That's the way the comet crumbles: Splitting Jupiter-family comets
aa r X i v : . [ a s t r o - ph . E P ] J u l That’s the Way the Comet Crumbles:Splitting Jupiter-Family Comets
Yanga R. Fern´andez ∗ University of Central Florida, Department of Physics, 4000 Central Florida Blvd.,Orlando, FL 32816-2385 U.S.A.
Abstract
Our current understanding of split, Jupiter-family comets is reviewed. The focusis on what recent studies of comets have told us about the nature of the splittingphenomenon. The goal is to not repeat the information given in recent reviews ofsplit comets, but to build upon it. In particular, we discuss comets that have sufferedsplitting or fragmentation events in the past few years. These include comets (a)57P/du Toit-Neujmin-Delporte, observed with a long train of fragments in 2002; (b)73P/Schwassmann-Wachmann 3, which split in 1995 and was extensively studiedduring its relatively close passage to Earth in 2006, during which dozens of fragmentswere discovered and studied; and (c) 174P/Echeclus, a Centaur and potentiallyfuture JFC, which split in late 2005 and was the first such Centaur observed to doso. We also discuss recent observations by SOHO of split comets that are likely ofshort-period. The Spitzer Space Telescope has observed many JFCs and provided uswith unprecedented detailed views of cometary debris trails, which may be thoughtof as a middle ground between “normal” ejection of micron-sized dust grains andthe cleaving off of meter-to-kilometer sized fragments. We will also discuss potentialbreakthroughs in studying splitting JFCs that may come from future surveys.
Key words: comets – splitting, comets – individual (57P, 73P, 174P), comets – evolution
PACS: ∗ Corresponding author. Tel.: +1 407 8236939; fax: +1 407 8235112.
Email address: [email protected] (Yanga R. Fern´andez).
Article published in Planetary & Space Science 57 (2018) 1218–1227
Introduction
The “typical” Jupiter-family comet (JFC) loses mass through a relativelyslow process of volatile sublimation and dust entrainment. Common mass lossrates when a JFC is near perihelion are roughly ∼ − kg/s (A’Hearn et al.,1995), depending on the size of the nucleus’s “active area.” This indicates thatnominally the cometary nucleus could disintegrate away probably only afterthousands of orbits around the Sun. This end-state of a JFC has never beendirectly observed, although such very tiny comets (meters in diameter) thatare about to disintegrate would be difficult to discover. In any case currentthinking (e.g. Meech and Svoreˇn, 2004) holds that most JFCs end their liveseither by plunging into the Sun, colliding with a planet, simply turning offand becoming asteroid-like, or by catastrophic fragmentation.This review will focus on this last option and its attendant large mass loss rate.The primary motivation for studying split comets is that they are laboratoriesfor understanding cometary structure and bulk mechanical properties. Sincethis can give us clues about how comets are put together, split comets can bean important window for investigating the details of planetary formation andspecifically the accretion of solids into icy planetesimals. As mentioned above,split comets also represent an aspect of cometary evolution. Since split cometsexpose previously buried material to the space environment and thus to ourtelescopes, they give us a way to probe chemical and thermophysical changesin cometary nuclei.In this paper, recent developments in our understanding of split comets will bediscussed. An excellent recent review of this topic is provided by Boehnhardt(2004), and there have been several likewise good reviews before that (e.g.Sekanina, 1982; Hughes and McBride, 1992; Sekanina, 1997; Boehnhardt, 2002).The goal of this paper is to not duplicate that earlier work but rather to pro-vide updates on work occuring in the interim. Conceptually, the question of “how do you know when a comet has split?” iseasy to answer. The observational manifestation of a splitting is a condensationthat appears away from the head of the comet but moving with very nearly thesame proper motion. However not all near-nuclear condensations are actuallyindicative of a split; the condensation may or may not hold a solid body andmay be simply (e.g.) a trick of perspective on jet features in the coma, or aclump of dust from an outburst. The overall dust production rate and thecomplexity of coma morphology can make the actual identification of a real2plitting problematic. In particular, a splitting is often associated with anoutburst or a brightening, potentially making it even more difficult to identifya fragment.Interestingly, a fragment may not appear even when a comet brightens dra-matically. For example comet 29P/Schwassmann-Wachmann 1 has frequentoutbursts (see, e.g., Jewitt, 1990), yet no fragment has ever been seen.On the other hand a comet that appears perfectly “typical” could have had arecent splitting, but the fragments may be more than a few arcminutes fromthe head, i.e. beyond the field-of-view of a typical CCD camera. Due to spo-radic monitoring of most comets, a fragment will not always be seen in thenear-nuclear region; it may not be discovered until well after the split, as wasthe case for fragment B of 57P/du Toit-Neujmin-Delporte (Marsden, 2002a),which was found 0.2 ◦ away from fragment A, the comet’s main head. Theeven more extreme case is the paternity of comets 42P/Neujmin 3 (discov-ered in 1929) and 53P/van Biesbroeck (discovered in 1954), which were found(Carusi et al., 1985) to be pieces of one comet that split after a close approachto Jupiter in 1850.Another problem in determining whether a split has happened is simply faint-ness. A 50-meter radius bare nucleus that is 2.0 AU from the Sun and 1.0AU from Earth – all reasonable numbers for a cometary fragment – will havean R-band magnitude of about 25 (for a geometric albedo of 0.04), which isbeyond the reach of many facilities. Even when such a fragment is activelyoutgassing and thus brighter, the observation itself may not be done in sucha way as to detect the fragment.All this makes determining the frequency of fragmentation among JFCs rela-tively difficult to measure. A list of published instances of JFC fragmentationis give in Table 1, but since many comets have long intervals (months or years)where no observations are obtained by either professionals or amateurs, thisis naturally a lower limit of the true roster. Hughes and McBride (1992) esti-mated that a JFC has 0.3% chance of splitting per perihelion passage, basedon the historical record of observed fragmentations. Chen and Jewitt (1994)observed a sample of 34 JFCs with CCDs in the late 1980s and early 1990s,and found 2 had split, corresponding to a ∼
