Observations of Comet Hale-Bopp

Comet 1995 O1 (Hale-Bopp ) was first reported during 1995 July while still well beyond the orbit of Jupiter. Although comets are notoriously fickle, coma activity in Hale-Bopp presently continues to be strong, suggesting that Marsden's prediction (1995), comparing it to the Great Comets of the 19th Century, remains valid. Hale-Bopp therefore seems to present an extremely rare opportunity, and the first such in the modern era of mm/submm telescopes, to investigate a bright long-period comet through all the stages of its development during its passage through the inner solar system.

We made the first mm-wave detection of Hale-Bopp with the JCMT in the CO 2-1 transition in early September 1995 when the comet was still at a distance of 6.7 a.u. from the Sun (Matthews et al , 1995). At this distance comets are too cold for the formation of the coma to be driven by the sublimation of water ice, and the detection of CO in Hale-Bopp was further vindication of the idea (suggested by observations of P/Schwassmann-Wachmann 1 by Senay & Jewitt, 1994 and Crovisier et al , 1995) that at such large distances the sublimation of more volatile ices such as molecular nitrogen or carbon monoxide drive the coma.

On the basis of these results, subsequently confirmed at IRAM (Rauer et al. 1995), and the exceptional promise of Hale-Bopp as a major comet, we submitted a request jointly to the CTAG and the UH TAG for observations covering four semesters beginning in February 1996. The Canadian program was awarded long-term status. The UH companion programs continue to gain time each semester. In the latter part of 1996 (and also throughout 1997) Hale-Bopp is largely a daytime object, and we have benefitted from a special arrangement by the Director, JCMT allowing such observations.

Observations so far

Using observations made during daytime `override' time in 1995 September through November we showed (Jewitt et al , 1996a; see also Biver et al , 1996a, and Weaver, 1996) that Hale-Bopp was undergoing an extremely rapid increase in its CO production rate Q(CO), such that the relationship Q(CO) proportional to R(-9.4) provided a good fit to the data. In retrospect it would appear that this was a surge in outgassing in response to increasing insolation, since subsequently the CO output dropped significantly before recovering. Some models suggest (eg: Prialnik, 1997) that such bursts are to be expected at the onset of early outgassing at large distances from the Sun, perhaps due to a runaway transition from amorphous to crystalline ice in the outer layer of the comet nucleus.

CO is easily released from the water-ice matrix, but its vaporization behaviour is quite different from that of other, more complex, trace constituents of cometary ices. One of the goals of our program was therefore to catch the onset of outgassing of different molecules. On 1996 April 8 we detected HCN 4-3 for the first time (Jewitt et al , 1996b), and subsequently CH3OH was detected (Womack et al , 1996; Biver et al , 1996b) in two different pairs of transitions.

We have continued to monitor both CO and HCN output intermittently as the telescope schedule, instrument availability, and weather has allowed. Except for the earliest work reported in Jewitt et al (1996a) we have used almost exclusively the CO 3-2 and HCN 4-3 transitions for this work. Recent CO 3-2 line strengths are about 0.8K Ta* (as compared to 0.1K Ta* at discovery in the CO 2-1 transition), while the HCN 4-3 line is now typically 3-4K Ta* . Some of our spectra illustrating the development of the line emission are shown below in Figure 1.

Figure 1: Spectra of CO (in green) and HCN (purple) observed on four epochs (from bottom to top, UT 1996 May 17, August 11, and November 30, and 1997 January 18). Spectra are offset vertically from one another for clarity, as appropriate, and aligned on a common velocity scale. The 1996 May 17 HCN spectrum is the discovery spectrum of HCN. The telluric CO line has been removed from the 1996 August 11 spectrum, resulting in the flat section of the spectrum. All observations were made in frequency-switched mode. The first three were obtained with the interim receiver B3i. The last was taken with the new receiver B3 using both polarization channels, and a double-sideband configuration which permits the simultaneous observation of CO 3-2 and HCN 4-3, now possible due to the change of the IF from 1.5 GHz with B3i to 4 GHz with B3.

