Show Summary Details

Page of

date: 28 October 2020

# Active Asteroids

• Henry HsiehHenry HsiehPlanetary Science Institute Tucson, Arizona Institute of Astronomy and Astrophysics Academia Sinica Taipei, Taiwan

### Summary

The study of active asteroids is a relatively new field of study in Solar System science, focusing on objects with asteroid-like orbits but that exhibit comet-like activity. This field, which crosses traditionally drawn lines between research focused on inactive asteroids and active comets, has motivated reevaluations of classical assumptions about small Solar System objects and presents exciting new opportunities for learning more about the origin and evolution of the Solar System. Active asteroids whose activity appears to be driven by the sublimation of volatile ices could have significant implications for determining the origin of the Earth’s water—and therefore its ability to support life—and also challenge traditional assumptions about the survivability of ice in the warm inner Solar System. Meanwhile, active asteroids whose activity appears to be caused by disruptive processes such as impacts or rotational destabilization provide exciting opportunities to gain insights into fundamental processes operating in the asteroid belt and assessing their effects on the asteroid population seen in the 21st century.

### Asteroids and Comets—The Classical Paradigm

Classically, asteroids and comets are considered to be distinct types of objects. Both belong to the category of small Solar System objects, which are essentially debris left over from the formation of the major planets, but they differ in several key aspects. In terms of appearance, a comet often has a distinctive fuzzy halo (known as a coma) around its central body (known as the nucleus), one or more tails of dust or gas streaming away from the nucleus, or both. The appearance of these features is typically referred to as cometary activity. Meanwhile, asteroids appear essentially as single points of light that look identical to stars, where the word “asteroid” itself means “star-like” (Herschel, 1801). In terms of their physical nature, comets are icy bodies whose comae and tails are the result of sublimation of volatile (i.e., icy) material when they approach closer to the Sun (Whipple, 1950, 1951; Sekanina, 1991), while asteroids are mostly composed of inert material such as rock, dust, or metal.

In terms of orbital dynamics, comets typically travel on highly elongated, often highly inclined (with respect to the plane of the Solar System) orbits that take them from the outer Solar System (beyond the orbit of Jupiter) to the inner solar system (in the vicinity of the Earth) and back again. Asteroids occupy much more circular, typically less inclined orbits that are mostly confined to the region between the orbits of Mars and Jupiter known as the main asteroid belt (also referred to simply as “the main belt” or “the asteroid belt”). Finally, most comets are believed to originate in the distant, cold Kuiper Belt and Oort Cloud (Oort, 1950; Fernández, 1980; Duncan et al., 1988, 1995; Levison & Duncan, 1997), beyond the orbit of Neptune, while asteroids are believed to have formed in place where they are currently found in the relatively warm main asteroid belt (e.g., Gradie & Tedesco, 1982; Petit et al., 2001; Bottke et al., 2005).

The dynamical distinction between asteroids and comets is often quantified using what is known as the Tisserand parameter, $TJ$, with respect to Jupiter. It can be calculated from an object’s semimajor axis, $aobj$, eccentricity, $eobj$, and inclination, $iobj$, using the following equation:

$Display mathematics$

where $aJ=5.2$ au is the semimajor axis of Jupiter. Comets coming from the outer Solar System into the inner Solar System typically have $TJ<3$ and are dynamically coupled to Jupiter, while most main-belt asteroids have orbits that keep them far from Jupiter’s gravitational sphere of influence and have $TJ>3$ (Krésak, 1979, 1980; Vaghi, 1973). In practice, despite its appealing simplicity, $TJ$ is an inexact dynamical classifier as it does not account for gravitational influences other than Jupiter and the Sun (e.g., the other major planets), or non-gravitational influences such as the radiative Yarkovsky effect and cometary outgassing, and also assumes Jupiter’s orbit to have zero eccentricity and inclination, which it does not. As such, caution should be exercised when evaluating the dynamical nature of objects with $TJ$ values extremely close to $TJ=3$ (e.g., Hsieh & Haghighipour, 2016).

#### “Transition” Objects

Planetary scientists have realized for some time that asteroids and comets may not be as distinct as they are classically portrayed. For instance, comets are not always active along their entire orbits, becoming inactive when their orbits carry them back into the outer Solar System where temperatures are too cold to drive sublimation. Repeated visits to the inner Solar System also have the potential to exhaust a comet’s supply of icy material, thus causing them to go dormant or even “extinct” (e.g., Hartmann et al., 1987; Krésak, 1987; Coradini et al., 1997), becoming so-called asteroids on cometary orbits (ACOs; e.g., Carvano et al., 2008; Licandro et al., 2006, 2008). Also, even when an object is active and thus could be classified as a comet, if that activity is extremely weak, it might escape detection by an insufficiently sensitive telescope, further complicating the task of observationally distinguishing comets and asteroids.