1% chance of a JFC splitting peryear. If these are accurate estimates of the splitting rate, one remarkable con-sequence is that over the course of the ∼ orbits that a typical JFC will beactive (Levison and Duncan, 1997), it can expect to split perhaps dozens tohundreds of times. As noted by Chen and Jewitt (1994), while JFC splittingsmay be perceived as being rarer than splittings by near-isotropic comets, agiven JFC will shed fragments many times during its active life.This finding suggests that the shape and rotation state of a JFC is intimately3ied to the specific fragmentation events it has suffered, since the amountof mass coming off in a fragment can be significant compared to the totalmass lost by normal outgassing in the course of an orbit. Suppose a JFC withperihelion at 1 AU and aphelion at 5.2 AU (Jupiter’s distance) has a massloss that is 200 kg/s at perihelion and is proportional to the inverse-square ofheliocentric distance. This comet will lose about 4 × kg in one orbit. If theeffective radius of the nucleus is 2 km, and the density is about 400 kg/m ,that is just 0.03% of the comet’s total mass. If the comet is “active” over 10%of its surface, the mass loss from those active areas erodes about 2 m deep.If an equivalent mass were to come off as a spherical fragment, however, thefragment would be 130 m in radius. This is a reasonable size for a “typical”fragment of a JFC (Boehnhardt, 2004). Such a splitting would represent asignificant change in gravitational field, angular momentum, and topography.As mentioned earlier, we have not discovered a population of very tiny comets.We do not have observational evidence of a population of deka- or hectometersized JFCs that could be made up of fragments. The discovery biases almostcertainly play a role in this, although, as Meech et al. (2004) showed, if oneaccounts for the biases it seems that the JFC size distribution really does falloff below ∼ § While tidal disruption as a mechanism for splitting comets is well understood,it is responsible for only a small fraction of observed splits. Briefly, a cometpassing close enough to a planet (usually Jupiter) or the Sun will feel dif-ferent gravitational forces on one end of its nucleus compared to the other.The difference in forces can be strong enough to overcome the body’s cohe-sion. The extremely low tensile strength and high porosity of cometary nuclei,as suggested most recently by the Deep Impact visit to comet 9P/Tempel1 (Ernst and Schultz, 2007; Holsapple and Housen, 2007; A’Hearn, 2008), in-4icate that a tidal force need not be that strong to successfully rip a cometapart. However, in only one case – comet D/1993 F2 (Shoemaker-Levy 9)– arewe very sure that tidal disruption is the cause of the fragmentation. Comet16P/Brooks 2 probably also suffered this fate but it was not discovered untila few years after the purported close-approach.For all the other comets in Table 1, tidal disruption could not have been thecause of the split. In these cases, we are no closer to understanding why aparticular comet splits when it does than we were when Boehnhardt (2004)wrote his review. He gives four other methods that could cause splitting: byfast rotation, by thermal stress, by internal gas pressure, and by impacts.Since that review was written, our understanding of the thermal propertiesof cometary nuclei has improved as a result of the Deep Impact experiment.For example there is stronger evidence now that the thermal conductivityof cometary nuclei is extremely low (Sunshine et al., 2007; Groussin et al.,2007), so low that the thermal pulse can penetrate only a few centimeters dueto diurnal heating. This means that on diurnal time scales thermal energymay not be transported effectively into the comet’s interior any faster thanthe surface can itself be excavated by normal cometary activity. The existenceof abundant CO – with its low sublimation temperature – in P/Tempel 1’scoma (Feaga et al., 2007) also indicates that the comet could not have beenentirely baked out. Of course these are results for just one comet, but if theyare indicative of the “average” thermal properties of a JFC, then perhaps itmay be less likely for JFCs to split as a result of thermal stress. On the otherhand, the low strength of a JFC means that one can conceive of localizedstructures (e.g. sheer cliffs in depressions or concentrations of less porous rock)where thermal stress in just a small volume could cause a much bigger volumeto break off. This relates to more fundamental questions of nucleus structureabout which we do not yet have much data. How well are the volatiles andrefractories mixed within the nucleus? Is the high porosity manifest in micro-or macro- scales? How common is significant topography? Fragment B was discovered in July 2002 well away from the main part of thecomet. Soon after, we (Fernandez et al., 2002) discovered eighteen more frag-ments (named “C” through “T”) along the line of variation (the projectedorbital path) and extending out 27 arcminutes from the comet’s head. A mon-tage of these individual fragments as seen on UT July 17 is shown in Fig.2, along with a plot of the fragments’ positions. The fragments had varying5rightnesses ranging from 20 to 23.5 mag in R-band. They also had widelyvarying degrees of condensation; some fragments were nothing more than blobsof dust with no central source, such as I and P. All fragments seemed to beactively outgassing, though with apparently varying production rates.Interestingly, there was no outburst in 2002 associated with this shedding ofmass. There was a significant outburst at the previous apparition, in 1996,when the comet was observed to be about 5 mag brighter than expected.However dynamical analysis of the largest fragments by Sekanina and Chodas(2002) suggests that they could not have broken off six years earlier.The fragments must represent a significant fraction of the comet’s total mass.How much mass, and what fraction is it? The size of the nucleus before frag-mentation is unknown, but Lowry and Fitzsimmons (2001) derived an upperlimit to the radius of 1.1 km. So for an assumed density of 400 kg/m , theupper limit to total mass would be approximately ∼ × kg. The activityof the fragments makes their size estimation problematic, so a mass estimate iseven more difficult. However we can make an order-of-magnitude analysis. Forexample, fragment G had an R-band magnitude of about 22.5, which at thedistance of 57P at the time, would correspond to a solid body with effectiveradius of about 130 meters. Such a body alone would be ∼ N ( > R ),the number of objects with radius bigger than R , proportional to only R − . .While this totally ignores the obvious activity of the fragments, if activityscales roughly with the fragment’s surface area, the power-law slope of thesize distribution will be unaffected. This slope means that the total mass m contained in fragments up to some size cutoff R is proportional to R . . Thissuggests that much of the fragmented mass is in the largest pieces such as B,E, and F.Interestingly, the CSD slope is shallower than that of large dust grains found inJFC trails, as measured by Reach et al. (2007a). They found N ( > R ) ∝ R − . for grains over 0.25 mm. Whether this is a clue as to the physical mechanismbehind fragmentation remains to be studied.The real significance of the discovery of the train behind 57P is that it was thefirst time so many fragments had been observed around a surviving, non-tidallydisrupted comet. Other comets have had as many fragments but only while6eing completely broken apart or after passage by Jupiter or the Sun (e.g.,comet C/1999 S4 (LINEAR), comet D/1993 F2 (Shoemaker-Levy 9), andthe Kreutz sungrazers). It motivates the question of whether shedding eventshappen more frequently than thought and are just being missed due to theextreme faintness of the fragments. Indeed, the sheer length of 57P’s trainsuggests that there were several fragmentation episodes in the past. If a JFCloses only fragments of magnitude 23, 24, or fainter, only deep imaging of thecomet during the course of monitoring will reveal them.The case of 57P also raises questions about the endurance of the smallestfragments. Sekanina and Chodas (2002) state that fragment F would have leftthe primary nucleus – i.e. fragment A – about 14 months before the discoveryobservation; at discovery, F was 6 arcmin from Fragment A. They predictedthe future motion of fragment F based on this model, but unfortunately ap-parently little if any data could be obtained by observers to corroborate orrefute the hypothesis. However, if true, it suggests that large fragments fartherdown the train could have been released at even earlier times. Fragment T is27 arcminutes from fragment A – 4.5 times farther than fragment F. T couldhave left the main nucleus (on its own or as part of another fragment) years inthe past. This would not necessarily be unprecedented, since as Boehnhardt(2004) notes JFC fragments can have long endurance (hundreds of days), andthere is indication that JFCs can split most anywhere along their orbit. Theestimation of the endurance of fragment T would also depend on at what rela-tive speed it left the nucleus. In any case, fragment T is a fairly small fragment:could it have survived for years, and are we just now seeing the end of its life?Is it really persistent despite its small size? It is also possible that fragmentT is a subfragment of a brighter fragment, say fragment S. Even so, the dis-tance between S and T is appreciable (3.2 arcmin) and suggests an endurancefor T of several months despite its small apparent size. Beech and Nikolova(2001) have modeled the survival times of fragments and show that small,clean icy fragments can survive for months after separation, although whetheror not fragment T can be clean “enough,” is uncertain. The images in Fig. 2show cometary dust, not gas, and so there must be some “dirtiness” to thefragments, which would shorten their lives (Hanner et al., 1981; Lien, 1990).The faintness of the fragments, and then later the fact that the comet was postperihelion and at unfavorable elongation, prevented detailed follow up. Theendurance of the the fragments was not directly measured. Therefore furtherobservations of the fragments and of the main comet itself at future apparitionswould be extremely useful to understand how this comet is evolving.7 .2 Comet 73P/Schwassmann-Wachmann 3 One of the most-widely observed split comets since D/Shoemaker-Levy 9 hasbeen 73P/Schwassmann-Wachmann 3, which first split in 1995 and approachedto within 0.07 AU of Earth in May 2006. This recent visit was a monumentallyimportant apparition scientifically since it would allow us detailed studies ofbright and relatively-fresh fragments – fragments whose ices had formerly beendeeply embedded within the comet’s nucleus but were now exposed to sunlight.In a way, it was similar to the Deep Impact experiment in that 73P providedus with a close-up view of subsurface pristine cometary material.A review of this comet’s behavior was given by Boehnhardt et al. (2002) andby Sekanina (2007), which we briefly summarize here. This comet split intoseveral fragments in 1995, some of which deactivated or disintegrated beforethe apparition was over. At the comet’s next apparition in 2000 and 2001,two of the 1995 fragments, B and C (the primary part of the comet), wererecovered, and two new fragments, E and F, were confirmed (although appar-ently E broke off from C during the previous orbit). F disappeared later in theapparition but the stage was thus set for the 2006 apparition: would B and Csurvive another perihelion passage? Would E have survived? Would there bea train of even more fragments?The result was that in April and May 2006 over 60 fragments were found– including B and C but not E. The proximity to Earth certainly helped; afragment of the same size as 57P-G (mentioned in the last subsection) wouldbe about 3 mag brighter at 73P’s geocentric distance. In any case some ofthe fragments were very short-lived, lasting for only a few days. Amateurastronomers contributed greatly to the census of fragments, in fact even findingfragments that disappeared within a day and so could not be followed up andformally named. Lists of the fragments and relevant info have been compiledby Sekanina (2007) and Birtwhistle (2008), and the former also discusses somepreliminary work on describing the cascading fragmentation.One of the many spectacular images from this apparition is shown in Fig. 3,which comes from the Hubble Space Telescope. Fragment B was a dynamicand rapidly changing fragment for much of the apparition, and this imageepitomizes this. It shows several subfragments (a.k.a. “mini-comets”) tailwardof fragment B itself. These pieces are probably dekameter in scale, and asequence of images from HST show these fragments moving down the tail,outgassing until they disintegrate away in timescales of only hours or days. Theimage field-of-view is only about 25 arcseconds. A wider scale picture is shownin Fig. 1, covering almost 5 degrees of sky along the comet’s orbital path. Thedynamics of the small fragments in Figs. 1 and 3 have been analyzed byReach et al. (2007b), who found that the HST fragments are strongly affected8y the non-gravitational reaction force due to outgassing – suggesting a highvolatile content. On the other hand, the “meteoroids” in the Spitzer image aremoving as would be expected simply from radiation pressure and solar gravity– suggesting a low volatile content. This perhaps could be explained as anevolutionary effect, where the HST fragments dry out to become the Spitzerfragments. This does require that the HST fragments would have to havesufficient size and sufficiently low rock-to-ice ratio to be able to survive for theseveral days that they are seen in the HST data. Unfortunately, identifying thefragments seen by HST on April 18 with some part of the extended emissionseen in the Spitzer mosaic on May 4 would be problematic since the spatialresolutions are so different and the time gap is so large.The variability of fragment B is in stark contrast to fragment C, the primaryfragment (Sekanina, 2007). That fragment remained relatively stable, withonly slowly varying activity, nowhere near the frequency and amplitude ofchanges seen from night-to-night (and sometimes within a night) in fragmentB. The two fragments must be roughly comparable in size, yet the specificsof B’s shape and the location of its volatiles has made apparently a hugedifference in the evolution.Many observations of comet 73P were obtained during its apparition, and anal-ysis on these rich datasets continue. We summarize here some of the excitingresults and apologize for oversights.Arguably some of the most important findings involve the composition. Over-all, 73P seems to be depleted in CH OH, C H , and C H , but has “typical”abundance of HCN (Villanueva et al., 2006; DiSanti et al., 2007; Kobayashi et al.,2007). This means that this ecliptic comet that came from the trans-neptunianregion is compositionally similar to C/1999 S4 (LINEAR), a carbon-chaindepleted Oort Cloud comet (DiSanti and Mumma, 2008). Furthermore it isunlike its fellow ecliptic comet 9P/Tempel 1 (DiSanti and Mumma, 2008).This matching of compositions across dynamical classes hints that there wassufficient mixing in the protoplanetary disk to allow individual icy planetesi-mals to accrete material from various regions. In other words perhaps there isnot necessarily a compositional distinction that exactly matches the dynam-ical distinction of ecliptic comets forming beyond Neptune and Oort Cloudcomets forming among the giant planets. Alternately, there could simply beinterlopers polluting the dynamical groups. In particular, we are only nowstarting to build up a statistically significant sample of the parent-moleculecomposition of JFCs (Mumma, 2008) to build upon the daughter-species workby (e.g.) A’Hearn et al. (1995) and Schleicher and Bair (2008). We can notethat 73P is a “depleted” comet in the A’Hearn et al. (1995) taxonomy and 9Pis “typical;” as the parent species of more JFCs are observed, 73P and 9P canbe placed into better context. 