The difficulty in appreciating the hazard due to asteroid impacts is that the likelihood of a large asteroid strike in any given year, or even over a human lifetime, is very small. However, the consequence of such a strike is very large, amounting to an existential risk to humanity depending on the size of the near Earth object (NEO). The difference between the asteroid hazard, however, and most other deadly natural events, such as hurricanes and earthquakes, is that asteroid strikes can be both more devastating and also potentially predictable and preventable. Prediction and prevention, however, require knowledge of the asteroid population. Discovery and characterization observations—those that first identify potential NEOs and determine their nature—are made with telescopes either on the ground or in space with radar facilities providing additional positional and characterization measurements. This chapter discusses the data needed to detect, track, and characterize NEOs.
In comparison to the planets, asteroids are small and dim. They shine by reflected light from the Sun and by thermal radiation from their warm surfaces. Therefore, the ability to detect an asteroid depends on how much light it reflects, how warm it is, and the wavelength region in which the detection is being made. In visible wavelengths, the brightness depends on the asteroid’s size as well as on the intrinsic reflecting ability (albedo) of its surface (most asteroids are very dark, reflecting only 10 percent or less of the sunlight that falls on them), and the phase at which it is observed. Most small asteroids are sufficiently faint that they cannot be seen from Earth unless they make a close approach and are then picked up as moving points of light against the background sky by a sufficiently sensitive instrument. Once a newly detected asteroid moves away from Earth, it quickly becomes dimmer and eventually undetectable until it makes another close pass, perhaps as many as several years or even decades later.
In addition to initially detecting an NEO, there are other variables needed to determine the possibility of an asteroid impact and its potential severity. One of the most important among them is the NEO’s orbit, which determines if it is likely to intersect the orbit of Earth. This information, in turn, helps estimate not only whether there will be an impact but also the timeline of impact, the warning time, and the velocity at which the impact occurs. In addition to orbit, crucial for determination of the impact velocity, location, and time, there are other physical characteristics, such as mass and density, that determine the energy, and therefore the severity, of an NEO’s impact. The final important component of NEO detection strategy is the length of time it would take to achieve the George E. Brown, Jr. Near-Earth Object Survey Act’s goals.
While an NEO’s discovery observation can typically be used to derive an approximate orbit, follow-up observations are required to refine this solution and enable the prediction of the NEO’s position in the future. Such observations may be made both during the asteroid’s first appearance as well as during reobservation on a subsequent close-pass by Earth, which may occur many years, even decades, later. It also may involve “precovery” data, which are archival images in which an object was not found at the time the images were taken but is identified in them at a later time with the help of subsequent observations. NEO discovery surveys benefit from dedicated follow-up observations to improve the objects’ orbits and to provide information on their physical properties.
Astrometric observations measure the course of an asteroid across the sky in order to refine its orbit. Without astrometric observations, the positional uncertainties from insufficiently constrained orbits will quickly grow, impeding future targeted observations and the accurate assessment of impact probabilities. The quality of astrometric observations is dependent on the ability to pinpoint the position of an asteroid, the uncertainties of the asteroid images, the number of stars with well-known positions on each frame, and other factors. In practice, a given system will optimize some of these factors at the possible expense of others. However, observations over a long time span (or “arc”) usually allow for the rejection of poor-quality astrometric measurements and allow high-quality orbits to be determined. While current and future optical surveys (e.g., Catalina Sky Survey, Pan-STARRS, Large Synoptic Survey Telescope [LSST]) provide astrometric follow-up for some serendipitously observed asteroids soon after discovery, targeted observations of select objects (e.g., objects showing nonnegligible impact probabilities) will be necessary before their positional uncertainties grow too large or the objects become too faint (however, targeted follow-up is pointless for objects with very large uncertainties). Well-timed and short, single-apparition observations are sufficient to improve orbits in most cases. If initial orbits are recognized as being problematic, they will have to be followed up during the discovery apparition or risk being lost. In cases where they are lost, they will have to await rediscovery (or be located in a prediscovery image) at which point they will be linked with the discovery observations and a much-improved orbit determined. LSST will have to perform its own follow-up because the NEOs it discovers will be too faint and too many for the existing follow-up assets. The plan is that LSST will return to the same region of the sky every few nights to ensure that each discovered object receives sufficient follow-up.
