COMET/ASTEROID PROTECTION SYSTEM (CAPS): A SPACE-BASED SYSTEM CONCEPT FOR REVOLUTIONIZING EARTH PROTECTION AND UTILIZATION OF NEAR-EARTH OBJECTS Daniel D. Mazanek, Carlos M. Roithmayr, and Jeffrey Antol (d.d.maza_ek_larc.nasa._s_v, c_m r(fiihmavr:a,_larc.nasa._, _larc.nasa._.ov) Spacecraft and Sensors Branch, ASCAC NASA Langley Research Center Hampton, Virginia USA Linda Kay-Bunnell and Martin R. Werner Joint Institute for Advancement of Flight Sciences (JIAFS) The George Washington University Hampton, Virginia USA Sang-Young Park (_Iarc.nasa._ Swales Aerospace, Inc. Hampton, Virginia USA Renjith R. Kumar (_iv:.ama-inc.com) Analytical Mechanics Associates, Inc. Hampton, Virginia USA ABSTRACT There exists an infrequent, but significant hazard to life and property due to impacting asteroids and comets. There is currently no specific search for long-period comets, smaller near-Earth asteroids, or smaller short- period comets. These objects represent a threat with potentially little or no warning time using conventional ground-based telescopes. These planetary bodies also represent a significant resource for commercial exploitation, long-term sustained space exploration, and scientific research. The Comet/Asteroid Protection System (CAPS) would expand the current detection effort to include long-period comets, as well as small asteroids and short-period comets capable of regional destruction. A space-based detection system, despite being more costly and complex than Earth-based initiatives, is the most promising way of expanding the range of detectable objects, and surveying the entire celestial sky on a regular basis. CAPS is a future space- based system concept that provides permanent, continuous asteroid and comet monitoring, and rapid, controlled modification of the orbital trajectories of selected bodies. CAPS would provide an orbit modification system capable of diverting kilometer class objects, and modifying the orbits of smaller asteroids for impact defense and resource utilization. This paper provides a summary of CAPS and discusses several key areas and technologies that are being investigated. Copyright© 2002bytheAmericanInstituteofAeronauticsandAstronautics,Inc. Nocopyrightisassertedinthe United StatesunderTitle17,USCode. The USGovernmenthasaroyalty-freelicenseto exerciseallrightsunderthecopyright claimedhereinforGovernmentalPurposes.All otherrightsarereservedbythecopyrightowner. 53rdInternational Astronautical Congress October 10-19,2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA INTRODUCTION Although the primary motivation for CAPS is to provide protection against impacting comets and An enormous number of asteroids and comets orbit asteroids, it is anticipated that the system and the Sun, ranging in size from pebbles to mountains. technologies developed would have many additional Fortunately, only a tiny number of these objects benefits extending to governments (U.S. and cross the Earth's orbit, and our atmosphere protects international), the commercial sector, the scientific us from small and structurally weak objects. Impacts community, and academia. The CAPS detection axe extremely infrequent events relative to a human system would provide an astronomical asset that lifetime, but have the potential for massive loss of could observe extremely faint or small targets (both life and property. Impacts have occurred in the past planetary bodies and extra-solar objects), providing and will occur in the future. The energy released an unprecedented level of scientific observations from an impactor capable of causing surface damage while surveying the entire celestial sky on a regular ranges from -10 megatons (MT) of TNT to billions basis. The CAPS orbit modification system could of megatons (1 MT = 4.185 x 1015Joules). A 10 MT enable exploitation of the vast economic resources impact can result from an object approximately 50 m available from NEOs, and promote synergistic in diameter, and is roughly equal to 700 Hiroshima technologies for other future space missions. size explosions. This class of impact is estimated to Technologies that will permit the future exploration occur every several hundred years (or possibly less) and colonization of the solar system (e.g., high and can cause regional destruction. An impact with power and thermal management systems, high thrust a 1km diameter object, capable of releasing roughly and specific impulse propulsion, and power 100,000 MT and resulting in a global catastrophe, beaming) are applicable also to the mitigation of can be expected to occur every several hundred Earth impacting comets and asteroids. Additionally, thousand years to a million years. An impact from a there is tremendous benefit in "practicing" how to 10 km object, like the one believed to have caused move these objects from a threat mitigation the great dinosaur extinction 65 million years ago, standpoint; developing the capability to alter the can be expected on an interval of 10million years or orbits of comets and asteroids routinely for non- greater. defensive purposes could greatly increase the probability that we can successfully divert a future Earth approaching asteroids and comets are impactor. The vision for CAPS is primarily to collectively termed NEOs (near-Earth objects). The provide planetary defense, and to provide productive goal of current search efforts is to catalog and science, resource utilization and technology characterize by 2008 the orbits of 90% of the near- development when the system is not needed for the Earth asteroids (NEAs) larger than 1km in diameter, infrequent diversion of impacting comets and currently estimated to number between 900 and asteroids. 