9everal people compared the two main fragments, B and C, to each other (e.g.Biver et al., 2006; Villanueva et al., 2006; Schleicher et al., 2006; Dello Russo et al.,2007; Kobayashi et al., 2007). The consensus result is that B and C have sim-ilar composition. This is an important finding since both fragments are rela-tively large fractions of the original comet and so show us fresh material thatformerly was very deep inside the nucleus. They should be excellent labora-tories for determining heterogeneity – i.e., whether large blocks of the comethave different compositions. No such effect was found, and this is in stark con-trast to comet 9P, where the heterogeneity was quite obvious after the DeepImpact flyby (e.g. Feaga et al., 2007). The fact that 73P seems to be both (a)homogeneous and (b) different from what may be currently considered “typi-cal” composition suggests that this comet had an atypical formation history.However this is speculative and, again, we are suffering somewhat from a smallJFC sample. In any case we clearly see the vital need for more surveys of JFCcomposition.One important additional result regarding composition was presented by Schleicher et al.(2006). They showed that the B and C fragments both have the same “de-pleted” abundance of CN, C , and C relative to water. As mentioned, sincethe fragments were outgassing relatively pristine material, this suggests thatthis depletion seen in many JFCs could not be an evolutionary effect, butrather is primordial. It may be indicative of the chemistry happening in theprotoplanetary disk at the location where at least some of the JFCs formed.The proximity of 73P motivated many observers to investigate the physicalproperties of the fragments. Of primary concern were the fragments’ sizes andmasses. Boehnhardt et al. (1999) observed the comet in 1994, before breakup,and estimated an upper limit of 1.1 km for the radius. Some of the most ex-citing data on the nuclei after break-up were obtained by Howell et al. (2007),who used radar to obtain Delay-Doppler maps of fragments B and C in May2006. 73P was at the time only the second comet so imaged. The data showthat fragment B is at least 0.2 km in radius and that fragment C is about0.5 to 1 km in radius. These results are consistent with earlier estimates ofthose two fragments’ sizes (Boehnhardt, 2002; Toth et al., 2003, 2005, 2006).While fragment C is the primary remnant of the comet, fragment B took asignificant fraction of the mass with it.Knowing the rotational states of the larger fragments could potentially giveinsight into the fragmentation process, the dekameter-scale structure of thecomet, and/or the cometary mass. So far, there have been reports only for frag-ment C (e.g. Storm et al., 2006; Toth et al., 2006). Fragment B was so activeand changing on such short timescales that obtaining either a photometrically-or morphologically-derived period may be challenging. Coma structures wereseen however (e.g. Bonev et al., 2008) so a sufficient baseline of observationscould prove fruitful. Rotation periods for the other relatively bright fragments,10uch as G and H, have not been reported to our knowledge.Images such as Figs. 1 and 3 make it clear that there is a continuum of sizesamong the fragments. An analysis of the size distribution of all fragments hasyet to be presented, but Fuse et al. (2007) have studied a group of fragmentsthat at the time had all just recently broken off fragment B. Their processedimage is shown in Fig. 4, and was obtained a few weeks after the HST image inFig. 3. They identify 54 fragments in their data, all of which were active, andmeasured the luminosity of each. Assuming that the activity is proportionalto the fragment surface area, they then derive a CSD power-law slope basedon their 54 fragments: N ( > R ) ∝ R − . . This is tantalizingly similar to therough CSD slope for 57P as discussed in § The Centaur 174P/Echeclus = (60558) 2000 EC was discovered by Space-watch in March 2000 (Scotti et al., 2000), and it orbits between 5.9 and 15.6AU from the Sun. Its current orbital intersection distance with Jupiter is 0.9AU and with Saturn is just 0.2 AU; as a Centaur it is likely to be significantlyperturbed on ∼ year timescales (Levison and Duncan, 1997) and may be-come a JFC. No cometary activity was reported for several years after itsdiscovery, and physical properties of the bare object were obtained by severalgroups (Rousselot et al., 2005; Lorin and Rousselot, 2007; Stansberry et al.,2008). Activity was first noticed in December 2005 by Choi et al. (2006a) andcontinued through May 2006 (Choi et al., 2006b; Weissman et al., 2006), whilethe comet was about 13 AU from the Sun.Bauer et al. (2008) present an analysis of contemporaneous ground-based vis-ible and Spitzer infrared imaging of the comet from February 2006. Their im-ages are shown in Fig. 5, and demonstrate why this comet should be counted Note that Fuse et al. (2007) call their “ q ” the CSD power-law slope but it isactually the differential size distribution’s slope; their 1 − q = − .
11s one that has split. The center of brightness of the cometary activity is noton the nucleus itself, but offset (by six arcseconds at the time the images weretaken). Furthermore there is a condensation embedded in the coma. However,imaging obtained in March 2006 (i.e. the following month) by Rousselot (2008)shows a more diffuse coma, and they state that the surface brightness profilesof the coma suggest that the dust is no longer coming from a central sourcebut rather from a diffuse source.One hypothesis to explain the observations is: Echeclus itself was mostly inac-tive, but perhaps active enough in one locale for a fragment to break off. Thisfragment stayed active, while the remainder of Echeclus continued to have noactivity. After a few months, the fragment itself began to disintegrate intosmaller pieces or subfragments. The subfragments remained active and so – atthe spatial resolution obtainable from Earth – the coma appeared to emanatefrom a distributed source. The problem with this scenario is explaining whythere would not be activity from the “hole” on the primary created by thedeparting fragment.An alternate hypothesis is that we are seeing a satellite of Echeclus that justhappens to be active. However, the motion of the fragment over the course ofseveral months suggests that it is moving hyperbolically (Choi et al., 2006b;Weissman et al., 2006). The apparent motion is too great for a bound orbit,given the expected mass of Echeclus. Also, Echeclus’s Hill sphere radius atthe time was roughly only 5 × km, i.e. about 6 arcsec on the sky. Furtherevidence against the satellite hypothesis is that a search in earlier deep imaging(when no activity was seen) yielded no such object (Rousselot, 2008).The fragment itself could be a few kilometers in radius (Rousselot, 2008) –compared to the ∼
40 km of the primary – and so a large impulse wouldbe needed for that much mass to be accelerated up to the ∼
15 to 30 m/srequired to reach escape velocity. While typical separation speeds are an orderof magnitude smaller (Boehnhardt, 2004), we note that a fragment that is flungoff the surface of Echeclus due to the primary’s rotation could be given sucha speed, since in that case the fragment’s speed would be proportional to theprimary’s radius. On the other hand, current evidence suggests that Echeclushas a relatively long rotation period of 27 hours (Rousselot et al., 2005).While a fragment remains the best explanation for 174P, further study ofthis enigmatic comet is certainly warranted. The main science goals wouldbe to (a) ascertain the nature of Echeclus’s activity so as to explain howsuch a large mass could have left the primary; (b) explain why activity isso tightly localized; (c) determine what the source of the activity is (CO?CO ? crystallization of amorphous H O ice?) and whether this plays a rolein making fragmentation more likely; (d) monitor the rest of the Centaurpopulation to determine whether any other objects suffer these events, and12hat the frequency is; and (e) infer what Centaur fragmentation implies forthe bulk structure and strength of the JFCs that the Centaurs become.
