e-Prints

The widely discussed use of US reconnaissance satellites during the Gulf War will strongly motivate future regional adversaries to seek ways of countering US space-based assets. The presumption that reconnaissance satellites can operate covertly is obsolete. Tracking US reconnaissance satellites can provide valuable support to a hostile country's concealment and deception programs. Iraq's ability to conceal both major weapons programs and many SCUD launchers is a warning of the serious consequences such programs can have. Space surveillance systems of the type likely to be acquired by Third World countries are inconspicuous and may well go undetected, while direct ascent ASAT rockets are within the reach of many countries. This article argues that fundamental reexamination of the functions and architecture of US overhead reconnaissance is needed, and should be done outside the traditional Cold War bureaucratic structures.

Mr. Thomson is employed as a senior scientist in a defence analysis company in the Washington, D.C. area. From 1972 to 1985 he was employed as an analyst by the US Central Intelligence Agency, where he wrote many documents, including contributions to several national intelligence estimates.

This material has been reviewed by the CIA. That review neither constitutes CIA authentication of information nor implies CIA endorsement of the author's views.

It has been public knowledge for many years that the U.S. possesses photoreconnaissance satellites in low earth orbit (LEO).[1] There are also sources which assert that American space-based reconnaissance resources include imaging radars and signals intelligence (SIGINT) satellites in various orbits, from LEO to geosynchronous orbit (GEO).[2] In the words of one Russian author, [3]

The orbital grouping which supported the operations of multinational forces [during Desert Storm] included more than 20 spacecraft of imaging ('KH-11', 'Lacrosse') and SIGINT ('Ferret', Chalet', 'White Cloud', 'Aquacade') reconnaissance...

While this information has long been in general circulation, up until the Desert Storm operation of 1991 there was an aura of remoteness about the subject. Despite occasional indications that reconnaissance satellites might be used for tactical purposes, there seems to have been a general perception that satellite- derived intelligence only pertained to the arena of 'superpower' affairs -- monitoring strategic arms control treaties or supporting political and military confrontation between the U.S. and the U.S.S.R.

The wide and often official publicity given satellite reconnaissance during Desert Storm marks a fundamental and extremely important break in the world's perception of the use of spacebased intelligence. 'Lessons learned' from the Gulf War by all the world's military commands must include an appropriate assessment of the tactical use of space.[4]

Heightened international interest in offensive, or at least proactive, use of satellite reconnaissance has been manifested in the increased activity by France in designing the Helios photoreconnaissance satellite. The recent interest of the United Arab Emirates in acquiring a one-meter-resolution reconnaissance satellite probably is due to the impression US imagery made on the Coalition partners during the Gulf War.[5]

Of greater concern for American intelligence planners is that Desert Storm has prompted states which might find themselves in conflict with the USA in the future to develop countermeasures against US space-based reconnaissance.[6]

The first order of business for a Third World entity hoping to develop countermeasures to space reconnaissance is to identify the satellites involved and to obtain as detailed as possible an understanding of the threat they pose. One route to obtaining such understanding would be agent espionage -- a powerful but chancy technique which, even when it works, is susceptible to counterintelligence and, worse, disinformation. Similar problems arise with leaks of classified information appearing in the open press -- perhaps some of the reported 'facts' are accurate, but it is difficult to sort out conflicting stories and to detect possible disinformation. More certain knowledge comes from technical methods of collection and analysis. The question then arises: can a country with limited resources develop a system to carry out effective space surveillance and space object identification (SOI)? This question is usually answered in the negative, based upon the misperception that to create such a system, a country would have to duplicate the very large and expensive phased array radars which the U.S. and the former Soviet Union constructed for ballistic missile and satellite detection and tracking.

There are, however, many commercially available technologies which can be used to construct systems to detect, identify, and track US reconnaissance satellites in LEO. Third World countries could use this capability to conceal sensitive activities and to attack satellites with rudimentary antisatellite (ASAT) weapons. Though not as imminent as the threat to LEO systems, world-wide progress in sensors, computation, and propulsion is so rapid that higher-altitude satellites may also become vulnerable in the early years of the next century.

Tracking Satellites

The Minimal Technology Approach

There is already an example of a 'satellite surveillance organization' which uses minimal technical means: the worldwide network of amateur satellite observers. Tracing their origins back to the Moonwatch program of the late 1950s, the participants in this network use stopwatches, sky maps, personal computers and (optionally) binoculars. Observing at twilight, the amateurs note the positions of satellites with reference to the star field, and then use programs based upon methods developed by nineteenth century astronomers to determine the satellites' orbital elements.[7] The amateur community believes that this technique, when carefully applied, may yield results at least as accurate as those generated by the large radars used by the US Space Command. Increasingly, the amateurs organize observing campaigns and report their results using computer bulletin boards or the Internet. Because the USA does not release the orbital elements of its classified satellites, American 'spy satellites' are regarded by the amateur community as a special challenge, and systematic efforts are often mounted to detect the payloads of classified launches. These searches are generally successful, and the amateur community maintains orbital elements for most classified U.S. vehicles in LEO.[9] Except for very low or maneuvering satellites, orbital elements do not change very quickly, and reacquisition after an observing hiatus of weeks or even months is fairly easy.[10] The identities and missions of the satellites are derived from press stories, supplemented, as discussed below, by analysis of orbits and visual appearance.

