The existence of asteroids was first suspected in the late 18^th Century. Astronomers noted that the Titus-Bode law, an empirical relationship between the size of successive planets' orbits around the Sun, suggested that there should be a 'missing' planet between Mars and Jupiter, orbiting the Sun at an average distance of 2.8 astronomical units. In 1801 the Italian astronomer Giuseppe Piazzi discovered a small body orbiting the Sun at 2.77 AU. This body, later named Ceres, was evidently too small to be a true planet, and when 3 further objects (Pallas, Juno and Vesta) were discovered in similar orbits over the following 7 years, it was recognized that there was a new class of small bodies orbiting in the 'empty' region. Known as asteroids, from the Latin for 'star-like' in reference to their appearance in telescopes, they mostly orbit in a 'main belt' some 2 to 4 AU from the Sun. The advent of astrophotography in the late 19^th Century greatly increased the discovery rate of asteroids to the extent that many modern astrophotographs inadvertently image dozens or even hundreds of them.

For many years asteroid research was neglected in favour of the more spectacular advances being made in stellar astronomy and cosmology, but since the 1960s there has been a resurgence of interest in the smaller bodies of the solar system. Better observational techniques and the possibility of actually examining asteroids in situ have led to a rapid advance in knowledge regarding asteroids. In 1991 the first high-resolution image of an asteroid, the 18 km by 9 km Gaspra, was obtained by the Galileo probe on its way to Jupiter (Fig 5.1). A number of good academic references on asteroid science have been published, notably the University of Arizona Press series [Gehrels79, Gehrels89, Gehrels94] whilst the introductory volume by Kowal [Kowal96] provides a particularly good overview.

Asteroids are catalogued by the Minor Planet Centre at Cambridge, MA. When an asteroid is discovered (which effectively means that it is observed for long enough that a rough orbit can be computed) then it is assigned a temporary designation. This consists of the year of discovery followed by a 2-letter code indicating the half-month of discovery and the order of discovery within that period. Thus the first asteroid to be discovered in 1995 would be 1995AA, the second 1995 AB and so on until the second half of the month when the sequence would continue 1995BA, 1995 BB etc; additional numerical suffixes are used if required. Once an asteroid has been observed for long enough to determine an accurate orbit then it is assigned a permanent number. This usually takes several years, and thus an asteroids number may bear little relationship to its actual discovery order. Many asteroids are also named once they are numbered, usually at the discretion of the initial discoverer.

Not all asteroids are confined to the main belt though. The Trojan Asteroids occupy Jupiter's stable Lagrangian points, 60° ahead of and behind it in its orbit. Several asteroids have large and often eccentric orbits taking them into the realm of the outer planets; there is evidence that this population of bodies merges into the Kuiper Belt of protoplanetary objects beyond Neptune. Of more local interest though, a significant number of asteroids have orbits partly or entirely within the notional Main Belt, in many cases approaching or even crossing the orbit of Earth. These latter objects are known as Near-Earth Objects, or NEOs.

NEOs are divided into 3 categories, named after the asteroids that typify each group. Amor asteroids are those whose orbits are larger than that of Earth, but which approach it closely without actually crossing it (formally, they are between 1.017 and 1.3 AU from the Sun at perihelion). Apollo asteroids have orbits larger than that of the Earth, but which actually cross it. Finally, Aten asteroids have Earth-crossing orbits that are smaller than the Earth's orbit. No asteroid has yet been discovered with an orbit entirely inside the Earth's, but this is probably due to the difficulty of observing such an object. Fig 5.2 shows the type of orbit of these objects.

Many of these objects - especially the Apollos and Atens - have the potential to have close approaches to the Earth. For such close encounters to take place, the asteroid's orbit must pass very close to or cross that of the Earth. Furthermore either it must be in an orbit with a very small inclination to the ecliptic, or one or both of its nodes must lie near the Earth's orbit. Note that the normal precession of orbits around the Sun due to perturbations by other planets ensures that the nodes of any Earth-crossing asteroid's orbit will periodically intersect the Earth's orbit.

It has been estimated that there are some 1500 NEOs larger than 1 km diameter [Gehrels94]. Some 200 such bodies are known to exist; the total population estimate is based on the average discovery rate. The number of asteroids of a given size is roughly proportional to r^-2, so there are believed to be in excess of 10⁵ NEOs of order 100 meters diameter and may millions of order 10 meter diameter.