The SOHO spacecraft has discovered over 1500 comets, almost all of whichare “sungrazers” (comets with perihelia less than about 0.06 AU). A largefraction of the SOHO comets are part of the Kreutz family, whose cometspass only about 0.005 AU from the center of the Sun. Since the Sun’s radiusis 0.00465 AU, many Kreutz comets (and almost all the ones discovered bySOHO, which are faint) disintegrate to dust very near perihelion if not earlier.The Kreutz comets are all presumably fragments of a long-period comet thatbroke apart thousands of years ago; see Marsden (2005a) for a recent review.Since the Kreutz comets are of long-period, we do not consider them here.However SOHO has discovered four other families of sungrazer comets, someof which may be of short-period. While the Meyer group (Meyer, 2002) cometshave orbital inclinations of about 72 ◦ , well out of the ecliptic, the Marsdengroup (Marsden, 2002b), the Kracht group (Kracht, 2002a), and the KrachtII group (Kracht, 2002b) have members with inclinations of 27 ◦ , 13 ◦ , and 13 ◦ ,respectively, perfectly normal for short-period comets. As with the Kreutzfamily, the members of each group have similar orbits and thus imply a singleprogenitor in the past. The perihelia of these comets are at about 0.05 AU,so, while not currently passing within the Sun’s Roche limit, perhaps stressdue to tidal forces or due to energy transport facilitated fragmentation athigher heliocentric distances. Such distant fragmentation has been suggestedby Sekanina (2002) for the Kreutz comets.Recent work incorporating the 12-year database of cometary astrometry fromSOHO has revealed possible linkages between some of the comets in these threegroups, as well as between some comets not belonging to any known group.This information is collected in Table 2. It is worth noting that the linking ofapparitions with short-arc orbits can be difficult, especially when independentcomets move on similar orbits anyway. There are several other possible linksamong comets within the three groups that have been reported (e.g. Marsden,2006; Kracht, 2008). Sekanina and Chodas (2005) present a detailed analysisof some of the Marsden and Kracht group linkages.Detailed physical studies of the sungrazing short-period comets are currentlylimited to observations by Sun-staring spacecraft. These often provide mag-nitudes and thus some secular light curve. In many cases, the comets aresufficiently faint that neither a tail nor an extended coma is visible, limitingthe amount of available information on gas and dust. The nuclei are probably13f order dekameters in radius (Marsden, 2005a); this means that even with animproved orbit, it would be feasible to observe such a comet well away fromthe Sun only when it is near Earth. For example, a 10-meter radius nucleus atopposition 0.1 AU from Earth still only achieves an R-band magnitude of 22.5.So even in a very favorable apparition (which would not happen very often tobegin with), it will not be easy to make detailed studies of such comets. Thedetailed study of a wide sample of sungrazing short-period comets will proba-bly have to wait for very deep and wide sky surveys, or for classically-scheduledtime at the largest telescopes. As comet 73P shows, comet fragments have a distribution of sizes. Indeed,Fig. 1 demonstrates that such a comet gives off fragments that are decimetersized and larger – fragments that are larger than what can be typically liftedoff a comet’s surface simply by gas drag (Gr¨un and Jessberger, 1990). Thesefragments remain in a trail in the orbit plane since radiation pressure actsslowly on them.The deep infrared observations by IRAS brought to light the existence ofcometary debris trails (Sykes and Walker, 1992), consisting mainly of mil-limeter and centimeter sized grains that are the largest solid bodies that comeoff the comet during “normal” activity. Such grains have often been found tocontain most of the mass that is contained in dust. The grains in the trails rep-resent an intermediate size scale between the visible-wavelength dust that istypically micron sized and smaller, and observed fragments that are dekametersized and larger.An infrared survey of debris trails has been performed by Spitzer. Reach et al.(2007a) present 24 µ m imaging of 34 JFCs in which a trail is unambiguouslyseen in 27 of them – about 80%. Most of these comets were observed while2 to 4 AU from the Sun, and months away from perihelion. Examples of theobservations presented by Reach et al. (2007a) are shown in Fig. 6.Most of the comets in their survey had not suffered splitting events in thetraditional sense; the particles were liberated by regular cometary activity.One could call these millimeter and centimeter sized particles “fragments”only in a liberal definition. Nonetheless, the high frequency of trails, coupledwith the fact that often the trail was seen all the way out to the edge ofthe field of view, indicates that JFCs are prodigious producers of millimeterand centimeter sized fragments into the interplanetary dust environment. AsReach et al. (2007a) point out, the total mass of these large grains is largerthan the total mass of grains seen at visible wavelengths.14 .6 Searches for Fragments Usually fragments are found serendipitously, but a few pointed searches forfragments have been performed. The survey of Chen and Jewitt (1994), men-tioned earlier, focussed on finding fragments near (i.e. a few arcminutes awayfrom) the primary comet. The faint meter and dekameter scale objects couldhave been missed, however. Beech et al. (2004) did a telescopic search formeter-sized (and larger) Perseid meteoroids while Earth was passing throughthe meteor stream; this would have detected fragments of the Halley-familycomet 109P/Swift-Tuttle. They sought to find objects a few arcminutes awayfrom the radiant before they collided with Earth’s atmosphere. The searchfound no objects, thus constraining the space density of meter-scale objectswithin the stream.Another survey has made use of the proliferation of CCD cameras with verywide fields-of-view. Stevenson and Jedicke (2007) have used the Megacam in-strument on the CFH Telescope atop Mauna Kea to search for fragments upto half a degree away from a comet’s head. They have observed about 12JFCs so far with no fragments seen to a limiting magnitude of about 24. Forthese comets this limit corresponds to fragments of size about 100 m in ra-dius. With such a large field-of-view, the search is not just a snapshot in time,but actually tests whether these comets have had any splitting event in theprevious several months. This kind of survey gives a more complete view ofthe fragmentation history and so should provide a good new measurement ofthe frequency of JFC splittings.
The interested reader is encouraged to study the recent review of split cometsgiven by Boehnhardt (2004). We have tried here to provide a summary ofrecent developments on the topic of split comets. But what does the futurehold for our understanding of this phenomenon? While splittings caused bytides can be predicted if the comet’s orbit is known, the large majority of JFCsplittings instead happen stochastically. Classically-scheduled observing runswill often fail to catch these events. Synoptic observations of comets howeverwill in general be more fruitful for understanding the full evolution of a cometbefore, during, and after a fragmentation event; unbiassed and well-sampleddatasets will be important.It is possible that all-sky survey projects in the coming decade will indeedprovide us with a great deal of such data. For example, Pan-STARRS andLSST will likely discover several hundred comets each in their first year of15peration in addition to the thousands of new Trojans, Centaurs, and trans-Neptunian objects they will find (Jewitt, 2003; Ivezi´c et al., 2008). In addi-tion, the temporal coverage (scanning the sky on week-long timescales), depth(reaching approximately 23rd magnitude), and angular resolution will ensurethat moderate-sized fragments of JFCs will be found more reliably. Further-more, we will be able to follow the photometric and dynamic evolution ofthe fragments for months and possibly years at a time. The datasets willrevolutionize our picture of fragment behavior and endurance (in addition torevolutionizing many questions about comets).Since the review by Boehnhardt (2004) was published, the most importantevent with regard to split JFCs was the apparition of comet 73P/Schwass-mann-Wachmann 3 in 2006. Following up other known split comets in thefuture would be worthwhile as well but 73P represents a unique case that iscrucial to study. Observations of 73P in forthcoming years can give us newdetails about its fragments and its continuing evolution. If fragments B andC survive, we will be able to watch their new surfaces age as they continue tobe exposed to the space environment. Will the resulting chemistry change theapparent production rates we measure in their comae? If so, how? Observationsof other JFCs at multiple apparitions have shown a constancy in relative comaabundances from orbit to orbit (A’Hearn et al., 1995), but likely none of thosecomets have had such a recent fragmentation event that produced a secondaryof comparable size to the primary. Physical observations of the continuedunraveling of fragment B and perhaps the eventual unraveling of fragment Cwould also give us more insight into cometary structure. Unfortunately, threeof the next five apparitions of 73P are unfavorable – only the visits in 2022and 2033 are good – but even limited data on each fragment at each orbitwould be useful.More generally, wide-field observations of all JFCs that will make close ap-proaches to Earth would be useful for understanding the frequency of frag-mentation and the endurance of the fragments, since the proximity will makesearches for meter-scale fragments feasible. For example, comet 103P/Hartley2 approaches to 0.12 AU in 2010, and comet 45P/Honda-Mrkos-Pajduˇsakov´acomes to 0.06 AU in 2011.