Table 1. Orbital elements resulting from amateurs' observations

KH 11-6
1 15423U 84122 A 91242.81022954 .00001448 00000-0 18305-3 0 01
2 15423 97.7329 312.0831 011909 269.7143 88.9210 14.76965400 09

KH 11-7
1 19441U 87090 A 91243.96423057 .00012900 00000-0 36433-3 0 05
2 18441 97.7030 354.2820 0435739 261.5101 93.5343 14.7764260 00

KH 11-8
1 19625U 88099 A 92 61.79769562 .00040000 00000-0 28467-3 0 03
2 19625 98.2311 125.4679 0565956 184.1621 175.3456 14.75708308 06

Lacrosse 1
1 19671U 88106 B 92 70.09292299 .00003000 00000-0 49942-3 0 04
2 19671 56.9684 302.7068 0009100 273.0752 86.9247 14.69700000 09

NOSS 2-1 (B)
1 20682U 90050 B 91363.03070479 .00000080 00000-0 14505-3 0 01
2 20682 63.4250 218.0475 0055000 154.3842 205.6158 13.40165000 08

NOSS 2-1 (C)
1 20691U 90050 C 91363.03089751 .00000080 00000-0 14505-3 0 06
2 20691 63.4250 218.0724 0055000 154.4292 205.5708 13.40165000 08

NOSS 2-1 (D)
1 20692U 90050 D 91363.03080878 .00000080 00000-0 14505-3 0 08
2 20692 63.4250 217.6006 0055000 189.1923 170.8077 13.40165000 08

Lacrosse 2
1 21147U 91017 A 92 39.07763545 .00002000 00000-0 35681-3 0 04
2 21147 67.9 880 259.1450 0006000 343.0000 17.0000 14.66882000 00

NOSS 2-2 (C)
1 21799U 91076 C 91354.18039860 .00000090 00000-0 16556-3 0 08
2 21799 63.4169 123.5891 0074714 159.8750 200.4212 13.40163259 02

NOSS 2-2 (D)
1 21808U 91076 D 91354.25508020 .00000092 00000-0 16434-3 0 01
2 21808 63.4284 123.4380 0090080 158.1630 202.2230 13.40162530 06

NOSS 2-2 (E)
1 21809U 91076 E 91354.47877270 .00000092 00000-0 16544-3 0 04
2 21809 63.4253 122.3765 0076370 186.0834 173.8234 13.40163120 08

These elements are provided as a service to visual observers. They are uploaded weekly to the Canadian Space Society's BBS in Toronto, Canada. This is a free BBS, operating 24 h/d, <= 2400 B, 8N1, phone 416-458-5907.

With the activities of the amateurs as a proof of concept, it becomes clear that virtually any country can organize a reasonably effective surveillance capability against objects in LEO. A natural model for this would be a corps of military attaches which could be assigned to a country's embassies to carry out observations in the same way as the amateurs. With a set of dedicated observers in many different cities around the world, the probability that any satellite would go unobserved for long due to unfavorable lighting conditions or weather is small. As will be discussed below, some SOI capabilities are within easy reach of many countries, but the number of large, classified U.S. satellites in LEO -- around half a dozen by the amateurs' count - is sufficiently small that precise identification of mission and capabilities is not particularly necessary. An assumption that they are all imaging reconnaissance platforms of some sort would suffice for the purposes of a concealment and deception (C&D) effort, or to rationalize an attack against them.

Surveillance Improvements

If a country wished to acquire a better LEO surveillance capability, there are several obvious improvements to the rudimentary system discussed above. Unsurprisingly, the increasing availability of charge-coupled device (CCD) detectors and inexpensive computers afford the greatest opportunities for enhancing an optical surveillance system. CCD 'chips' are the solid-state imaging devices used in videocameras, and, besides being much more sensitive than the eye or film, are easily interfaced with computers for control and data collection.

Another advantage of CCD cameras is that they are not restricted to the traditional twilight observing hours. Work in the U.S. in the early 1980s, and more recently in the former Soviet Union, has shown that CCDs used with telescopes of modest aperture can detect even faint objects in the daytime sky. [11] Implementation of this technique would increase the possibilities for quickly placing newly launched or maneuvering objects in track.