In Table 1 we give outgassing rates for CO and HCN (Q(CO) and Q(HCN)) for each of the spectra in Figure 1. R and Delta are the comet-Sun and comet-Earth distances respectively, T(ex) the estimated excitation temperature, and V(exp) the coma expansion velocity.

UT Date         R   Delta  T(ex)  V(exp)  Q(CO)     Q(HCN)

               a.u.  a.u.   K     m/sec   mol/sec   mol/sec



17-May-1996   4.38   3.75   15    400   3.3 x 10(28)   9.9 x 10(25)
 
11-Aug-1996   3.48   2.74   20    470   5.6 x 10(28)   3.5 x 10(26)
 
30-Nov-1996   2.14   2.93   40    600   1.1 x 10(29)   1.4 x 10(27)

18-Jan-1997   1.53   2.26   45    750   1.3 x 10(29)   2.7 x 10(27)

As is apparent even in Figure 1 the original dramatic increase in CO outgassing has slowed considerably, presumably as the near-surface supply of CO is depleted. At the same time the HCN line has become very much stronger. As Hale-Bopp approaches the Sun there is also a major increase in the widths of both lines. This effect was also seen in Comet Hyakutake during the campaign in Spring 1996.

As shown in Figure 1 the more recent HCN (and to a lesser extent CO and CH3OH) spectra show the development of a central dip in the profile, as was also seen in Hyakutake in March 1996. Earlier spectra show clear asymmetry in the line shape which results from anisotropy in the emission of gas from the nucleus. Shown in Figure 2 is the result of a fit to one clearly assymetric line profile in which a free parameter is the ratio of outgassing rates from the day- and night-sides of the comet nucleus. Within the baseline uncertainties the data and the model agree rather well.

Figure 2: A model of the HCN 4-3 line compared with observations from 1996 November 30. This shows the fit of an outgassing model in which the ratio of day to night emission rates is 3:1.

The HCN 4-3 line in Hale-Bopp is so bright that it is possible to carry out mapping observations with very short integration times. A test observation of this type is shown in Figure 3; although somewhat of a curiosity at this point, having been taken under relatively poor conditions, it nevertheless shows that such observations are within our capability and should allow for detailed mapping of the coma out to several beamwidths.

Figure 3: HCN 4-3 map of Hale-Bopp obtained on UT 1997 January 18 using receiver B3 under marginal conditions. In this "on-the-fly" (raster) map the integration time was 5 seconds per point and the sampling interval 5 arcsec. The total time used to obtain this map was about 17 minutes.

It is possible to observe the HCN and HNC lines simultaneously with the new receiver B3. One example of these data is shown in Figure 4. The HNC 4-3 line was first detected in Hale-Bopp using B3i on UT 1996 December 1 (Matthews, Jewitt & Irvine, 1996), and, as in Hyakutake (see Irvine et al , 1996) the ratio R = [HNC]/[HCN] is typically 6% or 7% when optical depth effects are taken into consideration. This argues for a origin at a temperature of a few tens of degrees, if the molecular abundances are primordial, consistent with an interstellar origin. Since comets are believed to be fragments of the protosolar nebula it is important to attempt to rule out significant photo-processing of HNC during close approaches to the Sun by measuring R as a function of heliocentric distance. So long as the HCN line remains at present levels of a few K Ta* this would appear to be a straightforward task.

Figure 4: HCN (upper spectrum) and HNC 4-3 observed in Hale-Bopp on UT 1997 January 18. These data were taken with the new receiver B3 in single-channel mode, using frequency-switching. Both lines were observed simultaneously using a double-sideband configuration. The calibration of these particular data is not yet complete; since the ratio of HCN to HNC and its variation with heliocentric distance is a key question in settling whether the HNC abundance is primordial, this matter needs to be settled rather carefully.