Acknowledgements
We thank W. T. Reach, J. M. Bauer, H. A. Weaver, and T. Fuse for theuse of their figures, and M. S. Kelley for helpful suggestions. We appreciatethe thorough reviews of this manuscript given by N. Samarasinha and ananonymous referee. 16 eferences
A’Hearn, M.F., Millis, R.L., Schleicher, D.G., Osip, D.J., Birch, P.V., 1995.The ensemble properties of comets: Results from narrowband photometryof 85 comets, 1976-1992. Icarus 118, 223–270.A’Hearn, M.F., 2008. Deep Impact and the origin and evolution of cometarynuclei. Space Sci. Rev., 138, 237–246.Bauer, J.M., Choi, Y.-J., Weissman, P.R., Stansberry, J.A., Fernandez, Y.R.,Roe, H.G., Buratti, B.J., Sung, H.-I., 2008. The large-grained dust coma of174P/Echeclus. Publ. Astron. Soc. Pac. 120, 393–404.Beech, M., Illingworth, A., Brown, P., 2004. A telescopic search for large Per-seid meteoroids. Mon. Not. Roy. Astron. Soc. 348, 1395–1400.Beech, M., Nikolova, S., 2001. The endurance lifetime of ice fragments incometary streams. Plan. & Space Sci. 49, 23–29.Birtwhistle, P., 2008. 73P/Schwassmann-Wachmann (fragment summary2006). URL .Accessed June 25, 2008.Biver, N., Bockel´ee-Morvan, D., Boissier, J., Colom, P., Crovisier, J.,Lecacheux, A., Lis, D. C., Parise, B., Menten, K., and the Odin team, 2006.Comparison of the chemical composition of fragments B and C of comet73P/Schwassmann-Wachmann 3 from radio observations. Bull. Amer. As-tron. Soc. 38, 484.Boehnhardt, H., 2002. Comet splitting observations and model scenarios.Earth Moon & Plan. 89, 91–115.Boehnhardt, H., 2004. Split comets. In: Festou, M. C., Keller, H. U., Weaver,H. A. (Eds.), Comets II, pp. 301–316. University of Arizona Press.Boehnhardt, H., Holdstock, S., Hainaut, O., Tozzi, G.P., Benetti, S., Licandro,J., 2002. 73P/Schwassmann-Wachmann 3 - One orbit after breakup: Searchfor fragments. Earth Moon & Plan. 90, 131–139.Boehnhardt, H., Rainer, N., Birkle, K., Schwehm, G., 1999. The nuclei ofcomets 26P/Grigg-Skjellerup and 73P/Schwassmann-Wachmann 3. Astron.& Astrophys. 341, 912–917.Bonev, T., Boehnhardt, H., and Borisov, G., 2008. Broadband imaging andnarrowband polarimetry of comet 73P/Schwassmann-Wachmann 3, compo-nents B and C, on 3, 4, 8, and 9 May 2006. Astron. & Astrophys. 480,277–287.Carusi, A., Perozzi, E., Valsecchi, G.B., Kresak, L., 1985. First results of theintegration of motion of short-period comets over 800 years. In: Carusi, A.,Valsecchi, G.B. (Eds.), Dynamics of Comets: Their Origin and Evolution,pp. 319–340. D. Reidel Publishing Co.Chen, J., Jewitt, D., 1994. On the rate at which comets split. Icarus 108,265–271.Choi, Y.-J., Weissman, P. R., Polishook, D., 2006a. (60558) 2000 EC . IAUCirc. 8656.Choi, Y.-J., Weissman, P.R., Chesley, S., Bauer, J., Stansberry, J., Tegler, S.,17omanishin, W., Consolmagno, G., 2006b. Comet 174P/Echeclus. CentralBureau Electr. Teleg. 563.Dello Russo, N., Vervack, R.J., Weaver, H.A., Biver, N., Bockel´ee-Morvan, D.,Croivisier, J., Lisse, C.M., 2007. Compositional homogeneity in the frag-mented comet 73P/Schwassmann-Wachmann 3. Nature 448, 172–175.DiSanti, M. A., Anderson, W.M., Villanueva, G.L., Bonev, B.P., Magee-Sauer,K., Gibb, E.L., Mumma, M.J., 2007. Depleted carbon monoxide in fragmentC of the Jupiter-Family comet 73P/Schwassmann-Wachmann 3. Astrophys.J. 661, L101-L104.DiSanti, M. A., Mumma, M. J., 2008. Reservoirs for comets: Compositionaldifferences based on infrared observations. Space Sci. Rev., 138, 127–145.Ernst, C. M., Schultz, P. H., 2007. Evolution of the Deep Impact flash: Im-plications for the nucleus surface based on laboratory experiments. Icarus190, 334–344.Feaga, L. M., A’Hearn, M.F., Sunshine, J.M., Groussin, O., Farnham, T.L.,2007. Asymmetries in the distribution of H O and CO in the inner comaof comet 9P/Tempel 1 as observed by Deep Impact. Icarus 190, 345–356.Fernandez, Y. R., Jewitt, D. C., Sheppard, S. S., 2002. Comet 57P/du Toit-Neujmin-Delporte. IAU Circ. 7935.Fuse, T., Yamamoto, N., Kinoshita, D., Furusawa, H., Watanabe, J., 2007.Observations of fragments split from nucleus B of comet 73P/Schwassmann-Wachamnn 3 with Subaru Telescope. Publ. Astron. Soc. Japan 59, 381–386.Groussin, O., A’Hearn, M.F., Li, J.-Y., Thomas, P.C., Sunshine, J.M., Lisse,C.M., Meech, K.J., Farnham, T.L., Feaga, L.M., Delamere, W.A., 2007.Surface temperature of the nucleus of comet 9P/Tempel 1. Icarus 187, 16–25.Gr¨un, E., Jessberger, E. K., 1990. Dust. In: Huebner, W. F. (Ed.), Physicsand Chemistry of Comets, pp. 113–175. Springer-Verlag.Hanner, M. S., Giese, R. H., Wiess, K., Zerull, R., 1981. On the definitionof albedo and application to irregular partcles. Astron. & Astrophys. 104,42–46.Holsapple, K. A., Hausen, K. R., 2007. A crater and its ejecta: An interpreta-tion of Deep Impact. Icarus 187, 345–356.Howell, E. S., Nolan, M.C., Harmon, J.K., Lovell, A.J., Benner, L.A., Ostro,S.J., Campbell, D.B., Margot, J., 2007. Radar and radio observations ofthe fragmented comet 73P/Schwassmann-Wachmann 3. Bull. Amer. Astron.Soc. 39, 486.Hughes, D. W., McBride, N., 1992. Short-period comet splitting. J. Brit. As-tron. Assoc. 