1. Introduction

Asteroid and comet impacts played a crucial role in the solar system formation and presumably in the evolution of life on Earth, the most recent such event – the Cretaceous/Tertiary (K/T) event, 65 million years (Myr) ago – being considered largely responsible for the extinction of dinosaurs (Alvarez et al. 1980). However, despite the current achievements of human civilization, we are not yet able to prevent a similar fate. If a similar size (approximately 10 km) asteroid was found to collide with Earth within a few years, we would be very hard-pressed to prevent the impact. The rarity of such "extinction-class" events - in average one every 100 Myr (Morbidelli et al. 2002) – is generally seen as a comforting factor, but this is uncertain and could be illusory (Asher et al. 2005). A good approximation (Ward and Asphaug 2000) to the number of impacts by objects of radius r > R_1/ is a power law R₁^-7.3 Thus, an impact by an object of 1 km radius would therefore occur in average every 500,000 years, with an equivalent energy of approximately 500,000 megatons of TNT. This is still a massively devastating impact, and it is difficult to evaluate precisely when the last event of that class actually occurred.

The threat is very real and extends to objects well below the size of the K/T event; therefore, it would be good policy to develop the means to protect the planet, the eco-system, and our civilization from annihilation. This topic of "Planetary Defense" is increasingly the subject of serious studies, and although a coordinated, long-term effort is currently lacking, there is constant progress in that direction (Hiss and Garretson 2008; Wie 2008a; Ulrich 2009). This is however not an easy endeavor, both for socio-political and technical reasons. It is difficult to justify the commitment of public funds to address a problem occurring on the scale of a Myr, a number far beyond human perception. While the much higher frequency of medium-class (30-100m) impacts makes the threat more prevalent and understandable to the public, limited resources could equally prevent the funding of a remediation program – e.g. hurricane or earthquake preparation may take priority. Technical challenges can also be formidable and require long-term development of costly technology. However, we believe there is a path out of this dilemma.

2. The Challenges

2A. Detection & Characterization

The first step in improving our safety from the impact of Near-Earth Objects (NEO) and other bodies consists of actively detecting, tracking and cataloging them. This aspect of the remediation strategy is probably the most urgent and (even when considering dedicated space telescopes) the least costly aspect of any planetary defense program. Identifying threats as early as possible greatly facilitates a response mission for threat mitigation. Any mitigation option will require several years to a couple of decades lead time, a concept which humanity may not come to grips with until a real crisis is at hand. There is here a potential opportunity for international cooperation and US leadership to establish an all-encompassing detection effort.

For example: procedures for the US Space Command to declassify observations of upper atmosphere explosions of small (< 50 m) asteroids to help model frequency and energy distributions; and more scenario and contingency planning among US government agencies as well as international and non-government organizations. Currently, no agency is in charge of Planetary Defense, and there is no concept of operations to counter or respond to an impact.

There are several technical approaches to a systematic detection campaign which are discussed at length elsewhere (e.g. Ivezić et al. 2006) Overall, the detection infrastructure must be expanded. Ground-based telescopes are needed world-wide, especially in the Southern hemisphere where efforts have been lagging. Space-based detection is also needed, and may be performed at reasonable cost, such that universities can develop and experiment with new detection equipment.

A current example is the Near-Earth Object Surveillance Satellite program from the University of Calgary. This demonstration effort has multiple goals, including detecting NEOs and monitoring high-orbit satellites (Hildebrand 2008).

Another important issue is characterization, the gathering of detailed information from potential colliders. This would facilitate the subsequent mission of deflecting the asteroid, by providing detailed information, e.g. composition and structure, which help determine the best approach to deflection and/or the timing and position of the devices to be used. This is particularly important for methods such as nuclear detonations, kinetic impacts, or laser/solar ablation. However, any precursor characterization mission significantly adds to the lead time required for a successful deflection and increases the risks. Therefore, a long-term NEO characterization campaign is highly desirable, in order to gather statistically significant data for accurate and predictive modeling of various deflection scenarios. This will require low-cost robotic missions – arguably some manned missions – to asteroids of different types, involving intrusive (drilling, sampling), or remote (co-orbiting) sensing. A sufficiently large number of missions are needed to obtain data from a statistically significant sample of NEOs. Given the distances and associated communication delays, extensive use of Artificial Intelligence (AI) is necessary. The missions must also have sufficient power and life-time to perform the necessary work; compact, high-performance (notably regarding efficiency and specific impulse) advanced propulsion systems are also a critical requirement.

2B. The Challenges of Deflection

The key task in threat mitigation is to provide enough impulse to the object to make its trajectory miss Earth by a reasonable margin of error, an especially difficult task for large objects. Generally speaking, the impulse required is not a monotonic function of time, but the overall trend remains that the force required becomes excessively large when the time before impact decreases. Even when there is sufficient warning (i.e. decades), it is still beneficial to apply as much force as possible; not only this gives a bigger deflection margin, but it allows us more time to make corrections if necessary. The time required to prepare, launch and execute a deflection mission can take several years in the best case scenario (Barbee 2005), assuming we even had the launch and spacecraft capability readily available. A prior characterization of the target would add several years to the overall mission. A rapid "Access-to-Space" would greatly improve the odds of a successful deflection, but would require new, highly robust launchers. Such an airplane-like operating tempo has long been an objective for both civilian and commercial space throughout the world, and is the subject of active research.