In addition to optical techniques, other options are available for relatively low-cost LEO surveillance. Besides phased array radars, there are other kinds of radar which are much easier and cheaper to build and operate. One such is the US Navy's Navspasur system. It uses powerful radio transmitters in Texas, Arizona and Alabama to radiate fan-shaped beams into space. Echoes from satellites passing through the beams are received at a number of sites across the U.S., and analysis of the signals provides the orbital parameters of objects down to a fraction of a kilometer.[12]

Independent of weather and lighting conditions, the NAVSPASUR radar has several other characteristics which are relevant here:

It was designed more than 30 years ago using the technology of the time. Although it has been upgraded since, it has performed well from the start. Equipment with the technical characteristics of the system's transmitter and receivers is widely available on the open market today, and building a transmitter site is a medium-sized civil engineering job.

Data reduction is well within the power of personal computers. In fact, amateur satellite watchers and ham radio operators process NAVSPASUR returns as a hobby.[21] (This suggests the possibility that an Nth country intelligence service might simply set up clandestine Navspasur receiver sites in the U.S., Canada, Cuba, or Mexico.)

Another, more modern, but still comparatively simple unconventional radar is Japan's MU (Middle and Upper atmosphere) instrument. This radar was built in 1984 to carry out a variety of meteorological and aeronomical studies. In support of the Japanese contribution to the US space station program, MU performs orbital debris surveys for 48 hours each month.[14] Though not as massive or complex as the American and Russian phased array radars, MU is considerably more sophisticated than Navspasur, and may be more characteristic of a system which a wealthier Third World country might acquire.

It has also been proposed that the signals from some powerful commercial radio and TV transmitters could be used for space surveillance and tracking if properly processed. This idea, while interesting, needs further study to determine its feasibility. In addition, there are several other techniques, some of them in the category of folklore, which have been suggested for HF detection of satellites, and which ought to be examined as possibilities which other countries could exploit.

The existence of systems such as Navspasur and MU implies that virtually any country that decides it needs a 24-hour, all weather LEO surveillance system can acquire one fairly cheaply, and probably without having to circumvent export restrictions.

Surveillance of geosynchronous orbit

While GEO is intrinsically more difficult to monitor than low orbit because of the much greater ranges involved, applicable technology is evolving rapidly. Already, CCD cameras are used to detect larger satellites in geosynchronous orbit. Advances in sensor and computing technology will make this capability accessible to more countries as time goes on. Use of compensated imaging techniques of various sorts will enhance the detection capabilities of even small telescopes. Moreover, GEO is becoming increasingly accessible to a range of nations, raising the possibility that space surveillance packages

could be placed in the vicinity of suspect satellites.

Space object identification

At the simplest level, analysis of orbits and launch-related information often gives clues to a satellite's mission. The amateur community uses this technique to make deductions about the payload of classified missions: for example, the sunsynchronous orbits of some of the objects identified as "KH-11" photoreconnaissance satellites appear to confirm some sort of visual imaging function. Monitoring of orbital elements can also detect when satellites maneuver, and gives clues to the missions of specific objects of interest. This technique was used by Nicholas Johnson in his "Soviet Year in Space" series to deduce the tasking of the former USSR's photoreconnaissance satellites.[15]

Accruing to the amateur community, visual characteristics of satellites are used to establish family resemblances -- for example, the naked-eye appearance of the two "Lacrosse" satellites is unique to them, and was used as evidence that the second, which was launched on a Titan, was of the same class as the first, even though the first had been launched by a Shuttle into a somewhat different orbit. The "NOSS" (Naval Ocean Surveillance System) satellites, believed by the amateurs to be US Navy electronic intelligence (ELINT) collectors, fly in unique and striking formations of three, making them easy to distinguish from all other satellites. [16]

An attractive supplement to orbital analysis and simple visual observation for SOI is the technique of short-exposure imaging, which makes it possible to obtain high-resolution images from telescopes of up to about a meter aperture without the use of adaptive optics.[17] This method is currently used in the astronomical community to obtain sharp images of the planets and other bright objects. Short-exposure imaging requires only conventional telescopes, CCD cameras able to take exposures of a few to a few tens of milliseconds, and a mount capable of smoothly tracking a satellite as it moves across the sky. The statistics of atmospheric turbulence, which blurs longer exposures, guarantee that there will be occasional, momentarily smooth patches of atmosphere through which moderate-aperture telescopes can realize their full resolving power. Post-exposure visual examination of the sequence of frames taken during an observing session uncovers the few 'good' frames, which may then be enhanced using standard imageprocessing techniques.