Physically, despite their classical characterization as dry, inert bodies, many asteroids have long been known to have contained water ice sometime in the past. Absorption features attributed to hydrated minerals have been detected on many primitive asteroids in both the infrared and visible portions of the electromagnetic spectrum (at 3 µm and 0.7 µm; Lebofsky, 1980; Lebofsky et al., 1981; Feierberg et al., 1985; Hasegawa et al., 2003; Vilas et al., 1994; Barucci et al., 1998; Rivkin et al., 2002), indicating that those objects once contained water ice, and suggesting that some asteroids might still contain ice. Hydrated minerals themselves have also been directly found in CI and CM carbonaceous chondrite meteorites that have been linked to C-class asteroids in the asteroid belt (e.g., Hiroi et al., 1996; Burbine, 1998; Keil, 2000). In recent years, direct spectroscopic detections of water vapor or water ice have been reported for a few large main-belt objects like dwarf planet (1) Ceres and asteroid (24) Themis (discussed further in “Observational Characterization of Activity” and “Observational Characterization of Inactive Nuclei.

In practice, as mentioned in “Asteroids and Comets—The Classical Paradigm”, the classical dynamical distinction between asteroids and comets, the Tisserand parameter, also has its complications. Effects such as non-gravitational orbit perturbations (such as the solar radiation–driven Yarkovsky effect or cometary jet activity), close planetary encounters, and dynamical resonances can sometimes cause asteroids to take on comet-like orbits and vice versa (e.g., Yeomans, 1994; Yeomans et al., 2004; Brož et al., 2006; Vokrouhlický & Farinella, 1998; Bottke et al., 2002; Hsieh & Haghighipour, 2016). Comet 2P/Encke is a notable example of an object that is observationally undisputed as a comet yet currently occupies a technically asteroid-like orbit ($TJ=3.03$), apparently due to gas-jet-driven dynamical evolution, complex secular resonances, or a combination of both (Marsden & Sekanina, 1973; Yeomans & Chodas, 1989; Steel & Asher, 1996; Fernández et al., 2002; Pittich et al., 2004; Levison et al., 2006).

Objects that have characteristics of both asteroids and comets have been (and sometimes still are) referred to as “transition objects” to convey the idea that they occupy a sort of “transition zone” between asteroids and comets. The term is somewhat potentially misleading, however, as it could imply that these objects are actually in the process of evolving from one type of object to the other. While this may sometimes be the case, it is not always true. Some objects are simply intrinsically both asteroid-like and comet-like. Contrary to the previously held classical paradigm that separated asteroids and comets into distinct groups, it has become increasingly common to refer asteroids and comets as occupying a continuum of small Solar System objects that can potentially possess a variety of combinations of physical and dynamical properties (Hsieh, 2017).

It should be noted that the multiple ways asteroids and comets can be classified have led to some disagreements within the scientific community about how to refer to so-called asteroid–comet continuum objects that share characteristics of both types of objects (i.e., which objects should be called “comets” and which should not). Some researchers prefer observational definitions (i.e., defining a comet as any object that has ever been seen to have a comet-like appearance, regardless of the cause) for its simplicity and lack of ambiguity. Perhaps largely because of this unambiguity, this is the criterion used by the International Astronomical Union’s Minor Planet Center, which maintains official catalogues of all known small Solar System bodies, when assigning asteroid or comet designations to newly discovered objects. Scientifically, however, other researchers prefer the physical definition (i.e., defining a comet as any object that exhibits activity due to the sublimation of icy material) for the added physical context provided by the use of the term “comet.” However, use of this definition requires some amount of scientific analysis (which can be—and sometimes is—inconclusive, especially soon after an object’s discovery when insufficient data may be available) to determine the source of a particular object’s activity. Still others choose to classify objects by their origins, considering only objects from the outer Solar System to be comets and anything originating in the inner Solar System to be asteroids. Even today, no real consensus exists. As such, the terms are effectively free to be used according to individual preferences, although for clarity, definitions should ideally be made clear in those individual cases. Unless otherwise specified, the term “cometary activity” in this article will hereafter refer to activity caused by the sublimation of ice, and “comet” will refer to any object exhibiting such activity, regardless of the object’s place of origin. Meanwhile, “comet-like activity” will refer to the presence of any features typically associated with cometary activity (i.e., comae or tails), regardless of the actual source of that activity.

### Comet 133P/Elst-Pizarro—The First Active Asteroid

When it was discovered to have a comet-like appearance in 1996 (Elst et al., 1996), Comet 133P/Elst-Pizarro (hereafter referred to as EP) represented a new kind of object on the comet–asteroid continuum, although this was perhaps not fully appreciated at the time. It was quickly recognized as being previously discovered in 1979 as asteroid 1979 OW7, but in 1996, clearly exhibited a narrow comet-like dust tail stretching over more than 200,000 km in length. Strangely though, its orbit was fully contained within the main asteroid belt, had a low inclination, and was mostly circular, with a Tisserand parameter of $TJ=3.16$, all of which pointed to the object being dynamically asteroidal.

The discovery of EP’s cometary activity was surprising because instead of being an asteroid on a comet-like orbit, which could be naturally explained as a previously active comet that has become dormant, the object was a comet on an asteroid-like orbit. As mentioned earlier, Comet 2P/Encke technically has an asteroid-like $TJ$ value, but only just barely. It is also an extremely active comet that, despite its asteroid-like $TJ$ value, still has a highly elongated orbit typical of other comets, making a scenario in which cometary jet activity has managed to push an ordinary comet onto a slightly asteroidal orbit relatively reasonable. Meanwhile, EP’s observed activity was extremely weak by comparison and hardly seemed strong enough to drive it onto such a thoroughly asteroidal orbit. An impact with another asteroid could potentially have caused the observed ejection of dust into space from EP, although dust modeling of the tail indicated that the dust was most likely released over the course of several weeks or months prior to the comet’s discovery, rather than in a single instantaneous burst, which is what would be expected if the observed dust ejection was caused by an impact (Boehnhardt et al., 1996).

