1000 km diameter).

C.B. Cosmovici et al. / Planetary and Space Science 96 (2014) 22–2828