If applied to SOI, CCD imaging can be used to obtain pictures of satellites in LEO with resolutions in the 30 - 100 cm range. Unclassified USAF CCD images of the Shuttle in orbit show the utility of this approach, which may also be used in daytime. If implemented by other countries, short-exposure imagery could provide at least a rudimentary SOI capability and further complicate deployment of US satellites.[18] In the longer term, the availability of adaptive optics techniques of various sorts will bring blur-free performance to larger telescopes, with proportionally greater resolving power.[19]

The explosive growth of optical technology at the end of the twentieth century is also opening up methods for extending optical SOI techniques to higher orbit. Of most interest for surveillance of GEO is the technique of optical interferometry, which the astronomical community is just starting to use. With this technique, it is possible to connect several telescopes into arrays with effective apertures of hundreds of meters; this would give ten-centimeter resolution at geosynchronous orbit.[20] Though not a prospect for immediate future, it is certainly not out of the question that optical interferometers will be built in the Third World in the first decade of the next century and be used for examination of satellites in high orbit, as well as for astronomy.

Other geosynchronous SOI techniques, such geosynchronous surveillance packages, will become increasingly available as more countries develop an ability to place satellites into GEO.

The above discussions of space surveillance and SOI options by no means exhaust the possibilities which are becoming available as new technologies reach the market. Even so, it is clear that the presumption that U.S. reconnaissance satellites can operate covertly is largely invalid today, and will become more so in the future.

Measures to counter satellite reconnaissance

If future adversaries develop the ability to track and characterize US satellites, it is important we understand what actions these adversaries can take against the satellites. Here too, there is a broad spectrum of possibilities. Those discussed below concentrate on measures which might be taken against the satellites themselves, and omit such serious possibilities as attacks on launch facilities and ground stations.

Concealment and deception

The accuracy of orbital predictions using visually-generated orbital elements and publicly available software is great enough (typically, satellites pass on track and within a minute or so of the predicted time), that a country or other entity wishing to organize a satellite warning service as part of a C&D plan could do so. Undoubtedly, the major part of the effort would be training operating personnel to take effective C&D measures when warned of an impending overflight. Implementation of a CCD-based daylight detection capability at several sites, or of a Navspasur-like space surveillance capability would help maintain currency of the warning information if satellites maneuvered or a new vehicle were launched.

Just as Desert Storm has shown the rest of the world the extent to which the U.S. relies on satellite reconnaissance for military support, Iraq's ability to conceal both major weapons programs (on the strategic scale) and SCUD launchers (on the tactical scale) should warn us of the serious consequences of competently conducted C&D programs.

Electronic warfare and laser attacks

Vulnerability of satellite systems to jamming and component damage of various sorts depends strongly on particulars of design, and so it is very difficult to make general statements about potential threats. However, the very fact that LEO satellites can be accurately tracked and are at short ranges (hundreds of kilometers) makes it possible for an enemy to concentrate large fluxes of energy on them. At least for imaging optical systems, this is worrisome in and of itself, because their very functioning depends on concentrating incident light on a small, sensitive detector.

Direct ascent ASAT

Although the USA has tended to think of ASAT threats in terms of the former Soviet co-orbital system, near- to mid-term Third World systems are much more likely to use a direct ascent scheme. This is due primarily to the much simpler and cheaper rockets required: for example, a sounding rocket capable of lifting two hundred kilograms to a thousand kilometers altitude weighs an order of magnitude less -- a few metric tons -- than a space launch vehicle capable of orbiting similar mass. Because manufacturing costs grow more rapidly than the weight of rocket systems, the cost of a direct ascent ASAT can be only a few percent of that of a co-orbital system. This means not only that direct ascent systems are more easily procured by smaller countries, but also that they can be produced and employed in quantity. Sounding rockets with the required characteristics are in use in a number of nations, and have formed the basis for the first-generation Japanese, Indian and developmental Brazilian space launch systems. Besides purpose-built rockets, modification of medium-range ballistic missiles to loft payloads to altitudes of several hundred kilometers is a practice that goes back to the late 1940s. Direct ascent ASATs can be fired from mobile launchers, making them much more survivable than co-orbital interceptors, which generally must use fixed facilities.

Direct ascent engagement opportunities against the LEO payloads tracked by the amateur community are reasonably frequent for sites at middle latitudes. Table 2 illustrates the opportunities for a hypothetical site near Tehran to launch within a one-week period in early 1992 against four of the satellites tracked by amateurs -- two "KH-11" and two "Lacrosse" satellites --under fairly stringent constraints on the geometry of the engagement. As can be seen, there is at least one opportunity against each of the satellites, and several against "Lacrosse 1." This difference is because, due to orbital timing, favorable periods alternate with unfavorable ones for a given satellite.