Boehnhardt et al. (1998) eventually proposed that the excavation and subsequent sublimation of “icy dirt” on EP’s surface by a recent impact could explain EP’s apparent comet-like dust emission behavior, although how such material could remain icy over billions of years at main-belt temperatures was left unexamined. Meanwhile, Tóth (2000) returned to the impact hypothesis, proposing that a cloud of debris from nearby asteroid (427) Galene could have been responsible for the observed prolonged dust emission event from EP, where such an event could have either been mimicked by multiple successive impacts or caused by a single large impact that seismically excited the object, leading to lofting of dust grains (which then were able to escape due to the object’s low gravity) well after the initial impact.

New observations in 2002 showing that EP had become active again (Figure 1; Hsieh et al., 2004; Lowry & Fitzsimmons, 2005) placed strong additional constraints on the mechanisms that could plausibly explain the object’s activity. The discovery of new activity made impact-related hypotheses far less tenable as two activity-generating impacts occurring twice on one specific object within the span of six years seems highly unlikely, given that such behavior is not observed at anywhere near that rate on other objects. On the other hand, cometary activity (i.e., activity caused by the sublimation of volatile material) could explain the repeated activity much more naturally, since comets commonly show repeated activity interspersed with periods of inactivity as their elongated orbits take them from cold to warm regions of the solar system and back again.

#### Main-Belt Comets

The revelation that EP could be a bona-fide comet led to two possibilities: (1) EP could be an ordinary comet from the outer Solar System that serendipitously evolved onto a main-belt asteroid-like orbit, or (2) EP could be an ordinary main-belt asteroid that happens to contain sufficient near-surface ice to drive cometary activity (Hsieh et al., 2004). In the first scenario (the “lost comet hypothesis”), objects such as EP might be expected to be rare, given the difficulty of finding plausible dynamical pathways from a classical comet-like orbit to an EP-like orbit (e.g., Ipatov & Hahn, 1997; Fernández et al., 2002). In the second scenario (the “icy asteroid hypothesis”), if EP is an ordinary asteroid, EP-like objects could be common and might be found by a search targeting the right asteroids with the right observations at the right time.

Seeking to test the prediction of the latter hypothesis that other EP-like objects should exist if EP is an ordinary main-belt asteroid, Hsieh and Jewitt undertook an dedicated observational campaign targeting small EP-like outer main-belt asteroids in order to search for cometary activity (Hsieh, 2009). In 2005, observations by the eight-meter Gemini-North telescope on Maunakea in Hawaii as part of this search successfully uncovered cometary activity in asteroid (118401) 1999 RE70, which was subsequently also given the cometary designation 176P/LINEAR (Hsieh et al., 2006). Coincidentally, this discovery was a month after the serendipitous discovery of Comet P/2005 U1 (Read), now designated 238P/Read, by the Spacewatch survey (Read et al., 2005). Dust modeling of both objects indicated that their dust emission activity likely occurred as prolonged events, consistent with sublimation-driven activity and inconsistent with impact-driven dust emission (Hsieh et al., 2009, 2011) The existence of three known EP-like objects in the asteroid belt up to that point and their discovery from relatively limited data (implying that many others could exist) led Hsieh and Jewitt (2006) to conclude that a new type of comet, which they dubbed “main-belt comets” (MBCs), had been discovered.

#### Disrupted Asteroids

Following the recognition of MBCs as a new class of cometary objects and the discovery of a fourth MBC, 259P/Garradd, in 2008 (Garradd et al., 2008), the fifth active object discovered orbiting in the asteroid belt represented yet another new addition to the inventory of types of active objects in the solar system. P/2010 A2 (LINEAR) (since re-designated 354P/LINEAR) was discovered in 2010 and received a cometary designation due to its comet-like diffuse appearance (Birtwhistle et al., 2010). High-resolution Hubble Space Telescope (HST) images of the object showed structures unlike any ever seen before for other comets (Figure 2a; Jewitt et al., 2010). Instead of being sublimation-driven activity from an icy asteroid, the observed dust emission appeared to be the result of an impact on the object by another asteroid (Jewitt et al., 2010; Snodgrass et al., 2010). This discovery was followed in the same year by the discovery of another object exhibiting unusual comet-like activity that was also ultimately attributed to an impact event, this time on the large, known main-belt asteroid (596) Scheila (Figure 2b; Jewitt et al., 2011; Bodewits et al., 2011; Ishiguro et al., 2011).

In 2013, an even stranger comet-like object was discovered. P/2013 P5 (PANSTARRS) (since re-designated 311P/PANSTARRS) was seen in high-resolution HST images to exhibit multiple dust tails (Figure 3), where this unusual morphology was attributed to rotational mass-shedding, or surface material essentially being flung off into space due to the extremely rapid rotation of the object (Jewitt et al., 2013; Scheeres, 2015). Specifically, it was suggested that the object could have been spun up over time by a radiative torque effect known as the Yarkovsky–O’Keefe–Radzievskii–Paddack (YORP) effect, which occurs when the reflection and re-radiation of solar radiation by an asymmetric Solar System object causes its spin state to change (Rubincam, 2000). Mass loss, perhaps in the form of sporadic avalanches or landslides, would then be expected if and when an object is spun up so fast that the centrifugal force caused by the object’s rotation exceeds the self-gravity and internal cohesion and strength holding it together (Jewitt et al., 2013).