As their name implies, NEOs can approach the Earth very closely. One of the earliest NEOs to be discovered, Adonis, passed within 2.4 million km of Earth in 1936, whilst the following year Hermes passed within 800,000 km. More recent monitoring with modern instrumentation has identified many more such close encounters, with the 1994 passage of 1994 XM₁ at just 100,000 km being a good example. Indeed, the recognition that there is a large population of such Earth-crossing objects has recently led to serious interest in the hazard presented by possible impacts by such bodies [Gehrels94]. Table 5.1 lists encounters with NEOs closer than 0.07 AU (10 million km) over the next 10 years.

Composition and Classification of Asteroids. Once spectroscopic examination of asteroids became possible it became evident that they exhibit individual differences. This was consistent with the observation that meteorites - which were suspected to be fragments of asteroids - fell into a number of broad compositional categories. By comparing the spectral response of asteroids with that of known meteorite samples it became possible to classify them into several groups [Gehrels89].

• C-Type, or Carbonaceous asteroids are carbon rich. They typically have a low albedo and very flat spectral response. C-type asteroids may be primitive undifferentiated remnants of the formation of the solar system. Some 75% of asteroids fall into this category.
• S-type or Stony or Silicaceous asteroids are redder than average, and appear to have a composition similar to stony-iron meteorites. They may be fragments of the crust or mantle of asteroids that became large enough to internally melt and differentiate. Most non-C-type asteroids are S-type.
• M-type or Metallic asteroids have high albedos and appear to be made of nickel-iron. They may be the cores of former large, differentiated asteroids. M-type asteroids are relatively rare.
• Other Types include E (Enstatite), R (Red) and U (Unclassifiable) asteroids, all of which are fairly rare.

These categories can be further divided into subclasses. For instance, research has identified some groups of asteroids very similar to one another. A number of asteroids are known that appear spectrally very similar to the large asteroid Vesta, and are believed to be fragments knocked off it during collisions.

In general though, there is still little hard data (other than similarities to types of meteorite) as to what these classes of asteroid are actually made of. Although the spectral signatures of some minerals can be confidently identified [Kowal96] there has traditionally been a dearth of 'ground truth' regarding asteroidal geology. Furthermore, remote spectroscopic studies give little information about the actual structure of asteroids. For such information, other forms of remote sensing are required.

Radar Observation of Asteroids. By using high-power radar transmitters and sensitive radio telescopes it is possible to obtain clear radar returns from nearby asteroids. The first such detection was of the near-Earth asteroid Icarus in 1968, whilst the first detection of a Main Belt object, Ceres, was in 1977. Observing asteroids by radar allows for several sorts of data to be obtained:

• Range and Range Rate measurement can be carried out very accurately at centimetric wavelengths. Ostro, one of the leading researchers in the field, quotes demonstrated state vectors of asteroids accurate to within 100 metres after only 3 days of tracking [Gehrels94]. Use of the Goldstone 70m Deep Space Network antenna allows asteroids of 1 km diameter to be tracked out to 0. AU, whilst the 300 metre Arecibo radio telescope can track such objects out to 0.4 AU, although over only a limited region of sky.
• Radar Reflectivity measurements can give an indication of the surface composition and texture of an asteroid. Surface composition can produce a direct effect on reflectivity; for instance, nickel-iron asteroids are naturally highly reflective at radar wavelengths. Surface texture indirectly affects reflectivity, as a rough surface will produce a different pattern and strength of reflection to a smooth one.
• Shape and Rotation Measurement can be carried out on asteroids by measuring the way in which the amplitude and Doppler shift of returns vary with time. For sufficiently good data (i.e. a high signal-to-noise ratio) it is even possible to use computer analysis of returns to synthesize images of asteroids. Fig 5.3a and b show radar images of asteroid 4179 Toutatis produced by Ostro and his colleagues. It is evident from these images that Toutatis - which is about 2 km by 4 km - is a highly irregular body. Radar studies of other NEOs have shown that such uneven shapes are common.

Structure of Asteroids. Little is known of the structure of asteroids, other than can be inferred from visual and radar images. There is some question as to what the irregular shapes of some bodies as seen in Fig 5.1 and 5.3a and b imply. On the one hand, they may show monolithic fragments of larger bodies that have been broken apart by collisions. Alternatively, they may be 'rubble piles' that have so little self-gravity that they have not aggregated into the near-spherical shape seen with larger objects such as planets.