If an enemy country can find and track satellites, and can fire significant payloads to their vicinity, the remaining part of the ASAT problem is to ensure that the payload, now regarded as a nonnuclear warhead, pass sufficiently close to the satellite to have a significant chance of damaging it, even though their relative velocity is several kilometers per second. (The relative velocity is not only a problem but also an advantage, as it ensures high lethality for any warhead fragments which collide with the target.[21]) In the context of attacks on very high-value satellites using inexpensive direct-ascent weapons, multiple launches against the same target are reasonable. End-game guidance is classically by far the most difficult part of the ASAT problem, and it is inappropriate to discuss specific possible solutions here. However, the general technological trends noted earlier -- very rapidly advancing and diffusing sensor, computation, and communications capabilities -- make it imprudent to assume that problems which were difficult to solve even ten years ago will remain outside the capabilities of all but the most advanced countries indefinitely.[22]

Implications

In the areas of space surveillance and SOI, we have seen that the potential for Third World acquisition of considerable capabilities is large and growing. The growth is fueled by the development of electronic and electro-optical devices in the worldwide civilian marketplace, and for this reason is certain to continue and to be available to any group with even limited financial resources. The fact that relevant technology development is now market-driven also has the very important consequence that the rate of change is characteristic not of major military systems --ten to twenty years-- but of consumer and commercial electronics -- three to five years. Plans for dealing with Nth-country space surveillance capabilities must take this very rapid evolution into account, and consider that the technical capabilities of a particular country will be not so much determined by its indigenous technical base as by market availability. Also significant is the inconspicuousness of many of the space surveillance techniques likely to be acquired, especially those using optical approaches. It is quite possible that a country (or other group) could equip itself with a very respectable surveillance and SOI capability without that fact ever becoming known to the USA 'Absence of evidence' should by no means be taken as 'evidence of absence.'

In a somewhat similar vein, the crucial part of direct ascent ASAT systems -- the terminal engagement guidance and fuzing mechanisms -- is dependent on the same very rapidly developing and proliferating technologies mentioned above, with the same implications for US planners. Moreover, the low cost of the boosters needed, the probably low cost of the associated guidance mechanisms, and the independence from fixed launch facilities makes it likely that an aspiring ASAT power will think in terms of multiple engagements against a single target, possibly using salvos of ASATS fired from different locations each time. Laser and electronic warfare attacks will also involve multiple engagements carried out over days or weeks. Finally, most of the development and even deployment of the ASAT systems sketched out here could, like the corresponding surveillance systems, be hard to detect and easily hidden by a determined opponent. Whether a full-up test involving launch of a missile against a real or simulated satellite -- perhaps a discarded booster rocket -- could be carried out covertly is an interesting question deserving further examination.

What is to be done?

The largely unanticipated collapse of the Soviet Union has radically changed the context in which the USA must think about matters related to national security. One of the most significant changes is the emergence of regional powers, typified but by no means limited to Iraq, with strongly pursued goals inimical to US interests. Finally, there is the ever accelerating pace of technological advancement, which places capabilities formerly considered to be 'advanced' in the hands those regional powers. The changed context necessitates revisiting questions which were believed, correctly or not, to have been well understood in the Cold War environment, and, more importantly, makes imperative revisions in the ways in which we think about them.

As we have seen, the conditions of the post-Cold War era may adversely affect the United States' ability to conduct satellite reconnaissance with the complete freedom it has enjoyed since the 1960s. The issues are complex both technically and politically, and must be thought out carefully. Potential solutions must be debated openly if viable responses are to be found. No pretense to a final solution can be made here, but some general approaches are worth reviewing.

Techniques for enhancing satellite survivability

Many papers and monographs have been written on the topic of satellite survivability[23]. An example is a lengthy monograph written by Steven Peterson 1991.[24] Although assuming superpower confrontation and conflict as the background, his discussion can be used as a point of departure for consideration of future problems.

Survivability techniques accomplish one of three purposes: they make the satellite or system hard to find, hard to hit, or hard to kill. Deception techniques hide the satellite or system. They include stealth or masking designs (reduced radar and infrared/optical signatures). Satellites can also be placed in deep-space storage orbits (even beyond geosynchronous) and maneuvered down as needed. When an enemy discovers and targets one of our satellites, we can make interception difficult or impossible by maneuvering the satellite or by ejecting decoys from the satellite... However maneuvers must be balanced against the need to replan subsequent satellite usage, the likelihood of temporarily disrupting the mission, and the decrease in satellite lifetime caused by the expenditure of limited fuel... other methods such as autonomy and encryption may also enhance survivability.

The U.S. could also choose to position defensive satellites (DSAT) near its high-value systems. DSATs would use active measures [lasers, missiles, electronic countermeasures] to eliminate attacking ASAT weapons.... Finally nuclear and laser hardening would prevent easy kills, lengthen system lifetimes, and complicate an opponent's damage assessment process...

Some comments on the above approaches are called for in light of our earlier analysis of the probable nature of Third World surveillance and ASAT systems.

Deceptive techniques may be extremely valuable on tactical timescales of hours to days, but not for longer periods. The variety of methods which may be used to detect and analyze satellites is so diverse, and the possibility of compromises of secrecy through espionage or press leaks so great, that any deception scheme must be assumed to have a finite, probably short lifetime.[25] This is acceptable in times of war or extreme crisis, but would be a waste of potentially valuable capabilities if undertaken in the open-ended timeframes of peace. Worse, reliance on strategic deception for hiding reconnaissance satellites carries with it the very real, and extremely dangerous, possibility of self-deception. If an enemy discovers satellites which the owner believes he does not know of, he can then use those satellites as channels of disinformation which will probably be believed -- the modern technical equivalent of 'turning' an agent. In addition, of course, discovered satellites can be targeted for attack.