Lessons from Comet 1996/B2 (Hyakutake)

Comet Hyakutake, discovered only in 1996 February, made a spectacular passage close by the Earth in March 1996 and provided an excellent "dress rehearsal" for the Hale-Bopp program. HCN and CO were both first detected (Matthews et al , 1996a; Senay et al , 1996b) at the JCMT in this object shortly after its discovery. Subsequently, in collaboration with other members of a JCMT target-of-opportunity consortium, we were able to map Hyakutake during close approach (about 0.1 a.u.) in the CO 3-2 and HCN 4-3 lines. We obtained the first detection ever in a comet of HNC, and the first sub-mm detection of CS in a comet (Matthews et al , 1996b). We were able to use two transitions of CH3OH to monitor the temperature of the coma, and we detected H2CO also. Finally, we obtained for the first time ever the mm/sub-mm dust spectrum and rudimentary mapping of a comet on two separate nights (Matthews et al , 1996c; Jewitt & Matthews, 1997).

There are some useful experiences gained from the Hyakutake campaign which can be applied to the next phases of our Hale-Bopp program. On the other hand, there are two major differences which should indicate caution in making such a comparison:

1. Hyakutake was a small comet (its nucleus less was than 3 km in size) whose close approach to Earth happily made mapping observations possible. Hale-Bopp is much larger (40 km, give or take a factor of two), but will never come closer to Earth than about 1.4 a.u. However, the considerable coma activity of Hale-Bopp is more than likely to compensate for the increased distance. Present line strengths in Hale-Bopp at more than 2 a.u. from the Earth are similar to those from Hyakutake at a geocentric distance of 0.2 a.u.

2. Hyakutake showed a wealth of mm/sub-mm spectral lines while in the vicinity of Earth's orbit, but once subjected to more extreme temperatures close to the Sun, most of the emission faded away. The molecules were being either pushed into higher excitation levels (with a corresponding increase in the partition function), or destroyed. In the case of Hale-Bopp there is the more favourable situation that it never gets closer to the Sun than about 0.9 a.u., and for this reason excitation temperatures should remain moderate, and Hale-Bopp should be a strong source of mm/submm spectral line emission throughout the course of the long-term program.

Continued observations of Hale-Bopp through 1997

During the early part of 1997, Hale-Bopp moves inbound from R = 1.4 a.u., through perihelion, to R = 2.0 a.u. Within this interval the predominant driver for coma formation is H2O (see for example Jewitt et al , 1996a).

Figure 5: Hale-Bopp in 1997: predicted H2O outgassing Q(H2O) and 850-micron continuum flux densities from the coma and nucleus, as a function of time. This figure uses recent outgassing rates, and assumes that the diameter of the nucleus is 40 km. Recent observations by Kreysa et al (1997) indicate that the flux density from the coma may be considerably greater than in this model.

In Figure 5 we show the predicted H2O sublimation rate as a function of time, together with the flux densities at 850 micron from the coma and nucleus. The behaviour of CO outgassing is perhaps less predictable, particularly after perihelion, since it will depend on the amount of CO in the surface layers of the nucleus.

For one thing it is to be expected that this model is simplistic in predicting smooth changes with time in the coma. A detailed model recently advanced by Prialnik (1997) makes specific predictions about the behaviour of CO, H2O and dust release rates from the nucleus of Hale-Bopp, which we should be able to test with our program of observations at the JCMT. In this model a runaway conversion of surface amorphous ice to a crystalline form drives a series of outbursts of CO (carrying dust to form the coma and to build a dust mantle on the surface of the nucleus) which begins when the inbound comet reaches heliocentric distances of around 6 or 7 a.u. That such outbursts mimic rather well that seen by Jewitt et al. (1996a) when Hale-Bopp was first detected lends support to Prialnik's model.