Objects like 311P/PANSTARRS, 354P/LINEAR, and (596) Scheila that exhibit comet-like mass loss that is attributed to processes other than the sublimation of volatile material are now often referred to collectively as disrupted asteroids (see Hsieh et al., 2012), although the term “activated asteroid” is sometimes also used. While impacts and rotational disruptions are currently the most commonly inferred causes of disrupted asteroid activity, other activity drivers may also act on specific objects, such as thermal fracturing, radiation pressure sweeping, electrostatic dust levitation, or binary interaction (Hainaut et al., 2014; Jewitt et al., 2015; Ye et al., 2019).

### Current State of the Field

Together, MBCs and disrupted asteroids comprise the group of objects now collectively known as active asteroids, which are defined as objects that have been observed to exhibit comet-like activity at least once but have asteroid-like orbits with $TJ>3.08$ and semimajor axes smaller than that of Jupiter (i.e., $a) (Jewitt et al., 2015). Within this classification, MBCs are active asteroids that have orbits fully confined to the main asteroid belt and whose activity has been determined to be at least partially due to sublimation of volatile ices (where in many cases, activity may also be triggered or facilitated by other processes such as impacts or rotational destabilization), and disrupted asteroids are active asteroids whose active events are due to mechanisms other than sublimation. The $TJ$ cutoff defined by Jewitt et al. (2015) differed slightly from the theoretical dividing line of $TJ=3$ between asteroids and comets (see “Asteroids and Comets—The Classical Paradigm”) because there are a number of Jupiter-family comets (JFCs; a subset of the classical comet population that is believed to originate in the outer solar system and has orbits with $2) that can occasionally occupy orbits with $TJ$

values slightly larger than 3, but that are not believed to originate in the inner Solar System like other active asteroids that have $TJ»3$ (see Hsieh & Haghighipour, 2016). This higher $TJ$ cutoff value excludes these objects. Using this definition, there are currently about 30 known active asteroids (Figure 4), a little over half of which are considered likely MBCs, and the remaining objects are likely disrupted asteroids, active asteroids whose activity mechanisms are unknown, or near-Earth active asteroids.

MBCs have attracted considerable interest due to their significance for the field of astrobiology. The Earth and the other terrestrial planets formed inside the snow line (i.e., the distance from the Sun beyond which temperatures were cold enough in the protosolar disk for water to condense as ice). As a result, they are expected to have formed mostly from dry material and should be dry themselves (e.g., Boss et al., 1998), making the origin of their current inventory of water and other volatile materials an open scientific question. Due to their icy composition, comets from the outer Solar System colliding with the Earth early in its history were initially suspected as a likely source of terrestrial volatiles. However, most measurements of deuterium-to-hydrogen (D/H) ratios in classical comets made to date (an admittedly small sample) appear to be inconsistent with ocean water (e.g., Altwegg et al., 2015), suggesting that an alternative source is required. Meanwhile, dynamical studies have suggested that objects from the outer asteroid belt region of the Solar System, where most MBCs are now found, could have impacted the Earth in sufficient quantities to provide significant amounts of water and other volatiles (e.g., Morbidelli et al., 2000, 2012; O’Brien et al., 2018).

While water ice is thermally unstable against sublimation over long timescales (i.e., over the dynamical lifetimes of most main-belt asteroids) at the temperatures found in the main asteroid belt, it is believed that ice could have been preserved over billions of years in MBCs in sub-surface reservoirs. In this scenario, present-day sublimation-driven activity is triggered by an impact or rotational destabilization event that removes some of the surface material insulating the sub-surface ice reservoir and exposing it to enough solar heating to drive sublimation, thus giving rise to activity (e.g., Hsieh et al., 2004; Schörghofer, 2008, 2016; Capria et al., 2012).

For their part, disrupted asteroids offer opportunities to study an array of processes that have long been studied using numerical modeling or laboratory experiments (e.g., Housen & Holsapple, 2003; Holsapple, 2009; Hirabayashi, 2015; Jutzi et al., 2015), but rarely in real-time under real-world conditions. As such, disrupted asteroids present invaluable opportunities to test the results of these models and experiments against real-world events. Studies of physical disruptions can also give insights into the internal compositional or structural properties of disrupted objects either by providing opportunities to directly observe excavated subsurface material (e.g., Bodewits et al., 2014) or use numerical modeling of the physics of disruptions to infer internal structural properties (e.g., Hirabayashi et al., 2015). Finally, appropriate characterization of the debiased rates of impact and rotational disruptions in the main asteroid belt will allow Solar System scientists to better understand its collisional environment and how collisions and rotational disruptions affect its size distribution (e.g., Hsieh, 2009; Jacobson et al., 2014).