1300. Devastating impacts can also occur from smaller NEAs, short-period comets (SPCs) in This paper provides a summary of CAPS and asteroid-like orbits, and long-period comets (LPCs) discusses several of the key areas that axe being which do not regularly enter near-Earth space since investigated: precision orbit determination, optical their orbital periods range from 200 years to million interferometry, and rapid rendezvous with laser of years. ablation for orbit modification. The paper concludes with a discussion of some key technologies that The Comet/Asteroid Protection System (CAPS) is a would permit a permanent, continuous detection future space-based system concept (25 or more years system, and an orbit modification system capable of from now) designed to detect and mitigate the entire altering the orbits of asteroids and comets in a range of threatening comets and asteroids. The controlled, rapid manner. initial focus is to determine the feasibility of protecting against 1 km class long-period comets, BACKGROUND including inactive nuclei. The system is designed also to protect against smaller LPCs, as well as While many aspects of the impact hazard can be NEAs and SPCs capable of regional destruction. addressed using terrestrial-based telescopes, the 53rdInternational Astronautical Congress October 10-19, 2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA ability to discover and track faint and/or small It is likely that the next object to impact the Earth comets and asteroids is tremendously enhanced, if will be a small near-Earth asteroid or comet. The not enabled, from space. Just as the Hubble Space most significant danger from smaller NEOs (several Telescope has expanded our ability to see the hundred meters in diameter) may result from ocean universe without the limitations imposed by the impacts, which can generate tsunamis capable of Earth's atmosphere, a space-based NEO detection massive destruction on distant shorelines. A system would allow us to expand the range of globally devastating impact with a 1 km class LPC comets and asteroids that we can observe, and will not be known decades, or even years, in advance provide coordinated follow-up observations. with our current detection efforts. Searching for, and protecting ourselves against these types of impactors A space-based detection system is capable of making is a worthwhile endeavor. A space-based detection observations on a continuous basis, without the system, despite being more costly and complex than various constraints (daylight, weather, etc.) imposed Earth-based initiatives, is the most promising way of on Earth-based systems, and NEO searches need not expanding the range of objects that could be be focused on the solar opposition point. If detection detected, and surveying the entire celestial sky on a systems can be designed to observe faint NEOs that regular basis. Current ground-based efforts should appear to be near the Sun, which is impossible from be expanded, and a coordinated space-based system the ground because the atmosphere scatters sunlight should be defined and implemented. CAPS is an during the daytime, it would be possible to see attempt to begin the definition of that future space- objects close to the Sun, and on the solar far side based system, and identify the technology where solar illumination conditions are favorable. development areas that are needed to enable its Additionally, it is critical to ascertain, to the greatest implementation. extent possible, the composition and physical characteristics of these objects. A space-based CONCEPT OVERVIEW approach can solve this aspect of the problem, both through remote observations and rendezvous The combination of the words "comet" and missions with the NEO. Finally, any attempt to "asteroid" in the CAPS acronym is intended to deflect an impacting NEO with any reasonable lead- convey the idea of utilizing a combined approach for time is likely only to be accomplished using a space- protection against both types of these cosmic based deflection system. projectiles. Conventional ground-based telescopes may provide little or no advanced warning of a It is recognized, and appreciated, that the currently collision with a small NEA or LPC, and developing funded terrestrial-based detection efforts axe a vital and maintaining separate space-based systems may and logical first step. Focusing on the detection of be impractical. Precise orbit determination