The amorphous/crystalline phase change boundary proceeds inward from the surface of the nucleus quickly at first, leading to a very porous outer layer. However, the process is self-limiting, and eventually the boundary essentially stalls at a few metres below the surface, by which time a porous dust mantle perhaps 10cm thick has formed and which serves to dramatically raise the temperature. From this time on, changes are more gentle; the CO (and dust) outgassing rate tends to flatten out. Beyond perihelion the CO output is predicted to remain fairly steady, and then slowly decline, at a somewhat lower level than the pre-perihelion rate, while the H2O outgassing continues to decline more rapidly. Around the same time the dust (and thus coma) output shows a sharp decrease. It is very likely that at this time a fairly thick and relatively impenetrable layer of crystalline ice has built up just below the surface of the nucleus.

The program proposed here consists of three main sections, aimed at testing models such as that discussed above. In this we take into account also our experiences in the Hyakutake campaign. We aim to (1) monitor the line emission characteristics of key constituents of the coma, (2) assess the chemical composition of the coma and its provenance via isotope ratios, and (3) determine the dust properties of the coma.

1. Monitoring key tracers of the physical and chemical state of Hale-Bopp.

HCN 4-3 (354.5 GHz);

CO 3-2 (345.8 GHz);

CH3OH and H2CO (to obtain temperature estimates)

With Hale-Bopp we have a unique opportunity to monitor the development of activity in a major comet from before water sublimation begins to perihelion and beyond. Systematic observations should show the transition from CO-dominated outgassing to H2O-dominated outgassing and back, and give greater insight into the complex physical phenomena which control the coma development of comets in the inner solar system. The total CO outgassing rate is itself a measure of the area of CO exposed to solar heating. Systematic observations will show how the CO is depleted by prolonged solar heating, and give clues concerning the sizes and lifetimes of active vents on the cometary nucleus. Further, variations in line shape, if observed with a suitable time increment (eg: 1 hour) can reveal the nuclear rotation rate and the development of active vents on the nucleus.

HCN is the most probable parent of CN, which gives rise to the strongest features in cometary optical spectra. We hope to be able to compare simultaneous observations of the HCN 4-3 transition with optical images in the CN lines taken at the UH 88-inch telescope in order to understand the connection between the two species.

2. Chemistry within the coma.

In particular, isotope abundances are of key importance to cosmogonical questions. The early detection of abundant CO shows immediately that Hale-Bopp formed at the low temperatures (< 50 K) characteristic of the solar nebula beyond 30 a.u.

a) We should be able to determine both 12C/13C and 14N/15N from observations of isotopomers of CO and HCN. Typically line temperatures of 2-3K Ta* are required in the main isotope lines for this to be feasible, and these have already been achieved, at least for HCN. Hence at present H13CN and HC15N offer the best possibilities, but require exceptional conditions.

b) Sulfur-bearing molecules such as SO and SO2, H2S and CS (both known in Hyakutake), H2CS.

c) Methyl group molecules such as CH3CN and CH3OH, both already known in Hale-Bopp , provide routes to the determination of excitation temperature.

d) Deuterium molecules, in particular HDO and DCN. HDO were detected at CSO at 464 GHz (Lis et al , 1996) in Hyakutake and provides an important route to the H2O abundance. It is likely only a matter of time before it is detected in Hale-Bopp . A second transition of HDO at 490 GHz should be considerably stronger and offers another way to determine the excitation temperature. We have recently attempted DCN 5-4, but we have not yet had the excellent conditions required to achieve a useful sensitivity.

e) Ions such as CO+ and H2O+ are known from optical spectra and are important targets. Also H3O+ and HCO+ present interesting possibilities.