As active asteroids became increasingly accepted as a new type of Solar System body and not simply a collection of individual oddball objects, the classification was expanded slightly to incorporate three unusual near-Earth objects—(2201) Oljato, (3200) Phaethon, and 107P/(4015) Wilson-Harrington—that have exhibited evidence of being active in the past and have orbits with $TJ>3.08$, formally making them members of the active asteroid population. Hereafter, this paper will focus on active asteroids in the main asteroid belt, but for more information on these objects, the reader is invited to refer to Jewitt et al. (2015) and references within.

#### Discovery

A major need for the advancement of active asteroid science is simply the discovery of more objects. A larger known population of MBCs will help with developing a better understanding of the abundance and distribution of MBCs in the asteroid belt, which in turn has implications for better constraining the total volatile content of both the present-day and primordial asteroid belt. More thorough sampling of the range of physical and dynamical properties of MBCs will also allow for an improved understanding of the conditions required to make MBC activity possible. Meanwhile, discovering more examples of impact and rotational disruption events will broaden the range of asteroid compositional types, sizes, and shapes that are represented, providing opportunities to determine how each of these parameters specifically affect disruption behavior. An improved understanding of the number and distribution of disruption events in the asteroid belt should also provide valuable insights into the size and distribution of the unseen population of small impactors (which are too faint to see from Earth) that may dominate the main belt in terms of total numbers, though likely not in mass, as well as whether rotational disruptions can be meaningfully correlated with any other circumstances such as family membership or highly collisionally active environments.

While a targeted campaign to observe specific objects to search for activity successfully discovered the third known MBC, 176P/LINEAR (Hsieh, 2009), such campaigns are time consuming to conduct and difficult to sustain. Untargeted surveys or searches of archival data are more realistic long-term and scalable options for conducting large-scale searches for active asteroids. Past and ongoing efforts along these lines include searches of archival wide-field image data from MegaCam on the 3.6-meter Canada-France-Hawaii Telescope in Hawaii and the Dark Energy Camera (DECam) on the 4-meter Blanco Telescope in Chile (Gilbert & Wiegert, 2009; Sonnett et al., 2011; Chandler et al., 2018), catalogued photometry from the Minor Planet Center (Cikota et al., 2014), and Pan-STARRS1 and Palomar Transient Factory survey data (Waszczak et al., 2013; Denneau et al., 2015; Hsieh et al., 2015). As of 2019, active asteroids are being discovered at the rate of about one to three each year, with the Pan-STARRS1 survey telescope being particularly prolific in terms of discovering these objects. Scheduled to come online in late 2022, the Large Synoptic Survey Telescope (LSST) should increase the discovery rate of new active asteroids (including both previously completely unknown active objects as well as activity associated with previously known objects) by perhaps an order of magnitude more (Jones et al., 2009), which should help to address many key questions regarding these objects.

#### Observational Characterization of Activity

In addition to discovery, another high priority for active asteroid research is observational characterization to determine the physical properties of known and newly discovered objects and their associated activity. At the most fundamental level, such characterization is needed to determine the most likely source of activity for an active asteroid in order to determine whether it should be classified as a MBC or disrupted asteroid; however, even once sources of activity are identified, more detailed physical characterization is still needed to determine other characteristics of the activity.

The most straightforward way to determine whether a particular object’s activity is sublimation driven would be to directly detect sublimation products in the object’s coma using spectroscopy. Such spectroscopic searches for sublimation products have been attempted for many active asteroids, but to date, most have been unsuccessful. The sole exception is dwarf planet (1) Ceres, for which H2O outgassing was detected on multiple occasions by the Herschel Space Telescope (Küppers et al., 2014). As a dwarf planet with a diameter of almost 1,000 km, however, Ceres clearly occupies a very different physical regime (e.g., with potential geological activity and perhaps a differentiated interior) than most other active asteroids (which typically have diameters of a few km or less). It should also be noted that observable dust emission has never been observed for Ceres as it has for almost all of the other active asteroids. Similar Herschel observations of MBCs 176P and 358P have not resulted in any similar definitive detections of water outgassing (de Val-Borro et al., 2012; O’Rourke et al., 2013).

For other active asteroids observed with ground-based telescopes, searches have typically focused on the (0-0) emission band of CN at 3883 Å, as this is typically a strong emission feature that is relatively easily observed by ground-based telescopes and so is often used as a proxy for estimating water production rates in classical comets. Despite often employing large ground-based telescopes, including the 10-meter Keck telescope in Hawaii, the twin eight-meter Gemini telescopes in Hawaii and Chile, the eight-meter Very Large Telescope (VLT) in Chile, and the 10.4-meter Gran Telescopio Canarias (GTC) in the Canary Islands, none of these searches have been successful to date (Snodgrass et al., 2017). Rather than indicating the absence of gas emission, however, it is believed that these non-detections simply indicate that sublimation rates are too weak to be detected with currently available facilities. It is also possible that while CN serves as a useful tracer of sublimation in classical comets, it may be depleted in MBCs (e.g., Prialnik & Rosenberg, 2009), suggesting that future searches should focus directly on H2O sublimation in such objects instead, rather than searching for proxy gas species.