of most large asteroids capable of global destruction is the NEAs and SPCs can be expected to be obtained best expenditure of limited resources. However, several orbital periods prior to a collision, provided various aspects of the impact threat axe largely that we actually have the ability to observe them. unaddressed by these efforts. There is currently no This would not be the case for impacting LPCs, specific search for LPCs, small NEAs, or small whose orbits need to be characterized very SPCs. Additionally, coordinated follow-up accurately over a small observation arc on their first observations are critical to limit the likelihood of observed perihelion passage through the solar losing a newly discovered NEO, and to determine system. If the situation does occur in which a small the object's orbit. One short coming of current NEA is first detected on its final approach, ground-based efforts is the difficulty in providing mitigation of the impact may prove problematic. In these follow-up measurements, which axe in part this case CAPS would at least provide an accurate provided by amateur astronomers. Looking for assessment of where the object would impact, and much smaller and fainter targets is likely to exceed enough warning time to allow some appropriate civil the capabilities of many asteroid and comet defense effort to be carried out successfully. "hunters" on Earth. 53rdInternational Astronautical Congress October 10-19,2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA The timely detection of LPCs, even those of resource utilization. We envision a future for significant size, presents many intractable problems. mankind in which asteroids and cometary bodies are LPCs can be extremely faint (albedos of -0.02) until routinely moved to processing facilities, with a the sublimation of their volatile frozen gases begins. permanent infrastructure that is capable and prepared Moreover, comets can remain in a dormant state to divert those objects that axea hazard. during their perihelion passage, or they can exhaust their volatiles and become extinct nuclei. Observing Detection LPCs at significant distances from the Sun is a formidable task. The ability to predict their orbits It is worthwhile to describe what is meant by accurately, and hence determine whether or not they "detection" in relation to the CAPS detection system. represent a threat, is dependent upon the number, Detection includes initial discovery, follow-up resolution, and spacing of observations of these observations, precision orbit determination, and objects. Finally, the comet's trajectory can be physical characterization. Although all aspects of significantly altered by non-gravitational forces if it the detection problem are critical, for objects with a becomes active, affecting our ability to predict its very limited observational period the accurate path and properly alter it. The ability to observe assessment of their trajectory is vital. faint LPCs and rapidly determine their orbits is consistent with protection against small, previously The initial benchmark for the CAPS detection undiscovered NEAs. A system capable of protecting system is to be able to identify an impactor with a against LPCs, placed properly in heliocentric space, diameter of 1km or greater, at a distance of at least 5 should also be capable of protecting against small Astronomical Units (AU) from Earth (1 AU _ 150 NEAs and small SPCs. million km), and objects as small as 50 m in diameter at a distance of 0.2 AU from Earth. In The baseline detection concept advocates the use of general, these distance limits would provide warning high-resolution telescopes with advanced detector times of approximately one year for a 1km LPC, and arrays, coordinated telescope control for NEO a few weeks to approximately a month for a small surveying and tracking, rapid spectral imaging for NEA that has not been previously cataloged. A NEO identification and characterization, and system possessing the sensitivity to observe 1 km interferometric techniques to obtain precision orbit objects at 5 AU would be capable of detecting many determination when required. Detection telescopes 50 m class asteroids significantly farther away than would be orbiting and/or lunar surface based, 0.2 AU, so the warning times for uncataloged NEAs providing surveys of nearly the entire celestial sky could be significantly longer. Ultimately, the ability approximately every 30 days. Orbiting telescopes to identify a LPC on an impact course at a distance could be placed in heliocentric orbits, including of 5 AU from the Earth may not provide sufficient Earth-Sun libration points, or around a planetary warning time, and observing further out may be body or moon. The CAPS detection system would valuable. Conversely, due to the unpredictable provide a high probability that impacting NEOs are nature of comets, both with respect to their orbits detected, and their orbits accurately characterized and structural integrity, it may not be prudent to take with significant warning time, even upon their first