Asteroids: Areas for Study

As is evident from the discussion above, there is much still to be learned about asteroids. In many cases more detailed knowledge can only be obtained by close investigation. Since the mid-1970s there have been several studies of missions to asteroids [Gehrels79, Gehrels89] which have led to 3 flybys to date: Gaspra and Ida by Galileo and Mathilde by the Near Earth Asteroid Rendezvous (NEAR) mission. NEAR flew past Mathilde in June 1997 en route to a rendezvous with the NEO Eros in 1999 [Farquhar95]. Such studies and missions have emphasized the following areas of study:

• Shape and Structure. Are asteroids monolithic fragments or aggregations of 'rubble'? High-resolution imaging of asteroids could help answer this question, especially if it included multispectral imaging to determine the mineralogical makeup of different regions of the asteroid. Such knowledge may be of practical as well as theoretical interest, should it ever become desirable to adjust the orbit of a NEO [Gehrels94]. If the target is suspected of being an extinct comet (as many NEOs are believed to be [Kowal96]) then imaging may show evidence of this, e.g. surface vents or icy deposits.
• Composition. Long-distance spectroscopy can give only an overall indication of surface composition. A high-resolution multispectral image would give much more information on composition. Techniques such as X-ray fluorescence and in-situ analysis would be even better as they would lead to direct information on surface composition.
• Asteroid Mass is difficult to determine on a flyby (although very easy on a rendezvous mission); the most promising method is to use Doppler tracking of the probe to measure any small velocity vector change. If imaging allows an accurate volume measurement then the target's density can be determined.
• Surface Structure can be determined by IR imaging, as the thermal properties of solid rock differ from those of dust or rock fragments [McEwen97].
• Asteroid Moons can be identified by imaging, although this involves imaging the area around the parent body. Ida's moon Dactyl was serendipitously discovered by Galileo during its 1993 flyby, but there is evidence that a significant proportion of asteroids may be binary. As seen from the radar images at Fig 5.3b, Toutatis may be a composite asteroid, whilst studies of double or multiple craters on Earth and other bodies suggest that many asteroids may have one or more moons.
• Magnetometry experiments can indicate the metallic content of an asteroid by the disturbance it creates in the local area of the solar magnetic field.

Comets

The initial proposal for this project cited comets as possible mission targets. In many ways comets are more promising targets for early detection than asteroids. They are intrinsically much brighter by virtue of the cloud of volatile material that surrounds them and trails away under solar radiation pressure to form the characteristic tail. As such they are usually detected some weeks or months before their closest encounter to Earth. They are highly dynamic bodies, whose interaction with the solar magnetic field and particle environment is a complex one. However, on further investigation a number of factors mitigated against choice of comets as a target for the more detailed design proposal.

Target Frequency. By contrast with the large population of NEOs (both known and awaiting discovery) relatively few comets pass close to the Earth. In the last few decades only 2 comets have passed within a few million km (IRAS-Araki-Alcock in 1983 and Hyakutake in 1996) so a short-range limited-duration probe such as this would, if optimised for a cometary encounter, probably have to wait a long time for a suitable launch opportunity.

Flyby Constraints. Comets tend to occupy highly-eccentric orbits [Kowal96] and indeed many are on orbits that are effectively parabolic. These long-period comets are on their first visit to the inner solar system, having been perturbed from the distant Öort cloud of cometary bodies by the gravitational influence of passing stars. Encounters with objects in such orbits will be at a very high relative velocity, possibly as high as 70 km/s. Although the larger size of comets means that the encounter duration will not be unduly shortened, it leads to a problem with the cloud of debris surrounding a comet's nucleus. As was found on the Giotto mission to comet Halley in 1986, hypervelocity impacts with even small fragments can cause severe damage. A cometary variant of the probe would thus need either extenal armour or would have to carry out a relatively distant flyby.

Science Payload. Although there is some overlap in the range of payload instruments required for comet and asteroid flybys, comet encounters have a number of scientific requirements not found in the latter mission. Particle and fields experiments are much more important, leading to the need to carry such instruments as magnetometers, particle density counters and dust impact measurement units. Many of these instruments are less or not appropriate for asteroid flybys, and so would be excess mass on most candidate missions; they also often require careful probe design in terms of electromagnetic compatibility to avoid the probe's systems or structure interfering with their operation.

In summary, a joint comet/asteroid probe design would probably not be an appropriate approach for a small, low cost mission, so separate designs would be required. Given the relative infrequency of close comet encounters, and the hazards they pose it was (with some regret) decided to concentrate on asteroids as targets for this mission. Nonetheless, the possibility of using a specially-designed low-cost probe for a longer duration mission to a more distant comet encounter should still be worthy of possible future study.