102, 265–268.Ivezi´c, ˇZ., and 93 colleagues, 2008. LSST: From science drivers to referencedesign and anticipated data products. arXiv:0805.2366v1 [astro-ph] 15 May2008.Jewitt, D., 2003. Project Pan-STARRS and the outer Solar System. EarthMoon & Plan. 92, 465–476.Jewitt, D., 1990. The persistent coma of comet P/Schwassmann-Wachmann18. Astrophys. J. 351, 277–286.Kobayashi, H., Kawakita, H., Mumma, M.J., Bonev, B.P., Watanabe, J., Fuse,T., 2007. Organic volatiles in comet 73P-B/Schwassmann-Wachmann 3 ob-served during its outburst: A clue to the formation region of the Jupiter-Family comets. Astrophys. J. 668, L75–L78.Kracht, R., 2002a. Comets C/1999 M3, 2002 E1 (SOHO). Minor Plan. Electr.Circ. 2002-E18.Kracht, R., 2002b. Comets C/2002 R4, 2002 R5, 2002 R6, 2002 R7, 2002 R8(SOHO). Minor Plan. Electr. Circ. 2002-S35.Kracht, R., 2008. Orbital links between the Marsden and between the Krachtgroup comets. URL . AccessedJuly 3, 2008.Lien, D. J., 1991. Optical properties of cometary dust. In: Newburn, R.L.,Neugebauer, M., Rahe, J. (Eds.), Comets in the post-Halley era, pp. 1005–1041. Kluwer.Levison, H. F., Duncan, M. J., 1997. From the Kuiper Belt to Jupiter-Familycomets: The spatial distribution of ecliptic comets. Icarus 127, 13–32.Lorin, O., Rousselot, P., 2007. Search for cometary activity in three Centaurs[(60558) Echeclus, 2000 FZ and 2000 GM137] and two trans-Neptunianobjects [(29981) 1999 TD and (28978) Ixion]. Mon. Not. Roy. Astron. Soc.376, 881–889.Lowry, S. C., Fitzsimmons, A., 2001. CCD photometry of distant comets II.Astron. & Astrophys. 365, 204–213.Marsden, B. G., 2002. Comet 57P/du Toit-Neujmin-Delporte. IAU Circ. 7934.Marsden, B. G., 2002. Comet C/2002 C3 (SOHO). Minor Plan. Electr. Circ.2002-C28.Marsden, B. G., 2004. Comets C/2004 V9, 2004 V10 (SOHO). Minor Plan.Electr. Circ. 2004-X73.Marsden, B. G., 2005a. Sungrazing comets. Ann. Rev. Astron. & Astrophys.43, 75–102.Marsden, B. G., 2005b. Comets C/2005 E3, 2005 E4 (SOHO). Minor Plan.Electr. Circ. 2005-E87.Marsden, B. G., 2005c. Comets C/2005 W4, 2005 W5 (SOHO). Minor Plan.Electr. Circ. 2005-X14.Marsden, B. G., 2006. Comets C/1996 X3, 1996 X4, 1996 X5, 1997 B5, 1997B6, 1997 B7 (SOHO). Minor Plan. Electr. Circ. 2006-C49.Marsden, B. G., 2007. Comet P/1999 R1 = 2003 R5 = 2007 R5 (SOHO).Minor Plan. Electr. Circ. 2007-S16.Marsden, B. G., 2008a. Comets C/2007 Y8, 2007 Y9, 2007 Y10, 2008 A3(SOHO). Minor Plan. Electr. Circ. 2008-B61.Marsden, B. G., 2008b. Comets C/2008 G5, 2008 G6, 2008 H2, 2008 H3(SOHO). Minor Plan. Electr. Circ. 2008-L29.Marsden, B. G., 2008c. Comets C/2002 Q8, 2008 E4, 2008 F1 (SOHO). MinorPlan. Electr. Circ. 2008-F32.Marsden, B. G., 2008d. Comet C/1999 X3 = 2004 E2 = 2008 K10 (SOHO).19inor Plan. Electr. Circ. 2008-S49.Marsden, B. G., 2008e. Comets C/2001 D1 = 2004 X7 = 2008 S2 (SOHO).Minor Plan. Electr. Circ. 2008-S82.Meech, K. J., Hainaut, O. R., Marsden, B. G., 2004. Comets nucleus sizedistributions from HST and Keck telescopes. Icarus 170, 463–491.Meech, K. J., Svoreˇn, J., 2004. Using cometary activity to trace the physicaland chemical evolution of cometary nuclei. In: Festou, M. C., Keller, H. U.,Weaver, H. A. (Eds.), Comets II, pp. 317–335. University of Arizona Press.Meyer, M., 2002. Comets C/2002 C3 (SOHO). Minor Plan. Electr. Circ. 2002-C28.Mumma, M. J., 2008. Chemical diversity of organic volatiles among comets:An emerging taxonomy and implications for processes in the proto-planetarydisk. In: Kwok, S., Sandford, S. A. (Eds.), Proceedings IAU Symposium No.251: Organic Matter in Space, pp. 309-310. Cambridge University Press.Reach, W. T., Kelley, M. S., Sykes, M. V., 2007. A survey of debris trails fromshort-period comets. Icarus 191, 298–322.Reach, W. T., Lisse, C. M., Kelley, M. S., Vaubaillon J., 2007. Rocket effect forfragments and meteoroids of the split comet 73P/Schwassmann-Wachmann3. Bull. Amer. Astron. Soc. 39, 524.Rousselot, P., 2008. 174P/Echeclus: A strange case of outburst. Astron. &Astrophys. 480, 543–550.Rousselot, P., Petit, J.-M., Poulet, F., Sergeev, A., 2005. Photometric study ofCentaur (60558) 2000 EC and trans-neptunian object (55637) 2002 UX at different phase angles. Icarus 176, 478–491.Samarasinha, N. H., 2007. Rotation and activity of comets. Adv. Space. Res.39, 421–427.Schleicher, D. G., Birch, P. V., Bair, A. N., 2006. The composition of theinterior of comet 73P/Schwassmann-Wachmann 3: Results from narrowbandphotometry of multiple components. Bull. Amer. Astron. Soc. 38, 485.Schleicher, D. G., Bair, A. N., 2008. Compositional taxonomy of comets andthe unique cases of 96P/Machholz 1 and 73P/Schwassmann-Wachmann3. In: LPI Editorial Board (Eds.), LPI Contribution No. 1405: Asteroids,Comets, Meteors 2008 (CD-ROM), paper ID 8174.Scotti, J. V., Gleason, A. E., Montani, J. L., Read, M. T., 2000. Eight TNOsand Centaurs. Minor Plan. Electr. Circ. 2000-E64.Sekanina, Z., 1982. The problem of split comets in review. In: Wilkening, L.L. (Ed.), Comets, pp. 251–287. University of Arizona Press.Sekanina, Z., 1997. The problem of split comets revisited. Astron. & Astro-phys. 318, L5–L8.Sekanina, Z., 2002. Runaway fragmentation of sungrazing comets observedwith the Solar and Heliospheric Observatory. Astrophys. J. 576, 1085–1089.Sekanina, Z., 2007. Earth’s 2006 encounter with comet 73P/Schwassmann-Wachmann: Products of nucleus fragmentation seen in closeup. In: Valsec-chi, G.B., Vokrouhlicky, D. (Eds.), Near Earth Objects, our celestial neigh-bors: Opportunity and risk, pp. 211–220. Cambridge University Press.20ekanina, Z., Chodas, P. W., 2002. Comet 57P/du Toit-Neujmin-Delporte.IAU Circ. 7957.Sekanina, Z., Chodas, P. W., 2005. Origin of the Marsden and Kracht groupsof sunskirting comets. I. Association with comet 96P/Machholz and its in-terplanetary complex. Astrophys. J. Supp. Ser. 161, 551–586.Stansberry, J., Grundy, W., Brown, M., Cruikshank, D., Spencer, J., Trilling,D., Margot, J.-L., 2008. Physical properties of Kuiper Belt and Centaurobjects: Constraints from the Spitzer Space Telescope. In: Barucci, M. A.,Boehnhardt, H., Cruikshank, D.P., Morbidelli, A. (Eds.), The Solar Systembeyond Neptune, pp. 161–179. University of Arizona Press.Stevenson, R., Jedicke, R. 2007. An observational limit of decameter scalefragment mass loss from comets. Bull. Amer. Astron. Soc. 39, 525. 