any defensive action until the object is much closer. observed near-Earth approach. The threat of impact may change significantly if the comet becomes active, or if it fragments into a The primary orbit modification approach uses a number of sizable objects. The extremely short spacecraft that combines a multi-megawatt electrical warning times for LPCs, the large changes in orbital power system, a high thrust and specific impulse velocity required to avert an impact, and the orbital propulsion system for rapid rendezvous, and a pulsed and compositional uncertainties make this aspect of laser ablation payload for changing the target's orbit. the impact hazard particularly difficult to solve. This combination of technologies may offer a future orbit modification system that could deflect The envisioned CAPS detection system would impactors of various compositions, and provide an feature large aperture (> 3 meters), high-resolution effective method for altering the orbits of NEOs for telescopes capable of imaging in the ultraviolet, 53rdInternational Astronautical Congress October 10-19, 2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA optical, and infra_red wavelengths. Coordinated between the target and the detection system. The telescope control for NEO surveying and tracking tracking telescopes could be used as receivers for the would be incorporated to maximize follow-up laser ranging system, or the return signal of faint observations, and baffling and/or shading would be NEOs could be enhanced through active illumination employed to permit observations close to the Sun. to aid in interferometry measurements. Figure 1 depicts a luna_r-based option with a detection node consisting of a wide field-of-view Deflection/Orbit Modification (FOV) survey telescope located in the center, and three narrow FOV tracking telescopes (telescope Figure 2 depicts a promising, advanced method of enclosures and/or baffling are not shown). Two orbit modification that combines a multi-megawatt detection nodes, located in the northern and southern electrical power system, a high thrust and specific lunar hemispheres, could provide nearly complete impulse propulsion system for rapid rendezvous, and sky coverage every month. Each telescope would a pulsed laser ablation payload for changing the have large area mosaic detector arrays target's orbit. This combination of technologies may (approximately 36K × 36K pixels), with the survey offer a future orbit modification system that could telescopes having a 1.0 × 1.0 deg. FOV and the deflect an impactor and provide an effective method for altering the orbits of objects for resource tracking telescopes having a 0.1 × 0.1 deg. FOV. utilization. Spectral imaging would be implemented as early as possible in the detection process. Advanced Besides rapid and controlled trajectory modification, detectors capable of rapid identification of NEOs and one of the goals of CAPS orbit modification is to be their spectral signal could greatly simplify operations effective against NEOs of various compositions. and minimize the requirements on the tracking Asteroids range from primarily stony to mostly telescopes. If NEOs could be uniquely identified in metallic, with various proportions of each type of multiple survey images, a preliminary orbit could be material, and may contain deep, powdery regolith determined with minimal risk of "losing" the object. which can affect deflection efforts, particularly The tracking telescopes would be used in an landing and attaching to the object. Comets contain interferometric mode when higher precision astrometric observations are needed to confirm an a mixture of non-volatile materials and large amounts of frozen volatiles. When a comet becomes object has an impacting trajectory. Finally, active active, these volatiles create a diffuse cloud laser ranging could be used to provide range and surrounding the nucleus called the coma. This range-rate data to augment precision orbit variety of compositions and environments makes the determination. Active laser ranging is preferable to issue of mitigation difficult, and suggests that orbit radar systems due to the potentially large distances modification methods that can move the NEO without landing on it may be highly advantageous. Since time may be critical, this approach could diminish the need for detailed physical characteristic observations to be made before dispatching a deflection effort. Ultimately, a spacecraft capable of rapid interception of an incoming impactor is extremely beneficial, and several approaches for modifying the NEO's orbit could be incorporated into the deflection system. One of the most commonly cited methods for deflecting or pulverizing a threatening NEO on its final approach is the use of a nuclear detonation (Ref. [1]). However, there are many issues associated with this mitigation technique (fragmentation, radiation, etc.), Figure 1.Depiction of Detection System and it is unlikely that the CAPS goal of controlled Using aLunar-Based Approach. orbit modification can be achieved