Knowing the orbit allows an impact time to be determined if an impact is to occur. Knowing the orbit of an object allows that object’s velocity to be calculated for any point in the orbit. This is part of the motivation for asteroid searches in the first place—to determine whether an object has an orbit that intersects Earth’s orbit (and is thus a threat) or not, and to allow the impact velocity to be calculated if it does. For discussions of the NEO population as a whole, an average or typical velocity is usually adopted to represent the destructive power of a hypothetical impactor.
The severity of an asteroid impact is a function of its incoming energy, which is directly proportional to its mass. Small impactors are affected sufficiently by atmospheric drag that they fall at a relatively slow terminal velocity of a few hundred kilometers per hour—a potential problem for anything that happens to be at the spot where they land, but not an issue for broader areas. Larger, more massive objects are less affected and strike the ground at tens of thousands of kilometers per hour, fast enough to create a shock that more closely mimics an explosion than a rock dropped from a rooftop. The impact energy associated with an asteroid impact can be calculated from the asteroid’s mass and velocity. As a result, these are the two most important drivers for planetary defense studies.
Direct measurements of mass are more difficult to make than those of possible impact speed. In those cases where NEOs have satellites, the system mass can be determined remotely by determining the orbit of the NEO satellite. In some cases, mass can be inferred by making very precise positional measurements and measuring discrepancies between those positions and predicted positions due to the effects of nongravitational, mass-dependent forces (like the “Yarkovsky force”). In the general case, however, NEO masses can only be directly measured during spacecraft visits, which are rare.
Due to the need for mass estimates in general cases, indirect methods for estimating masses have been developed, using measurements or estimates of the NEO volumes and densities. Although there are additional uncertainties for NEOs with highly irregular shapes, in the general case, NEOs are usually treated as spheres. Because the volume of a sphere can be easily calculated from its radius or diameter, these are used as proxies for volume. For historical reasons, “size” usually refers to the diameter of an NEO rather than its radius. The language of the George E. Brown Act, which established NASA’s mandate to find asteroids of a certain size, is motivated by the use of asteroid diameters as a proxy of their destructive power.
While accurate diameter measurements—serving as a proxy for mass in combination with inferred bulk densities—have the highest importance for planetary defense purposes, additional physical properties can be constrained with a range of observational methods, improving the former estimate. Depending on the method, useful information may be obtained with relatively few observations or could require repeated observations over the course of many years. Focusing intensive efforts on the truly threatening objects, out of the roughly 30,000 NEOs, will benefit from an estimate of the diameter and mass in the initial detection, particularly if the facility needed for this estimate is of limited lifetime.
Finding: In addition to detecting NEOs and determining their orbits, it is necessary to estimate their masses to quantify their destructive potential. An NEO’s diameter is the most readily available indicator of its mass.
NEOs are too small to appear as more than point sources in telescopic data. In favorable cases, radar can be used to obtain a direct measurement of asteroid size. In the great majority of cases, however, the size of an asteroid must be calculated from its brightness and its distance, with the latter determined from its orbit. The population of NEOs has a wide variation in how reflective their surfaces are, which leads to uncertainty in this size-measurement technique. However, the uncertainties are much smaller when the measurements are done using emitted infrared light rather than reflected visible light (the reason for this improved accuracy is that albedo has only a weak effect on the emitted thermal flux). This difference underpins some of the reasoning for preferring infrared systems for asteroid surveying. (See Chapter 5 for further information.)
Finding: The accuracies of asteroid diameters derived from thermal-infrared measurements and simple modeling usually far exceed those based on the measured visible brightness alone.
The cause of the largest uncertainty in impactor destructive power in the general case, which cannot be easily reduced from remote data, is the uncertainty in object density. Considering the total range of possible densities (~1 to 8 g/cm3—see below) this factor of ~8 in possible densities translates to a factor of ~8 uncertainty in mass for NEOs with no data from which to constrain their densities. If some compositional information is present or rare compositions are excluded, there is still roughly a factor of 2 uncertainty in mass, which corresponds to a factor of 2 in impact energy. As a result, this factor of 2 in impact energy is used as the acceptable level of uncertainty in other measurements. The velocity and orbit uncertainties are easily dismissed as too small to contribute significantly to the overall uncertainties.
In order to have the size uncertainty be less important than the density uncertainty for likely impactors in terms of determining the mass of an object in the general case, the volume uncertainty must be less than a factor of 2. Given the well-known relationship between size and volume, given the established shape, that means the size uncertainty must be less than the cube root of 2, or roughly 1.26. This means that a ~25 percent uncertainty in size must be achieved by the NEO characterization systems considered here.