Practically speaking, it is difficult to envision researchers being able to characterize the detailed chemical or isotopic composition of a MBC in the foreseeable future without a spacecraft visit. An intriguing possibility has recently been raised, however, that some objects that currently have JFC-like orbits may actually have originated in the asteroid belt (Fernández & Sosa, 2015; Hsieh & Haghighipour, 2016). If true, these objects may represent bona fide samples of main-belt ice that actually come close enough to the Earth and Sun where they can be spectroscopically characterized in far more detail than any of their main-belt counterparts. Further dynamical and observational analyses of these objects are needed to better assess the likelihood that they actually originated in the main belt, but if this hypothesis proves to be correct, it could potentially transform MBC research.

Given that direct spectroscopic detection of gas emission by MBCs as a means for confirming sublimation-driven activity appears to be currently out of reach, astronomers must use indirect means for inferring the sources of activity observed for active asteroids. Finson-Probstein-style dust modeling (Finson & Probstein, 1968) is a frequently used tool for this purpose, where trajectories of ejected dust particles under the influence of solar gravity and radiation pressure are determined and then matched against observations to infer grain size distributions and emission event durations (e.g., Hsieh et al., 2004; Licandro et al., 2013; Moreno et al., 2013, 2016; Agarwal et al., 2016; Kim et al., 2017). As discussed in “Comet 133P/Elst-Pizarro – The First Active Asteroid” in the case of EP, long-duration emission is considered indicative of a likely sublimation-driven emission event, while impulsive (i.e., short-duration) emission is considered more likely to be the result of a disruptive event such as an impact or rotationally triggered mass loss event.

Meanwhile, as also discussed in the case of EP, recurrent activity during a subsequent orbit passage (where most MBCs are seen to become active near perihelion) is considered an extremely strong indicator of sublimation-driven activity (see Hsieh et al., 2012). Use of this technique normally requires waiting for up to another full orbit period (~5 to 6 years for most MBCs) following the initial discovery of an object’s activity, and so usually cannot be applied immediately to determine the source of an active asteroid’s activity. Serendipitous archival observations of an object, if obtained at the right point in a previous orbit, can sometimes be used to confirm the appearance of activity on two distinct occasions (e.g., Hui & Jewitt, 2015), but such cases are rare.

Dust modeling can also potentially constrain other useful parameters, such as dust production rates, grain-size distributions, emission directionality (if any), ejection velocities, emission start times, and so on (e.g., Ishiguro et al., 2011, Jewitt et al., 2014). Furthermore, for objects that exhibit recurrent activity, independent determinations and comparison of these parameters for each active apparition have the potential to provide useful insights into activity attenuation for MBCs. In practice, however, unique solutions for all of these parameters are not always possible to achieve, as the large number of potential free parameters in dust models means that they are almost always under-constrained by available observational data. To some extent, potential degeneracies between model solutions can be resolved with data sets containing high-signal-to-noise images obtained over time (e.g., several weeks or months, to allow for viewing geometries to change and for the morphology of ejected dust to continue to evolve; e.g., Hsieh et al., 2009, 2011), making campaigns to obtain such data sets very useful for active asteroid characterization.

#### Observational Characterization of Inactive Nuclei

Physical characterization of the nuclei of active asteroids (i.e., the central body from which activity originates) is another important component of efforts to better understand these objects. Determination of an object’s nucleus size provides a means for conducting sensitive searches for recurrent activity (i.e., by measuring an object’s brightness, comparing it to the object’s predicted brightness if inactive, and searching for photometric excesses that could indicate the presence of unresolved activity; e.g., Hsieh et al., 2014), where identifying activity as early as possible for a newly reactivated object is useful for triggering follow-up observations to chronicle the rise of that activity. Knowing an object’s nucleus size also allows for precise measurements to be made of ejected dust masses (e.g., MacLennan & Hsieh, 2012; Hsieh et al., 2018b), which can then be used as an independent constraint on dust production rates to be compared with dust modeling results.

Given that MBC activity is intimately tied to composition (i.e., since the existence of sublimation indicates the presence of ice), it is also useful to determine compositional properties (e.g., taxonomic types) for active asteroid nuclei in order to better understand the correlation between surface composition and activity and to try and identify the expected properties of other icy main-belt objects (e.g., Licandro et al., 2011). Direct spectroscopic detections of absorption features attributed to surface ice frost have actually been reported for asteroids (24) Themis and (90) Antiope (Rivkin & Emery, 2010; Campins et al., 2010; Hargrove et al., 2015), which are both members of the Themis asteroid family to which several other MBCs also belong (see Hsieh et al., 2018a). The small sizes of all of the currently known MBC nuclei (diameters of < 5 km) compared to both of these objects (diameters of > 100 km), however, mean that they are also much fainter and more difficult to observe, making searches for similar ice features on MBC nuclei challenging to say the least.

For both main-belt comets and disrupted asteroids, determinations of rotational properties are also highly desirable. Determination of rotation rates can help determine the plausibility of destabilization by fast rotation at close to an object’s critical limit being a significant factor in driving or helping to drive dust emission (e.g., Hsieh et al., 2004; Sheppard & Trujillo, 2015; Kleyna et al., 2019). Orbital obliquity (related to an object’s pole orientation in space) is also a key parameter in thermal models on which subsurface ice survivability has a strong dependence (e.g., Schörghofer, 2008).