with this 53rdInternational Astronautical Congress October 10-19,2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA technique. Moreover, there is a great deal of time. The quality of the result is affected by errors uncertainty as to how effective a nuclear explosion in the measurements, the spatial and temporal would be against a porous or non-monolithic object spacing of observations, the number of observations, that is effectively a gravitationally bound "rubble and by the number and placement of observatories. pile" (Ref. [2]). Compatible secondary mitigation Using the method of least squares and a batch filter modes could be carried as an additional payload for a (Ref. [3]), we have studied the effects of these deflection mission, or a phased approach using factors in order to identify trends, and make a rendezvous and intercept trajectories with various preliminary determination as to the number, payloads could provide a robust defense. Regardless placement, and resolution of optical instruments of the mitigation method used, rapid engagement of required to form an effective system for determining an object is critical for preventing an impact from a orbits of comets and asteroids. newly discovered LPC or asteroid. An accurate determination of the orbit of a It is essential to understand that the issues associated dangerous body is necessary in order to know if, with detection and deflection of an impactor axe when, and where on the Earth's surface a collision intimately connected, particularly if we are not will occur. The objective of our study is to predict afforded decades of warning time that would be and then prevent such an event; therefore, the likely for large NEAs. The requirements for the conditions required for orbital collisions are central detection system could be significantly reduced to our analysis. For an object in heliocentric orbit given an extremely robust deflection capability. with radius of perihelion rp and eccentricity e, the However, due to the enormous amounts of energy necessary condition for a collision in the ecliptic required to move these massive bodies, any plane at a heliocentric distance of rk is that the additional warning time is an extremely valuable argument of perihelion cosatisfy the relationship asset. rk\ e where the positive sign indicates the condition is imposed atthe ascending node, and the negative sign is associated with the descending node. Generally speaking, the longer the warning time (the interval between the time an object's orbit becomes known and the time of collision), the better the chances of being able to plan and execute action to prevent a collision. We assume a pre-perihelion collision with a warning time of less than one half the object's orbital period, and apply the time-of- flight equation (see Eq. (4.2-9) of Ref. [4]) in a straightforward manner. For the cometary orbits Figure 2.Depiction ofRendezvous Spacecraft with Laser Ablation Payload. studied, warning times range from 2years in the case of an object with rp= 1AU and aphelion ra= 15AU (orbit period of 20 years), detected at a distance of 7 ORBIT DETERMINATION AU from the Sun, to 9.5 months when rp= 0.1 AU and ra= 50 × 103AU (orbit period of 4 × 106years), Orbit determination is the process of using a and the comet is detected at 5 AU. Warning time collection of measurements obtained by observation does not change appreciably as a function of r_in the to calculate a set of orbital elements, six quantities range 1000 < r_< 50 × 103AU, and a reduction in rp that give (either implicitly or explicitly) the position by a factor of 10 reduces warning time by and velocity of an object at a particular instant of -6- 53rdInternational Astronautical Congress October 10-19, 2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA approximately the same amount as a reduction of 1 are obtained by increasing the data arc to 98 days, at AU in detection distance. Warning times for NEAs which point all three values of _ are less than 2 range from approximately 90 days for an asteroid in lunar distances (1 lunar distance = 384,400 km). a 0.9 x 1.4 AU orbit detected at a heliocentric With detection distance held fixed at 6AU, telescope distance of 1.3 AU, to 7 days in the case of an resolution is seen to have a marked effect on asteroid with a 0.2 x 3.0 AU orbit spotted 1.1 AU accuracy, especially for short data arcs, but the from the Sun. length of the data arc is again the most important factor and all three values of _ axe below 1 lunar Orbital elements determined on the basis of distance after 98 days. Unfortunately, longer data observations generally differ from the true orbital arcs yield shorter warning times. elements, which are unknown in practice but are specified in our analysis. A set of six elements p= 0.20 arcsec produced by a batch filter can be compared to the set Oh_elvations centered around5a.u. of true elements; however, it is more convenient to 180 013_el_,atiol-lScentered around6a.u. 160 ..................................................... compare a single parameter if at all possible. We introduce such a parameter s called the erroneous 140 ......... 