Finding: A search program that can measure NEO sizes to 25 percent uncertainty or better with the same observations used to discover them and obtain their orbits is preferable to separate search and characterization programs, unless separate systems can complete the survey more quickly or cost effectively than a single program that does both.
Meteorite data can be used to make density estimates. Meteorite densities are dependent on composition, with a porous carbonaceous chondrite having a density as low as 1 g/cm3 and a solid rock of the most common composition (“ordinary chondrite”) having a density of 3.0 to 3.5 g/cm3. Other important compositions have densities from 2.0 to 7.5 g/cm3 when they are solid chunks, although less than 4 percent of meteorites are high-density irons (density 7.5 g/cm3). Kilometer-size asteroids are expected to rarely if ever be solid chunks, however, and they can have 30-50 percent void space in their volume, bringing down the overall density. Density measurements of asteroids have a weighted average value of 2.62±1.23 g/cm3, also suggesting that most asteroid densities fall in a relatively small range.1 However, neither measure may be completely apt; meteorite densities are biased toward higher-density objects that can better survive atmospheric entry, and the available asteroid densities are generally for objects much larger than the NEOs of interest here. Nevertheless, both suggest that extreme values for asteroid density are rare.
If no other information is available, the factor of ~8 in density between solid iron and porous carbonaceous chondrite can be reduced only by making probabilistic arguments, such as those in the preceding paragraph. If compositional information is available from reflectance spectroscopy, the probable density range can be narrowed significantly. If albedo information is available, the likely density range can be similarly narrowed, although with somewhat less certainty.
There are several types of characterization observations that can be made. Photometric observations of asteroids measure the amount of solar light that is reflected by their surfaces. Quantifying their brightness over time and along their orbits over many apparitions constrains their surface albedos when combined with diameter measurements from thermal infrared observations, as well as their shapes and rotational properties. The derived albedo can be used to infer the targets’ composition and hence their bulk densities. In order to measure accurate asteroid albedos, repeated measurements over many apparitions along their orbits are required. The measurement of asteroid shapes and rotational properties through brightness variations due to their irregular shapes and rotations also requires a large amount of highly accurate photometric observations over many apparitions. Both properties, combined with accurate thermal-infrared observations, can significantly improve the physical characterization of asteroids and are able to improve bulk density estimates. Accurate photometry can be obtained for a small number of large and bright asteroids with easily accessible small telescopes (~1 meter aperture), but fainter asteroids require larger facilities. Often, the absolute magnitude of an NEO (H) is determined using data optimized for finding its position, and can have uncertainties of 30 percent or more. In addition, the brightness of an object changes as viewing angle changes, captured in a parameter called G. With concerted effort, improved values of H and G can be determined for NEOs, which in turn improve estimates of size.
Spectroscopic observations in visible and near-infrared light measure the amount of solar light that is reflected by the surface of the asteroid as a function of wavelength. The observed reflectance spectrum provides a spectral classification of an asteroid. Spectral classes have been linked to meteorite analogs, enabling fairly accurate inference of the bulk density and albedo. Quantitative spectral analysis can also retrieve information on surface particle size and mineralogy. For the majority of asteroids, such observations require large telescopes (>4 meter aperture) and long observations, except when they are very close to Earth. Single apparition observations are sufficient to put constraints on spectral classifications.
Spectrophotometry-photometric observations using different broadband filters provide a simple means to roughly constrain an asteroid’s spectral classification.2 While enabling a less-detailed taxonomic classification than spectroscopy, spectrophotometry enables the characterization of fainter, and hence smaller, asteroids with generally smaller telescopes. Single apparition observations are sufficient to put constraints on spectral classifications.
Polarization measurements of the Sun’s visible and near-infrared light reflected by the surface of asteroids directly provide the albedo and thus, via the albedo, a rough constraint on the density. They are also indicative of
1 B. Carry, 2012, Density of asteroids, Planetary and Space Science 73:98-118.
2 Z. Ivezić and 31 coauthors, 2001, Solar system objects observed in the Sloan Digital Sky Survey commissioning data, Astronomical Journal 122:2749-2784, doi:10.1086/323452; M. Jurić and 15 coauthors, 2002, Comparison of positions and magnitudes of asteroids observed in the Sloan Digital Sky Survey with those predicted for known asteroids, Astronomical Journal 124:1776-1787, doi:10.1086/341950; B. Zellner, D.J. Tholen, and E.F. Tedesco, 1985, The Eight-Color Asteroid Survey: Results for 589 minor planets, Icarus 61:355-416, doi:10.1016/0019-1035(85)90133-2.
spectral classification and provide additional information on surface properties like particle sizes and mineralogy. Polarimetric observations require well-timed multi-apparition observations and medium-size telescopes (>2 meter aperture) in order to provide useful information. Hence, only a small sample of asteroids can be investigated using this method.