54.09Storm, S., Samarasinha, N., Mueller, B., Farnham, T., Fernandez, Y., Kidder,A., Snowden, D., A’Hearn, M., Harris, W., Knight, M., Morgenthaler, J.,Lisse, C., Roesler, F., 2006. Time variability of component C of the frag-mented comet 73P/Schwassmann-Wachmann 3. Bull. Amer. Astron. Soc.38, 504.Sykes, M. V., Walker, R. G., 1992. Cometary dust trails. I - Survey. Icarus 95,180–210.Sunshine, J. M., Groussin, O., Schultz, P.H., A’Hearn, M.F., Feaga, L.M.,Farnham, T.L., Klaasen, K.P., 2007. The distribution of water ice in theinterior of comet Tempel 1. Icarus 190, 284–294.Toth, I., Lamy, P. L., Weaver, H. A., 2003. Hubble Space Telescope obser-vations of the nucleus fragment 73P/Schwassmann-Wachmann 3-B. Bull.Amer. Astron. Soc. 35, 985.Toth, I., Lamy, P., Weaver, H. A., 2005. Hubble Space Telescope observationsof the nucleus fragment 73P/Schwassmann-Wachmann 3-C. Icarus 178, 235–247.Toth, I., Lamy, P., Weaver, H., A’Hearn, M., Kaasalainen, M., Lowry, S., 2006.Bull. Amer. Astron. Soc. 38, 489.Villanueva, G. L., Bonev, B.P., Mumma, M.J., Magee-Sauer, K., DiSanti, M.A., Salyk, C., Blake, G. A., 2006. The volatile composition of the splitecliptic comet 73P/Schwassmann-Wachmann 3: A comparison of fragmentsC and B. Astrophys. J. 650, L87–L90.Weaver, H. A., Lisse, C. M., Mutchler, M. J., Lamy, P., Toth, I., Reach,W. T., 2006. Hubble Space Telescope investigation of the disintegrationof 73P/Schwassmann-Wachmann 3. Bull. Amer. Astron. Soc. 38, 490.Weissman, P. R., Chesley, S. R., Choi, Y. J., Bauer, J. M., Tegler, S.C., Ro-manishin, W.J., Consolmagno, G., Stansberry, J. A., 2006. Motion of theactivity source associated with active Centaur 174P/Echeclus (60558). Bull.Amer. Astron. Soc. 38, 551. 21 able 1Known Split JFCsComet When3D/Biela 184016P/Brooks 2 1889, 199551P/Harrington 1994, 200157P/du Toit-Neujmin-Delporte 200269P/Taylor 191573P/Schwassmann-Wachmann 3 1995, 2001, 200679P/du Toit-Hartley 1982101P/Chernykh 1991, 2005108P/Ciffreo 1985120P/Shoemaker-Holt 1 1996141P/Machholz 2 1987, 1989174P/Echeclus 2006205P/Giacobini 1896, 2008D/1993 F2 Shoemaker-Levy 9 1992P/2004 V5 (LINEAR-Hill) 2004List adapted from work by Boehnhardt (2004).Note that 174P is also a Centaur. able 2Potentially Periodic SungrazersComets Group P Refs.C/1999 J6 = C/2004 V9 Marsden 5.49 1C/1999 M3 = C/2004 L10 Kracht 4.95 1,2C/1999 N5 = C/2005 E4 Marsden 5.66 3C/1999 N6 = C/2004 J4 or C/2004 J18 Kracht 4.81 1,2C/1999 R1 = C/2003 R5 = C/2007 R5 Kracht II 3.99 4C/1999 X3 = C/2004 E2 = C/2008 K10 (none) 4.22 5C/2000 O3 = C/2005 W4 Kracht 5.32 6C/2001 D1 = C/2004 X7 = C/2008 S2 (none) 3.78 7C/2002 Q8 = C/2008 E4 Kracht 5.52 8C/2002 R1 = C/2008 A3 Marsden 5.37 9C/2002 S11 = C/2008 G6 Kracht 5.54 10All comets are named SOHO. P = orbital period in years.References: 1 = Marsden (2004), 2 = Sekanina and Chodas (2005),3 = Marsden (2005b), 4 = Marsden (2007), 5 = Marsden (2008d),6 = Marsden (2005c), 7 = Marsden (2008e), 8 = Marsden (2008c),9 = Marsden (2008a), 10 = Marsden (2008b). ig. 1. Mosaic of comet 73P/Schwassmann-Wachmann 3 as observed by the SpitzerSpace Telescope over May 4 to 6, 2006, at a wavelength of 24 µ m. About threedozen fragments of this split comet are visible here, and almost every one has itsown cometary tail. The fragments themselves all lie on the comet’s projected orbit.Courtesy W. T. Reach of Caltech. ig. 2. Information about the fragments of 57P. At top are the 18 fragments dis-covered by Fernandez et al. (2002) on July 17, 2002. Each panel is 44 arcsec across.Short white segments indicate the location of each fragment. Note the wide varietyof morphologies and condensations. At bottom is a schematic showing the locationof each fragment with respect to the head - fragment A. The location of fragmentB (Marsden, 2002a) is also shown. ig. 3. Hubble Space Telescope image of fragment B of comet 73P, taken on April18, 2006 by Weaver et al. (2006). The fragment itself is the condensation at upperleft; the image was taken just after the fragment had shed several fragments of itsown. These are the dozens of condensations farther down the tail. The image isabout 25 arcsec across, and HST’s spatial resolution is just 8 km/pixel (i.e. 0.05arcsec/pixel). Courtesy H. A. Weaver of JHU APL. ig. 4. Detail of an image (after unsharp masking) of fragment B taken by Fuse et al.(2007) on May 3, 2006, with the Subaru telescope. The image is their Figure 1. Thefield size is 96 by 68 arcsec. Fragment B itself is in the upper left, and Fuse et al.(2007) report finding 54 fragments in the image. Compare the time and scale toFig. 3. White streaks are trailed stars. ig. 5. Images of comet 174P from February 24, 2006, adapted from Figs. 1 and3 in the work by Bauer et al. (2008). The left panel shows the R-band image fromthe Table Mountain Observatory 0.6-m telescope; the right shows the 24- µ m imagefrom the Spitzer Space Telescope. In each panel, the two arrows indicate the mainbody of Echeclus itself and the condensation of the coma. Note that the orientationsare slightly different; in the TMO image, equatorial north is up, while in the SSTimage, north is 21 ◦ to the right of up.Fig. 6. Spitzer images of four representative JFCs taken from the workby Reach et al. (2007a). Clockwise from top-left: 10P/Tempel 2, 48P/Johnson,129P/Shoemaker-Levy 3, and 67P/Churyumov-Gerasimenko. All show a linear trailpopulated by millimeter (and larger) scale grains, though with different meridionaland longitudinal variations. While some of the comet-to-comet variation is due toobserving geometry, there are some intrinsic differences in the trails.to the right of up.Fig. 6. Spitzer images of four representative JFCs taken from the workby Reach et al. (2007a). Clockwise from top-left: 10P/Tempel 2, 48P/Johnson,129P/Shoemaker-Levy 3, and 67P/Churyumov-Gerasimenko. All show a linear trailpopulated by millimeter (and larger) scale grains, though with different meridionaland longitudinal variations. While some of the comet-to-comet variation is due toobserving geometry, there are some intrinsic differences in the trails.