8 predicted miss distance, defined to be the magnitude o_120 ................................................................................................ of the difference between an object's determined position and the specified position, evaluated at the ,_100 -.."............................................................................................... c specified time of collision. This scalar metric allows dv_o 8o us to choose easily between two observatory configurations that have obtained different sets of >_6o ,< :: : : measurements. 20 A single hypothetical cometary orbit is used to study 40 : ; how the accuracy of preliminary orbit determination 0 10 15 20 25 30 35 40 45 50 based on three observations, each involving two Interval between observations (days) measurements of angular position, is affected by the Figure 3.Average c for Various Detection Distances. length of time of the data arc, the distance at which the object is detected, and the resolution of the A similar approach is used in an extensive analysis telescope. A single observatory (not necessarily of preliminary determination of long-period comet coincident with the Earth) is assumed to travel in the orbits. The data arc is fixed at 66 days, heliocentric ecliptic plane in a circular orbit of radius 1 AU. detection distance is about 6 AU, and a resolution of Each of the six measurements is regarded as the sum 0.1 arcsec is assumed. Hypothetical orbits resulting of a true angle and an error produced with a pseudo- in collision are constructed for 1,008 objects. A random number algorithm, uniformly distributed single observatory is again given a circular between the limits of the telescope's angular heliocentric orbit of radius 1AU in the ecliptic. The resolution, +9. In a Monte-Carlo approach, an orbit effect of observatory position is analyzed by is determined 100 times using a different random allowing the initial true longitude L0 of the number seed in each trial, and an average erroneous observatory to take on 4 values, each differing by predicted miss distance _ is recorded. The length of 90°. Thus, atotal of 4 x 1008 = 4032 cases axe the data arc is varied between 10 and 98 days, examined. When _ is shown as a function of heliocentric detection distance takes on the values of inclination i, one observes that the best orbit 5, 6, and 7 AU, and the resolution 9 takes on the determination is obtained when the orbit planes of values 0.05, 0.10, and 0.20 arcsec (seconds of arc). the observatory and comet are perpendicular, With 9 fixed at 0.2 arcsec, Figure 3 shows how the whereas nearly coplanar orbits yield the poorest accuracy of orbit determination as measured by results. This relationship to inclination stems from becomes better as detection distance decreases, with the fact that, with three observations, an orbit is the greatest improvement occurring at the smallest unobservable (can not be determined uniquely) when data arcs; however, the most dramatic improvements 53rdInternational Astronautical Congress October 10-19, 2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA it is coplanar with the observatory's orbit. With rp = employ the better resolution, as shown in Fig. 4. In 1 AU, retrograde orbits are harder to determine the best case, _ is reduced from 0.14 to less than accurately than prograde orbits. Accuracy with rp = 0.01 Earth radius. 0.1 AU is noticeably poorer than with rp = 0.4 or 0.7 AU, and even poorer with rp = 1AU. A decrease in 4 rais associated with a decrease in orbit determination accuracy. The 4032 values of _ are sorted from 3.5 ............... !................ !................ !................ !................ !................ largest to smallest, the worst and best cases are identified, and the effects of data arc length and L0 _I2° : ............i...............i...............i.................i................. are then studied for these two cases. Observatory location has a significant effect on accuracy, i i i i i 2 particularly in combination with short data arcs, but : : : : : i i i i i the effect becomes less pronounced for data arcs longer than about 70 days. In the worst case, _ is < i i i i i less than 4 lunar distances (for all 4 values of L0) i................ i................ _.............. i................ i................ i................ after 98 days, whereas in the best case _ is better o.5 .