Hiding satellites in very high or unusual orbits may be an effective alternative to 'launch on demand', but the operational and budgetary impacts could be a major impediment. Optical reconnaissance satellites of conventional design cannot go very high without losing needed resolution. For example, the Hubble Space Telescope would have a ground resolution of about 10 centimeters (4 inches) from 500 km altitude, but only one meter from 5000 km, and eight meters from geosynchronous orbit. Unconventional designs such as segmented and multiple mirror telescopes or optical interferometers can in principle overcome this problem, but will be difficult and expensive to design, assemble, launch and deploy in the near to mid term. (But optical interferometers, which can consist of a loose collection of small telescopes operating in a coherent array may be attractive in the next century.) Cost is also a factor, since it is much more expensive to put a given amount of mass into high Earth orbit than into LEO.

Hardening, in principle a good idea to 'prevent easy kills', needs careful assessment in the new environment. Since the most likely kill mechanism is hypervelocity impact, which is extremely difficult to protect against, it may be that the most cost-effective solution will involve some level of subsystem redundancy to increase satellite survivability against small particles, but no major attempts to protect against larger impacts. Laser and EW hardening will continue to be desirable, as long as they do not unreasonably compromise overall system design (including indirect effects via the budget).

DSATs do not seem very promising if the defended satellite is subject to multiple attacks by low-cost weapons which disperse small, inert projectiles.

Deterrence, often mentioned in Cold War ASAT discussions, may be difficult to carry out effectively against a Third World country, particularly in wartime.[26] Deterrence in kind -- destroying an attacker's satellites in retaliation for loss of one of one of the user's -- is infeasible if the attacker does not operate satellite systems. Deterrence by massive attacks against the attacker's country, or by radical escalation of ongoing military actions, must be carefully considered in terms of the political situation at the time. The USA could bring on itself severe moral condemnation from the world community if it were to take drastic military action in response to the destruction of a system which is acknowledged to be a tactical reconnaissance asset. Even if the USA were willing to risk such general condemnation, the delicate nature of the Coalition during the Gulf War showed that there are practical reasons for refraining from military escalation.

Carrying any particular defensive measure to the extremes needed to provide good protection for a single satellite is likely to be extremely expensive, and thus precludes having more than a very few satellites on orbit at any one time. This is a situation tailor- made for catastrophic failure through any of a number of causes, whether enemy action, mechanical failure, or simply bad luck.

As an alternative to defending individual satellites, then, it is reasonable to look at defending the function of the system of which the satellite is a part. Again, Peterson discusses these issues:

Overall, the above approaches offer significant survivability to individual systems components. Other, broader approaches require consideration of the system architecture as a whole...

[By taking] this top-down perspective, system designers and planners can realize additional survivability by sizing and positioning satellite segments in nontraditional ways. Most of these efforts hinge on satellite proliferation, especially through the use of smaller satellites dedicated to one mission.

Designers could allocate satellite capability to a distributed network rather than to a few high value satellites, thus reducing reliance on any single satellite... Proliferation would be difficult for complex systems such as high-resolution reconnaissance, [but could] be suitable for communications relay, nuclear detection, and tactical reconnaissance...

...Proliferation acknowledges that some attrition is likely during wartime. Thus an important aspect of any survivability planning requires addressing the requirement to replace lost assets.

Replacement requires available ground spares and a capacity for making quick-response launches... Current U.S. satellites tend to be large and heavy, and require extensive periods of ground preparation and on-orbit checkout. While efficient in peacetime, these systems do not lend themselves to rapid replacement. Thus, the need for rapid replacement drives the need to build smaller satellites.

Survivable, flexible launch systems are also essential. Options include mobile ground, air or sea systems capable of launching small payloads quickly... One possible solution is the Pegasus air-launched rocket,... [which can] place a 400-pound payload in low earth orbit.

Organizational problems and coping with the future

It is clear that the developing situation with regard to the vulnerability of our space-based security assets, and possible remedial actions, is extremely complex. Careful assessment of emerging surveillance and ASAT technologies, as well as of available responses, must be carried out and must be reexamined on a regular basis due to the rapid changes taking place in the technical and political world. A broad, unbiased look must be taken at the functions performed by satellites, the environments in which the satellites will have to operate, and the potential for rapid change and surprise. Fortunately, the opportunities for carrying out such an examination now exist. No longer subject to the life-and-death urgencies of the U.S.-Soviet confrontation, we can now undertake studies to understand the technical, operational, political and budgetary implications of the changed environment. Unfortunately, the organizations and practices which grew up around national security space systems during the Cold War are now a serious impediment to analyzing and meeting the challenges which will face us in the coming years and decades.