3. Characterising the dust coma

Observations using UKT14, in particular our work on Hyakutake (Jewitt & Matthews, 1996), show that cometary dust masses are much greater than expected. Submillimetre observations are sensitive to particles of typical size around 1 mm, and these contain a large part of the dust mass. The specific goals we have are:

a) Determination of the dust mass and dust mass loss rate as a function of heliocentric distance. These data will be used to determine the physics of the dust ejection from the nucleus. We would expect a significant difference also in these quantities on opposite sides of perihelion; a recent detailed model (Prialnik, 1997) shows a fairly precipitous decline beginning immediately following perihelion, in contrast to the gas output.

b) Determination of the plane-of-sky morphology of the dust coma at 850/450 microns. The morphology places strong constraints on the dust particle size, as a result of sorting imposed by solar radiation pressure.

c) Determination of the frequency dependence of the opacity, which in turn informs us about the grain size. In Hyakutake we found that the opacity index was Beta = 0.84 +/- 0.11, typical of that found for dust envelopes around young stars. There are three key issues to be resolved in the case of cometary comae: (1) do all comets show the same value of Beta? (2) If not, can Beta be used to characterize origins in different parts of the protosolar cloud? (3) Does Beta vary with heliocentric distance? The last of these points can be answered to some degree in Hale-Bopp by time-resolved observations using SCUBA.

As shown in Figure 5 we have used recent outgassing rates to estimate the 850-micron continuum flux density of Hale-Bopp. Continuum observations were omitted from the original submission since SCUBA was not being offered in semester 96A. Although there are considerable uncertainties involved, by the beginning of March 1997 we find that the emission from dust in the coma should reach 100 mJy, and peaks at around 200 mJy at perihelion. The nucleus should be detectable also, peaking at around 40 mJy. This assumes that the size of the nucleus is 40 km, although there is a factor of about two uncertainty in this number.

Very recently (February 5) however, Kreysa et al (1997) have reported the first detection of continuum emission with a flux density of about 100 mJy from the coma of Hale-Bopp at 250 GHz with the IRAM telescope. Extrapolation to 850 microns using the value of Beta we found for Hyakutake indicates a flux density of about 250 mJy would be seen at this wavelength at present. This suggests that our model of Hale-Bopp considerably underestimates the dust mass in the coma. If this detection is confirmed then we might expect the 850-micron flux density at perihelion to be about 2 Jy.

References

Biver, N., et al , 1996a; Nature, 380, 137

Biver, N., et al , 1996b; IAUC 6386

Crovisier, J., et al , 1995; Icarus, 115, 213

Irvine, W.M., et al , 1996, Nature, 383, 418

Jewitt, D., Matthews, H.E., 1997; A. J., in press (for March)

Jewitt, D., Senay, M.C., Matthews, H.E., 1996a; Science, 271, 1110

Jewitt, D., Senay, M.C., Matthews, H.E., 1996b; IAUC 6377

Kreysa, E., Altenhoff, W., Haslam, C.G.T., 1997; IAUC 6555

Lis, D., et al ,1996; IAUC 6362

Marsden, B.G., 1995; Science News, 148, 103

Matthews, H.E., Jewitt, D., Senay, M.C., 1995; IAUC 6234

Matthews, H.E., Senay, M.C., Jewitt, D., 1996a; IAUC 6318

Matthews, H.E., et al , 1996b; IAUC 6353

Matthews, H.E., et al , 1996c; IAUC 6363

Matthews, H.E., Jewitt, D., Irvine, W.M., 1996, IAUC 6515

Prialnik, D., 1997, preprint

Rauer, H., et al , 1995; IAUC 6236

Senay, M.C., Jewitt, D., 1994; Nature, 371, 229

Senay, M.C., Matthews, H.E., Jewitt, D., 1996a; IAUC 6312

Senay, M.C., Matthews, H.E., Jewitt, D., 1996b; IAUC 6335

Weaver, H.A., 1996; Nature, 380, 107

Womack, M., et al , 1996; IAUC 6382

H.E. Matthews (JAC/HIA), D. Jewitt (IfA, Honolulu), & M.C. Senay (U.Mass).

For further information and observations of comet Hale-Bopp, Dave Jewitt has an excellent homepage HERE