- than 0.1 lunar distance, or about 6 times the Earth's radius. i i i i i 3 Number of Observations with Improved Resolution Our results involving preliminary orbit determination corroborate statements made in Refs. [5] and [6], Figure 4. Observations with Mixed Resolutions for Worst Case Orbit. pointing out that the length of the data arc is the single most important factor in determining the accuracy of the orbit solution. The number and The benefits of configurations with 2, 3, and 4 precision of the measurements, the object's observatories (9 = 0.1 arcsec) are studied with the proximity to the observatory when the measurements aid of the worst and best cases. The positions of two are obtained, and even the use of radar observatories are constructed such that they have measurements, are all secondary to the length of the heliocentric circular orbits of radius 1 AU in the data arc. ecliptic plane, and their true longitudes L0 are always 180° apart. Three observatories are placed in similar Improved orbit determination with multiple orbits, with their true longitudes phased by 120°. In observations is studied for the worst and best case the case of four observatories, the first two have comets. A number of observations, ranging from 3 orbits identical to the configuration of two to 99, are taken in equally timed increments over a observatories just described, and the remaining two period of 98 days, and a resolution of 0.1 arcsec is are in similar coplanar orbits perpendicular to the assumed. With a single observatory, _ is reduced in ecliptic with the ascending node _ taking on values the worst case from 3.8 to 0.22 lunar distances, or of 0°, 45°, and 90° in order to determine what effect, about 13 Earth radii. In the best case, _ is reduced if any, _ has on orbit determination. In connection from 2.5 Earth radii to 0.5 Earth radius. The with the worst comet orbit, two or more improvement is pronounced for the lower numbers observatories offer an improvement in 2 of nearly a of observations, and gradual for the higher numbers. factor of 10 over that from a single observatory. Further improvements afforded by measurements With 99 observations each, two observatories yield made with a mixture of resolutions are examined; less than 0.016 lunar distance, slightly less than 1 9 = 0.01 arcsec for initial observations and 9 = Earth radius. Three observatories are only 0.0001 arcsec for some number of final observations, marginally better than two observatories. Four where the total number of observations is 11. Using observatories are not substantially better than three, the method of weighted least squares, _ is reduced and 2 is relatively insensitive to the value of _ for in the worst case from 3.5 Earth radii when all 11 the members of the four-observatory configuration observations are obtained at the poorer resolution, to that have orbits normal to the ecliptic. Similarly, in 0.25 Earth radius when the final 6 observations the best case substantial improvement in 2 is -8- 53rdInternational Astronautical Congress October 10-19,2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA obtained by employing two observatories phased by accuracy of 1 Earth radius is achieved in only 110 180° instead of a single observatory, but the addition days. of a third or fourth observatory does not appear to provide any significant improvement in _. Io1 J , , t i J :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: Large data arcs axe unlikely to be available for objects on a final approach, in which case the I.._::i:_ :............................... Time.of .Fl[ghtl_Z4Z.days................ frequency and resolution of the observations become °_ o[ .........."_:............. _, 1 Lunar Distance i 1o i:•:•:•.:•:• very important. Significantly improving the :::::::::::: ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: [:5 ............ . ............. ............ .............. .............. !............. !.............. resolution of a single observatory would likely ......!....................................................!..i...........!..i................... require the telescope diameter to become excessive. _ ...........i..............:.............:.............:.............:..............i............... Every order of magnitude decrease in the resolution _ : : : : : : 13=0.1 ..... :: :: requires a telescope with a diameter 10 times larger. _IO I :::::::::::::::::::::: ::::::::::::::::::::::: ==================================================== Optical interferometry is another approach to :::::::::::::::::::::::::::::::::::::::::=::=::=::=::=::=::=:::=::==========================================:::i:::::::::::::: improving the resolution by combining the light ............ ............ p=0.000t from multiple telescopes of a more moderate size. Multiple observatories widely spaced in heliocentric 10 0 1oo 150 200 250 300 350 space axe unlikely to lend themselves to this Time (days) advanced observational technique because of over- Figure 5.Improvement inWarning TimeResulting from resolution and disappearance of fringe patterns. If Improvement inAngular Resolution. multiple telescopes at a single location can be combined into an interferometric system, the effective system resolution can be greatly increased We have examined 180 orbits of hypothetical NEAs when needed for precise orbit determination of a of the interior kind with rp< 1AU andra= 1AU, as particular object. well as 1350 orbits of exterior NEAs with rp< 1AU and ra > 1 AU. The focus of the analysis is on Using a Kalman filter, which allows observations to asteroids discovered less than one orbit period prior be incorporated