#### Theoretical and Computational Analyses

In addition to discovery and observational characterization efforts, various computational analyses play key roles in research on active asteroids. These include dynamical analyses, thermal modeling, and disruption modeling. A key question following the discovery of the first main-belt comet was whether it was possible that the object was simply a classical comet from the outer solar system that had somehow dynamically evolved onto a main-belt-like orbit (see “Comet 133P/Elst-Pizarro—The First Active Asteroid”). While subsequent dynamical analyses showed that EP and many other MBCs are likely native to the main belt (e.g., Ipatov & Hahn, 1997; Fernández et al., 2002; Jewitt et al., 2009; Haghighipour, 2009; Hsieh et al., 2013), more recent dynamical integrations show that there is a non-negligible possibility that some high-eccentricity, high-inclination MBCs could in fact be JFCs that have been recently temporarily captured onto main-belt-like orbits (Hsieh & Haghighipour, 2016). There have also been suggestions that icy outer Solar System objects may have been implanted in the asteroid belt very early on in the Solar System’s history as a consequence of planetary migration (e.g., Levison et al., 2009; Walsh et al., 2011). The exact rates at which the outer Solar System object contamination of the main belt suggested by these various works might occur and whether any of the currently known MBCs is actually the result of one of these processes remains unclear, however. These findings nonetheless highlight the need for continued dynamical analysis of MBC origins.

Another dynamical aspect of active asteroids, and MBCs in particular, that has attracted attention is their association with dynamical asteroid families. Dynamical asteroid families are groups of asteroids clustered in orbital element space that are believed to have originated in catastrophic fragmentations of larger parent bodies sometime in the past (Nesvorný et al., 2015). Almost all MBCs are associated with asteroid families (Hsieh et al., 2018a), some of which are relatively young (e.g., millions of years old rather than billions of years old; Nesvorný et al., 2008; Novaković et al., 2012). This finding suggests that young asteroid family members could be more likely than other asteroids to become active MBCs, perhaps because they may contain more near-surface ice than other asteroids due to their origin in the recent fragmentation of icy parent bodies.

Family membership is also significant because members of a specific family are generally believed to share similar compositions, having all originated from the same parent body (e.g., Masiero et al., 2015). So, if an icy MBC is found to be a member of a family, other members could also be icy, meaning that targeted searches of other members of families to which MBCs have been found to belong could reveal other MBCs. Many more MBCs and young families need to be discovered, however, to assess the robustness of this finding and to properly judge its implications for using the MBC population to estimate the total ice content of the main asteroid belt.

Consideration of thermal conditions, properties, and evolution comprise another important component of active asteroid research, and are important for understanding the nature of ice, or the absence thereof, in asteroids. Thermal modeling by Schörghofer (2008, 2016) showed that the survivability of ice in a main-belt asteroid depends strongly on the object’s obliquity and latitude on the object, but that subsurface ice could be expected on a MBC-like EP at depths of a meter or less, making it well within the reach of a small impactor, a conclusion similar to that reached by Capria et al. (2012). Meanwhile, modeling by Prialnik and Rosenberg (2009) indicated that even if common cometary ices are initially present, many of them may be severely depleted in main-belt objects, leaving water as the primary remaining volatile material, a finding with significant implications for efforts to detect sublimation products in MBC activity (see “Observational Characterization of Activity”).

Lastly, analytic, numerical, and even laboratory modeling of disruptions provides a framework for probing the internal structure and composition of active asteroids as well as the nature of their disruptions. By modeling the dependence of certain observable properties (e.g., mode of disruption, i.e., low-level mass-shedding versus complete disintegration, and the size distribution, axis ratios, and relative velocities of any resulting fragments) of rotationally disrupted asteroids (e.g., Figure 5) on key initial properties such as the amount of internal cohesion and initial spin period; and comparing the results of this modeling to observed data, one can infer the ranges of initial properties that the original body may have had (e.g., Scheeres, 2015; Hirabayashi, 2015; Hirabayashi et al., 2014, 2015; Zhang et al., 2018). For impact disruptions, numerical and laboratory simulations of asteroid impacts can provide information about, for example, how total ejecta mass scales with the size of an impactor for given target material, the expected morphology of an oblique impact, or how icy material responds to impacts (e.g., Kawai et al., 2010; Housen & Holsapple, 2011), which can then be used to interpret observations of real-world impact disruption events and assess the plausibility of MBC activity triggering scenarios (e.g., Ishiguro et al., 2011; Jewitt et al., 2011; Haghighipour et al., 2016, 2018).

#### In Situ Characterization

A final method of studying active asteroids that has actually yet to be employed and so about which there is presently nothing to say in terms of actual results is in situ characterization (i.e., characterization by a spacecraft in close proximity to an active asteroid). Several NASA and European Space Agency (ESA) missions targeting MBCs have been proposed to date (Meech & Castillo-Rogez, 2015; Jones et al., 2018; Snodgrass et al., 2018), although so far, none have been selected for flight. Interest in such missions remains high among scientists interested in MBCs, however, given the transformative insights likely to result from such a mission if other recent comet missions are any indication (e.g., A’Hearn et al., 2011; Fulle et al., 2016; Barucci & Fulchignoni, 2017).