Radar observations provide an opportunity to obtain highly accurate astrometric data and shape and rotational information for a small sample of asteroids that come close enough to Earth to be observed with this technique. While this limitation prevents its broader use for asteroid characterization in a large-scale survey, it can be used to understand the range of properties in the overall asteroid population by characterizing individual objects. Radar has been used to make measurements of a representative sample of NEOs down to an H magnitude of 28. High-precision, single-apparition radar data are extremely useful, and multi-apparition data, where available, are able to improve bulk density estimates through precise astrometric measurements that enable orbit migration due to nongravitational effects to be determined. Observations of binary NEOs directly yield bulk densities.3
Finding: Characterization—that is, determining the physical properties of NEOs such as their diameters and densities—is critical for a full understanding of the impact hazard. Characterization observations can include radar and visible-infrared photometry and spectroscopy. Most often, not all are available.
The timescale over which measurements will be made is considered in the context of the George E. Brown Act that includes a timeframe for discovery of NEOs >140 meters. It is apparent that the timeframe specified in the act cannot be completed by 2020, yet an infrared platform operating in conjunction with ground-based visible light telescopes both current and under construction will allow the completeness goal to be met in a timeframe constrained by physics, and development time needed to employ the technology that will achieve the completeness goal stated in the congressional mandate.
Another factor, although one that is not straightforward to quantify, is the “complementarity” of a new system to existing systems. A survey system with capabilities unmatched by existing systems, whether in observable areas of the sky, detection limits, with observing biases that are minimal or at least different from existing systems, and so on, is preferred to one that simply supersedes an existing system but still leaves portions of the survey space uncovered.
The discussion is focused on L1 as it is currently considered a very feasible location for a survey telescope. While it is true that locations at different heliocentric longitudes offer other advantages, it might be more challenging to maintain operations due to limited bandwidth and other aspects. L1 offers a good compromise. The overlap in search volume with ground-based surveys should be very limited as is shown in Figure 4.1: While ground-based surveys tend to focus on opposition observations, thermal-infrared space-based telescopes typically aim at quadrature to maintain cooling of the spacecraft.
Last, the cost of a system is another factor that can be used in assessments, but only in conjunction with other metrics. For instance, in general, a low-cost system that would take several decades to complete the survey would not be preferable to a hypothetical higher-cost system that could complete it faster—assuming reasonable costs; the committee is not claiming that cost is unimportant, only that taking many decades to conduct a survey in order to save costs could have deleterious effects, including cancelation and the inability to recruit talent to undertake a project with no clear end point. On the other hand, small investments in existing or planned astrophysical assets could allow them to contribute meaningfully, if incrementally, to completing the survey even if they do not play a large role. The scope of this study does not allow for new costing models to be developed. The committee therefore has had to rely on what public information does exist as well as indirect information where available.
3 D.J. Scheeres et al., 2019, The dynamic geophysical environment of (101955) Bennu based on OSIRIS-REx measurements, Nature Astronomy 3:352-361.
Because our knowledge of asteroids is still evolving, and because the properties of asteroids are essential for gauging the destructive potential of any impact, as well as evaluating the effectiveness of any proposed deflection method, it is necessary to incorporate scientific input into the development of survey techniques and technologies.
Finding: While planetary defense missions are not science driven, significant scientific input is essential to optimally design the planetary defense task.
The most successful campaign to protect Earth from the consequences of a large impact will involve strong contributions both from scientific study of the solar system and engineering approaches to asteroid discovery and deflection. Pure science will inevitably profit from a robust asteroid detection and defense program, while the proper design of deflection strategies must be informed by the latest scientific discoveries about the nature of asteroids and comets. A good example of how this synergy between pure and applied science might work is the Double Asteroid Redirect Test mission, which is a NASA-funded mission to change the orbit of an asteroidal satellite in space, demonstrating a possible mitigation technique.