into the orbit determination solution to collision. Preliminary orbit determination is as they become available, the worst case comet orbit performed with 4 observations of 0.1 arcsec has been analyzed using observations from a single resolution taken over a 33 day data arc ending 16 observatory with various angular resolutions to days before collision. Initial observatory true demonstrate how improved resolution results in longitude is again varied in increments of 90°, longer warning times. Figure 5 shows the amount of resulting in a total of 6120 cases involving asteroids. time required to reduce _ to 1 lunar distance, and Of the 720 interior cases, 697 values of _ are less then 1 Earth radius, assuming all observations axe than 0.05 lunar distance, or about 3 Earth radii. possible. The ultimate purpose of the CAPS Accuracy suffers for orbits with low inclinations, detection system isto maximize warning time, which which is not unexpected because orbits that axe we can now take to be the difference between the coplanar with the observatory require a minimum of time until collision (747 days in this case), and the 4 observations if they axe to be determined uniquely. time required to reduce _ to 1 Earth radius or less. Of the 5400 exterior cases, 5389 yielded _ less than In this comparison, one observation is taken every 7 1lunar distance. Relatively poor orbit determination days from a single observatory with angular is exhibited once again for orbits with low resolutions of 0.1 or 0.01 arcsec until _ isreduced to inclinations. Accuracy generally decreases for 1 lunar distance. At this time, the frequency of smaller values of rp,and for larger values of r_. With observations is increased to once per day. It takes each type of asteroid, 9 cases contained some approximately 300 days to reduce _ to 1 Earth number of trials out of 100 that exhibited radius with 9 = 0.1 arcsec, and 140 days with 9 = convergence problems. 0.01 arcsec. If multiple telescopes with an initial resolution of 0.01 arcsec can provide an effective The results indicate that, as long as the interferometric resolution of 0.0001 arcsec, an measurements are available, it may be possible to -9- 53rdInternational Astronautical Congress October 10-19, 2002 IAC-02-IAA. 13.4./Q.5.1.01 Houston, Texas USA make reliable forecasts with two observatories intensity pattern shown in Fig. 7. The position of the whose angular resolution is on the order of 0.1 internal delay line that gives the maximum intensity arcsec, or with a single observatory whose resolution is equal to the external path delay and determines the is better by 1 to 3 orders of magnitude. Additional value of D. This delay position is then used with the study of these two alternatives will have to weigh the equation advantages in geometry and redundancy of multiple D=lBIcosO observatories against the expense of putting them in place and maintaining them; a single observatory near Earth could be easier to maintain, but two to solve for 0, which gives the angle between the widely spaced observatories may offer observational object pointing vector S and the baseline orientation advantages, particularly for LPCs. vector B. The baseline vector orientation in inertial NEO ASTROMETRIC 1NTERFEROMETRY Increasing the angular resolution for measurements of NEOs increases the precision of determining their relative position in the sky which, in turn, decreases the error in the orbit determination. Analysis shows that the relationship between resolution and _ is approximately linear, where dividing the resolution by two, for example from 0.2 to 0.1 arcsec, will result in a decrease in _ by a factor of two. Also, improving the resolution can result in significantly more warning time, increasing the likelihood that mitigation efforts will be successful. Therefore, any method that can provide higher resolution angular measurements of NEOs will be beneficial to orbit determination, especially when observational time is limited. Astrometric interferometry is one method that has the potential to improve angular resolution by several orders of magnitude. Theor_ Figure 6.Basic Interferometry Layout. Interferometry is the measurement of interference fringes produced by combining the light from two or more telescopes that observe the same source. Figure 6 shows the basic setup of an interferometer. delected + The vector between the two telescopes is known as the baseline B. The unit vector from the centerline of the interferometer to the source being observed is called the position vector S. The observable in astrometric interferometry is called the delay D. Since the target being observed does not usually lie on the perpendicular axis of the interferometer, the o light from the source will reach one telescope before external deny -internal delay the other as shown in Fig. 6. In order to interfere the same photon, a delay line must be introduced in the telescope that receives the photon first. As the delay Figure 7.Interferometric Intensity Pattern. line is varied, the interferometer measures an -10-