Among the most acute of the institutional problems is the system of secrecy associated with reconnaissance satellites. The structure of special classification channels, "compartments" within the channels, and levels of access within the compartments within the channels has long since passed the legitimate needs of security and now appears as full-blown organizational dysfunction. Although this system evolved through conventional bureaucratic dynamics in conditions of self-defined priorities, the nation will be ill- served to continue its use.[27] Recent moves to relax the security environment -- for example the acknowledgment of the existence of the National Reconnaissance Office -- are commendable, but far from adequate.[28] At a minimum, the special 'security' channels and compartments associated with overhead reconnaissance should be reexamined, the principal missions and characteristics of reconnaissance satellites declassified, and planning for future systems opened to general discussion and critique, much as is the case for other weapons systems.

In addition, while there may continue to be a need for national-level reconnaissance systems, serious consideration should be given to giving the Services responsibility for developing, procuring and operating tactical reconnaissance assets. Freed from excessive secrecy, subject to the real-world requirements of operators, driven to efficiency by budget pressures, and open to commercial innovation and competition, it would be surprising if space-based reconnaissance systems did not see orders-of-magnitude improvements in their cost and utility -- improvements which could, in turn, benefit national-level systems. Conversely, if remedial measures are not taken and planning for the future is left solely in the hands of the established

Cold War institutions, the USA will almost certainly fail to keep pace with the world of the future and, sooner or later, find itself denied the use of the reconnaissance assets on which it has come to depend so much.

REFERENCES

[1]President J. Carter, KSC, 1 October 1978; Weekly Compilation of Presidential Documents, 9 October 1978, 16841687;

Angello Codevilla Informing Statecraft: Intelligence for a New Century The Free Press, New York N.Y., 1992

John Ranelagh The Agency: The Rise and Decline of the CIA Simon and Schuster, New York, N.Y., 1987

R.S. Cline The CIA Under Reagan Bush and Casey Acropolis Books, Washington, D.C., 1981

[2] A. Kuznetsov 'Tekhnicheskiye vozmozhnosti kosmicheskikh sredstv SShA'(Technical Capabilities of U.S. Space Assets) Voyennaya mysl', No. 8-9, 1992

Jeffrey T. Richelson America's Secret Eyes in Space Harper & Row, New York, 1990

Philip Klass 'NSA 'Jumpseat' Program Winds Down As Soviets Shift to Newer Satellites' Aviation Week & Space Technology April 2, 1990;

A. Andronov 'Programma ekspluatatsii MTKK "Shattl"' [Using the Shuttle] Zarubezhnoye voyennoye obozreniye, August, 1992

A. Andronov 'Amerikanskiye sputniki radioelektronnoy razvedki na geosynchronnykh orbitakh' [American Geosynchronous SIGINT Satellites] Zarubezhnoye voyennoye obozreniye, No.12, 1993

A. Andronov 'Kosmicheskaya sistema radiotekhnicheskoy razvedki VMS SShA "Uayt klaud"' [The U.S. Navy's "White Cloud Spaceborne ELINT System] Zarubezhnoye voyennoye obozreniye, No.7, 1993

[3] Kuznetsov op cit, Ref.2

[4] Kuznetsov op cit, Ref 2; and L.N. Doda 'O demaskiruyushchikh priznakakh kosmicheskikh sredstv razvedki' [Reconnaissance Satellite Operations as Indicators of Military Operations] Voyennaya mysl', No.10, 1992

[5] 'Matra Marconi Space Proposes Expanded Program for Helios' Aviation Week & Space Technology, May 20, 1991

Final Report to the President on the U.S. Space Program, January, 1993 USGPO, Washington, D.C. (ISBN 0-16-041608-6)

[6] Michael E. Baum 'Defiling the Altar -- The Weaponization of Space' Airpower Journal, Spring, 1994

[7] D. King-Hele Observing Earth Satellites Van Nostrand Reinhold, New York, 1983

P.R. Escobal Methods of Orbit Determination Wiley, New York, 1965

R.R. Bate, D.D. Mueller, J.E. White Fundamentals of Astrodynamics Dover, New York, 1971

D.L. Boulet Methods of Orbit Determination for the Microcomputer Willmann-Bell, Richmond, Virginia, 1991

G. Fjermedal New Horizons in Amateur Astronomy Perigee Books, New York, 1989

The orbital elements of a satellite are a series of numbers which mathematically describe its orbit. Using the orbital elements, commonly available personal computer programs can predict where the satellite will be days and even weeks in advance.

[8] King-Hele op cit, Ref 7

[9] Canadian Space Society Computer Bulletin Board, File N2L105.ZIP, excerpted in Table 1.

[10] Desmond King-Hele A Tapestry of Orbits Cambridge University Press, Cambridge, 1992

[11] E. Rork, S. Lin, A. Yakutis Ground-based electrooptical detection of artificial satellites in daylight from reflected sunlight Lincoln Laboratory, Massachusetts Institute of Technology, 1982

'Porog obnaruzheniia ISZ na dnevnom nebe' [Satellite detection threshold in the daytime sky] B. Davydov Kosmicheskie Issledovaniia, Vol. 28, Sept.-Oct. 1990

[12] King Hele, op cit Ref 7

[13] CompuServe Astronomy Forum, Data Library 3, File NAVSPA.THD

[14] T. Sato, H. Kayama, A. Furusawa, I. Kimura 'MU Radar Measurements of Orbital Debris Journal of Spacecraft and Rockets, V28, N6, Nov-Dec, 1991

[15] Nicholas Johnson The Soviet Year in Space, 1988 Teledyne Brown Engineering, Colorado Springs, 1989

[16] Canadian Space Society Bulletin Board, File N2L-156.ZIP

Project Operations Branch NASA Satellite Situation Report, Vol.32, No.4, Dec, 1992 NASA/Goddard Space Flight Center Greenbelt, MD

TRW Space and Technology Group TRW Space Log, Vol.27, 1991 Redondo Beach, CA

Andronov, 'Kosmicheskaya sistema' op cit, Ref 2.

[17] D. Fried 'Probability of getting a lucky short-exposure image through turbulence' Journal of the Optical Society of America, Vol.68, No.12, Dec. 1978

G. Coupinot, J. Hecquet, R. Futaully 'Optical astronomical angular resolution'

Journal of Optics, Vol.21, No.1, Jan.-Feb. 1990

[18] J. B. Rafert (ed) OL-AG Phillips Laboratory Malabar Test Facility User Manual

U.S. Air Force Phillips Laboratory, 1991

[19] T. Stewart McKechnie 'Atmospheric turbulence and the resolution limits of large ground-based

telescopes' Journal of the Optical Society of America, Vol.9, No.11, Nov. 1992

N.A. Massie, T. Lawrence, M. Shao, P. Finch and Y. Oster 'Stalking Satellites in High Resolution'

Lasers and Optronics, June 1990

[20] H. MacAlister, W. Bagnuolo, W. Hartkopf 'Multiple telescope optical interferometric array' Amplitude and intensity spatial interferometry; Proceedings of the Meeting, Tucson, AZ, Feb. 14-16, 1990 Society of Photo-Optical Instrumentation Engineers, Bellingham, 1990

Beckers, J.M. Interferometric Imaging with the Very Large Telescope Journal of Optics, Vol.22, No.2, March-April 1991

[21] Interagency Group(Space) Report on Orbital Debris for the U.S. National Security Council U.S. Government, 1989

[22] G.H. Canavan An Entry-Level Conventional Radar-Driven Rocket Anti-satellite

Los Alamos National Laboratory, Los Alamos, 1991

[23] Johnson op cit, Ref 15;

S.R. Peterson Space Control and the Role of Antisatellite Weapons

Air University Press, Maxwell AFB, 1991

A. Wohlstetter and B. Chow 'Recommended Changes in U.S. Military Space Policies and Programs'

Memorandum for the Commission on Integrated Long-Term Strategy, U.S. Department of Defense, undated (ca. 1988 from internal evidence)

R .B. Giffen US Space System Survivability; Strategic Alternatives for the 1990s National Security Affairs Monograph Series 82-4 National Defense University Press, Washington, D.C., 1982

P.B. Stares Space and National Security The Brookings Institution, Washington, D.C.,

J. M. Collins Military Space Forces; The Next 50 Years Pergamon-Brassey's, McLean, Va, 1989

G. L. Bennett Survivability Considerations in the Design of Space Power Systems Proceedings of the 23rd Intersociety Energy Conversion Engineering Conference, American Society of Mechanical Engineers, New York, 1988

Norman Friedman 'Smart Weapons, Smart Platforms: The New Economics of Defense' in Science, Technology, and Security in the New International Order American Association for the Advancement of Science, Washington, D.C., 1991

[24] Peterson op cit, Ref 23.

[25] Canavan, op cit, Ref 22;

Bennett, op cit, Ref 23

Bruce D. Berkowitz Calculated Risks Simon & Schuster, New York, 1988

[26] Wohlstetter and Chow op cit, Ref 23.

[27] H.H. Gerth and C.W. Mills From Max Weber: Essays in Sociology Oxford University Press, New York, NY, 1946

S. Bok Secrets Vintage Books, New York, 1989

M. Halperin Bureaucratic Politics and Foreign Policy The Brookings Institution, Washington, D.C., 1974

[28] DoD Memorandum for Correspondents, No.264-M, September 18, 1992

Report of the Secretary of Defense to the President and the Congress January, 1993 U.S. GPO, Washington, D.C. 1993, p.92