High priorities for in situ characterization of a MBC and its activity by a spacecraft that cannot be performed by Earth-bound facilities include mapping of surface structure, topography, geology, and mineralogy (with a focus on hydrated minerals and ices), detailed compositional characterization of the object’s gas and dust emission, in situ monitoring of diurnal and orbital activity cycles, precise determination of the object’s global properties such as size, shape, density and thermal properties, characterization of the object’s plasma environment and magnetic field (if any), and determination of the object’s D/H ratio and ratios of other isotopes such as oxygen (see Snodgrass et al., 2018). As many of these priorities require a target object to be active when visited by the spacecraft, target candidates have typically been limited to the small number of MBCs that have exhibited reliable repeated activity, severely limiting the number of available trajectories for such a mission. In the future, discovery of many more MBCs by LSST may alleviate this problem somewhat, although confirmation of recurrent activity will still take time to achieve for most new discoveries as discussed in “Observational Characterization of Activity.”

While in situ characterization of the activity of a disrupted asteroid would likely be similarly revelatory as for a MBC, the unpredictability of active events coupled with their typically short duration means that it is probably nearly impossible to design a mission capable of catching a disrupted asteroid while active. That said, a mission to a disrupted asteroid known to have been active in the past would still be extremely enlightening, as the object’s surface will be known to have been recently disturbed, likely bringing normally hidden interior material up to the surface where it would be directly observable. In addition to characterizing freshly excavated interior material on a disrupted asteroid, a mission would also be able to make detailed observations and measurements of a disrupted asteroid’s shape, topography, and spin state, all of which would likely provide significant insights into the nature of the object’s previously observed disruption.

### Conclusions and Future Research

Perhaps the main conclusion that can be drawn from active asteroid research in the 2000s and 2010s is that the present-day asteroid belt is a far more active and dynamic place than one might have once imagined from the classical view of it. Rather than being completely inert bodies, asteroids have been found that exhibit sometimes spectacular observable activity (e.g., Figure 6) as a result of processes such as volatile sublimation, impacts, and rotational destabilization, or even as a result of multiple processes operating simultaneously. With still relatively few of the objects currently known, continuing to discover and characterize more active asteroids remains a high priority in order to improve the ability of researchers to discern trends in the properties of these objects, account as completely as possible for the range of active behaviors exhibited by objects in the population, and reach the best possible understanding of the scientific implications of the derived global properties of the different types of active asteroids.

For example, a clearer picture of the abundance and distribution of active MBCs will enable better estimates to be made of the total number of icy bodies in the asteroid belt, while similarly clarifying the abundance and distribution of impact disruption events should provide insights into the size and distribution of the unobservable small impactor population. These are only possible, however, if sample sizes are large enough for those abundance and distribution determinations to be considered statistically robust. Current and future wide-field surveys (especially LSST) will likely account for most new discoveries (as they have already been doing for the last decade), but development of a more diverse portfolio of automated mechanisms for detecting activity (e.g., Hsieh et al., 2014) than is currently typically used would be very helpful in maximizing discovery rates.

Despite enormous progress over the last decade in particular, many aspects of active asteroids still remain poorly understood. A successful spectroscopic detection of gas species in an active MBC has yet to be achieved and may continue to prove elusive until new facilities, such as the next generation of extremely large (25- to 40-meter) telescopes become available, or until an active MBC can be visited by a spacecraft. On a similar note, virtually nothing is known about the composition of MBC ice and gas emission (i.e., whether they in fact consist nearly exclusively of water, as predicted by Prialnik & Rosenberg, 2009 or contain noteworthy levels of other volatile species). Since various ratios of volatile species have already been used to define taxonomic groups for comets that might be related to formation regions (e.g., A’Hearn et al., 1995; Mumma & Charnley, 2011; Cochran et al., 2012; Dello Russo et al., 2016), it would be extremely interesting to determine the abundances of volatile species in MBCs in order to assess whether they are similar to previously identified comet taxonomic groups, or if they exhibit intrinsic differences that could indicate a different source region, such as the asteroid belt. In terms of isotopes, measurement of the D/H ratio (as well as other isotopic ratios for other species such as oxygen) in MBCs is considered an extremely high priority from an astrobiology perspective (see “Current State of the Field”).

While continuing (and intensifying, where possible) efforts to identify and characterize new disruptions will perhaps do the most for advancing the study of disrupted asteroids, other research needs include acquisition of lightcurve data for determining the rotation rates and axis ratios of suspected rotationally disrupted asteroids to assess whether those properties are consistent with rotational disruption. Continued numerical modeling analyses of individual rotational disruption events (addressing such issues as what the characteristics of objects are before and after disruptions, and what types of morphologies are expected from different types of destabilization events) will also be extremely useful. Meanwhile, more detailed statistical studies of the expected rates of both impact and rotational disruption events will help provide a framework for interpreting observed incidence rates for such events (once enough events have been discovered so that these rates are statistically significant), enabling the evaluation and refining of initial model parameters.

The time since the discovery of the first active asteroid in 1996 has been extraordinarily exciting for Solar System scientists as it has seen the rise of an essentially new field of study involving objects in a region of the Solar System that many would likely have previously characterized as fairly well studied and understood. Given significant progress on other related fronts, such as the development of new wide-field surveys and enhancements of existing surveys—and ongoing improvements in computational power and techniques that allow for ever-more detailed numerical modeling studies—the next decade should see even greater advances. This story has just begun to unfold.