2C. The Nuclear Stand-Off Option

If time is critical, one must use "hard-push" deflection scenarios, i.e. application of large forces during short durations, and nuclear detonations are the most effective means of delivering the energy required for this impulsive maneuver. In the nuclear stand-off approach the weapon is detonated at some distance from the NEO, and the energy (mostly X-rays and neutrons) is used to ablate material on the surface of the asteroid, with the subsequent recoil providing the impulse for deflection (Dearborn 2004). The approach a-priori requires little characterization of the NEO, and no rendezvous procedure with the target – two significant time-saving features. Because the energy is delivered mostly as radiation, the impulse can be distributed over a large area (in contrast to high velocity impactors) and to objects of any shape (including rubble piles and sand piles). Furthermore, if detonated at sufficient distance, impulse could be provided to NEO satellites, i.e. composite objects. However this approach is not the most efficient: a lot of energy can be re-radiated and the impulse-to-energy ratio can be small – scaling approximately as: heating a small mass to very high temperatures is not an efficient way to provide the necessary impulse to the asteroid1/T^2/1: heating a small mass to very high temperatures is not an efficient way to provide the necessary impulse to the asteroid.. The same observations apply to the case of laser ablation.

The fact that a nuclear stand-off approach is still an attractive option mostly results from the enormous amount of energy available in thermo-nuclear devices. Since we are not concerned with having a limited amount of mass available, the best use of the available energy would be to heat a greater amount of material, but to temperatures just high enough to vaporize it. However, the energy threshold required depends on material properties (composition, structure, porosity, etc.), and this may again require a prior characterization mission, unless there is enough information from the systematic campaign of NEO characterization.

It has been suggested (e.g. Gennery 2004; Holsapple 2004) that neutrons could be more effective and their yields could be optimized with the appropriate warhead design, but this again requires a good characterization of the NEO properties, notably porosity and composition. If we consider, for example, an object with a diameter of 1 km, using an average range of 150 kg/m² for neutrons, average geometric irradiation factors and a stand-off distance of half the object's diameter, one needs a neutron yield of about 5-10⁷ J in order to achieve the material-independent value of specific energy deposition of 100 MJ/kg. Since in an optimized device the neutron yield is approximately 10% of the total energy, this is equivalent to a weapon yield of 1,200 megatons (Mt-TNT), or 130 times the yield of the most powerful weapon developed by the US (or equivalently, hundreds of existing stockpiled devices). The X-ray yield being typically larger by at least one order of magnitude, it is still debatable whether neutrons provide a better coupling efficiency.

2D. Explosive Fragmentation

Buried detonation would be a more effective approach, since the entire energy of the device can be absorbed by the target and re-radiation of high-temperature plasma is no longer a net-loss, since it is absorbed by the surrounding material. The effect most easily obtained, i.e. requiring the least energy, would be fragmentation of the object. Since at the threshold all the debris would remain on essentially the same trajectory, for fragmentation to be effective the debris must have a significant relative velocity as a result of the detonation, so that by the time of impact they are widely scattered. The minimum energy for fragmentation of a 1 km-diameter asteroid was approximated by Ahrens and Harris (1994) as 1 Mt-TNT; this figure is easily increased by one order of magnitude to provide the required scatter velocity to the debris (depending on the time-to-impact). This still makes a deeply buried nuclear detonation about two orders of magnitude more efficient than stand-off detonation and brings the required yield down to values where stockpiled weapons are useful. However, a buried device will require: (a) a rendezvous mission instead of simply intercepting the NEO; (b) drilling into the core to significant (≈100 m) depth, an especially difficult prospect for hard (nickel-iron) objects, or (c) a deep impactor with delayed fuse, possible only for very soft targets. Therefore, a buried nuclear detonation would not require development of new warheads, but the mission time would be significantly longer and would still be difficult for some targets.

It has been repeatedly claimed that nuclear detonations and kinetic impacts are not desirable, since fragmentation of the target would lead to an increased level of damage from multiple impacts. This is, however, highly debatable. Complete break-up of the target would lead to a spreading of the NEO material (eventually forming a "ring"), only a fraction of which would intersect with Earth (Sanchez, Vasile and Radice 2008). The effects of an impact also do not scale linearly; a large asteroid impact creates a global catastrophe for any impact location, while smaller objects have a greater chance of hitting oceans and regions of low-population densities. Finally, smaller debris would burn-up at high altitude in the atmosphere. It has been suggested that the flash heating would be equally detrimental, and this is mostly dependent on the volumetric density of the debris at impact, which becomes very low after sufficient time has elapsed for the spreading of the material. By comparing the energy flux of the debris with the solar flux, one could estimate the time required after break-up (presumably 1-2 yr) for safe interaction with the debris cloud; detailed calculations are still needed. Another issue is that by fragmentation, an impact that would otherwise devastate a large country could be transformed into a series of smaller events spread around the globe, but still capable of inflicting thousands of casualties. This scenario has political consequences, since the damage would be shifted from one country to another; in a sense, the deflection mission could be construed as an act of war.

Nevertheless, the nuclear option can easily handle small objects (30-50m) and is our only chance to deal with extreme (>1 km) threats. In between, the situation requires more careful study.

2E. Other Options

Deflection by kinetic impact is also possible for relatively small asteroids (Holsapple 2004; Sanchez et al. 2009). Kinetic energy being not conserved in inelastic collisions, conservation of momentum must be used to evaluate the impulse transferred to the asteroid ΔI = MΔV = m_pv_p. Due to constraints on initial launch mass M_o and time to intercept Δt the spacecraft must satisfy some minimal performance requirements. For any vehicle speed (e.g. v_p =10 km/s) and time to intercept (e.g. Δt ≈ 1 yr), the mass in orbit and minimum power requirements can be estimated from: