Document II-3

Document title: J.E. Lipp, R.M. Salter, Jr., and R.S. Wehner, et.al., "The Utility of a Satellite Vehicle for Reconnaissance," The Rand Corporation, R-217. April 1951, pp. ix, 1-21, 28-39.

Source: National Security Archive, Washington, D.C.

After Rand had recommended advanced study into the uses of satellites for strategic reconnaissance in November 1950, the Air Force authorized Rand to undertake further research. The results of a Rand study on "The Utility of a Satellite Vehicle for Reconnaissance" were presented to the Air Force in April 1951. They demonstrated the viability of the concept and recommended further research. This recommendation eventually led to the much larger "Project Feed Back" [II-7] in 1954. These excerpts from the over-135-page report contain a general discussion in terms of orbits and instruments of the feasibility of satellite reconnaissance.

Utility of an earth-circling space vehicle as a reconnaissance device is considered here in detail. A satellite (initially placed on its orbit by rocket power) which televises ground scenes and weather information to surface receiving stations is investigated. Particular attention is given to the television, communication, and electrical-power-supply problems, since these are the major determining factors in payload utility of a reconnaissance satellite. Some important corollary aspects namely attitude control and equipment reliability, are also discussed.

In order to round out the study, performance and weight estimates of the rocket vehicle required to carry a television payload are included.

The general conclusion of the report is that television satellites are feasible and that they would be useful if built and operated. Various essential lines of research in television, auxiliary power, and reliability are indicated....

The basic feasibility of satellites from the point of view of rocket performance was considered in a previous group of RAND reports, Refs. 3 through 14. That investigation pointed to several important conclusions. First, the engineering of a rocket vehicle of adequate performance for use as a satellite would require but minor development beyond the then-existing technology. Secondly, the payload would have to be small (not more than 2000 lb) to keep the gross weight within reason; hence destructive payloads are not likely to be economically worthwhile for many years to come. Thirdly, returning the vehicle to earth intact would be difficult and should not be attempted in the early versions.

The above factors indicated that the payload would be restricted to instrumentation and communication equipment and prompted the RDB (Technical Evaluation Group) and the Air Force to request that further attention be given to the question of utility. RAND's effort since 1947 on the satellite study has been closely tied to the payload-its description and military usefulness. Most attention has been directed toward reconnaissance, since that is a field in which a satellite may very well show advantages over other types of vehicles.

It now appears fortunate that reconnaissance was selected for the first payload investigation. As will be seen later in the report, pioneer reconnaissance (general location and determination of appropriate targets) and weather reconnaissance are suitable with the resolving power presently available to a satellite television system. These two classes of reconnaissance have also been growing in importance to the Air Force because of the

vastness of Russia and the difficulty of gaining information by conventional means.

To explore further the possibility of reconnaissance by means of a satellite, it is necessary to investigate the various constraints imposed in conducting such an observation from a remote, unattended vehicle.

The first step in such an analysis logically considers the movement of the satellite as a vehicle with respect to the targets to be viewed. Consideration must be given to the degrees of freedom at our disposal in the type and position of orbits and to the frequency of the satellite within the orbit. This approach, from a macroscopic standpoint, gives rise to information on how often and under what conditions the satellite can be placed over a given target. This is discussed in Section I, "Satellite Orbits and Ground Coverage."

Naturally following this step is the microscopic inquiry into the feasibility of viewing a target from the satellite. Television has been selected as the only practical way known at present for transmitting back to earth that which can be seen from the vehicle. Thus an evaluation of television-camera-equipment capabilities, along with a discussion of associated problems of transmission of the picture, is presented in Section II, “Reconnaissance by Television."

[2] Moreover, since there is an intimate interdependence between the type of reconnaissance desirable and the most fruitful way of obtaining such reconnaissance, some simultaneous consideration should be given to the presentation of satellite position (orbits), television scanning, and picture quality. This may be found at the conclusion of Section II.

The remaining problems can be classed as attendant ones peculiar to obtaining remote television broadcasts from the satellite. In order to scan the surface of the earth with the television camera, the vehicle must be properly oriented (attitudewise) with respect to the earth's surface. This is covered in Section III, "Orbital Attitude Measurement and Control."

The television and attitude control equipments require electrical energy that must be supplied by an auxiliary powerplant. A discussion of this powerplant, as well as the estimated power and weight requirements it must meet, is given in Section IV, “Auxiliary Powerplant."

Section V, "Reliability of the Satellite," includes an analysis of the anticipated reliability of the television and auxiliary equipment. This is a particularly important problem, since the equipment must operate automatically for a long period in an inaccessible location.

Finally, the characteristics of the vehicle itself necessary to place the television payload in a given orbit around the earth are presented in Section VI.

The several appendixes furnish correlative data and extensions of the remarks concerning some of the more salient features resulting from the study of the technical feasibility of utility of satellites for reconnaissance.

[3]

I. Satellite Orbits and Ground Coverage

This section presents a general discussion of the pertinent facts about orbits which are essential to the utility of a satellite as a reconnaissance vehicle and of the problems concerning the establishment of a rocket-vehicle satellite on an approximately circular oblique orbit relative to the earth. Since the primary utility aspect considered is reconnaissance, the effect of orbits on scanning (i.e., viewing) angles, as well as some discussion of the limitations imposed by optical and radio transmission requirements, are included.

Orbits Generally

A satellite is defined as an attendant body revolving about a larger one; a moon and a man-made object revolving about the earth are thus satellites. The earth itself is a satellite of the sun. The shape of a satellite orbit, which can be either circular or elliptical, is dependent principally on the initial conditions of velocity, position, and direction of mo

A circular orbit is of course the most desirable for an artificial satellite. Any marked deviations or eccentricity would cause some portion of the flight path to pass through more dense atmosphere and thus decrease the endurance of the satellite (for the likely range of orbital altitudes).

In order to remain on an orbit, the velocity of a satellite must be such that its centrifugal force is sufficient to overcome the earth's gravitational forces upon the satellite at the orbital altitude. Initial trajectory control is required to be such that the velocity is at least that necessary for a circular orbit at the design altitude, and the path angle is within 1/2°. These limits are attainable with present control equipment.

298

RAND's previous studies were devoted primarily to equatorial orbits, which are still of prime interest for preliminary, experimental satellite flights. However, it is obvious that a reconnaissance satellite must be placed on an oblique orbit to view targets of military interest most efficiently.

[4] Review of Orbital Features

Figure 1 illustrates the orbit of a satellite placed on a circular path. Such a path, if unperturbed, would maintain a fixed orientation in space (in this case, as in all others to be discussed here, the centers of the satellite's path reference frame coincide with that of the earth's). Thus the satellite reference frame would move around the sun with the earth but would not be affected by the earth's own rotation. Further, the position of the satellite orbit relative to the sunny side of the earth would change with the earth's seasons.300 Figure 2 depicts this relative change of orbital position for a hypothetical satellite whose orbit is undisturbed by external influences.

Orbital Regression and Resultant Periods

As pointed out in Ref. 14, the orbit is affected by the presence of other astronomical bodies, such as the sun and the moon, and by the shape of the earth. The effect of the sun and the moon on a satellite orbit is nearly identical with their effect upon free-water surfaces of the earth (tides) and results in approximately a 3-ft orbital variation.

The oblate shape of the earth, however, exerts a much larger influence on the satellite rocket. The earth's polar diameter is about 25 mi less than its equatorial diameter. Although a polar orbit will have a vertical variation of approximately 1 mi (an orbit around the equator will have a negligible variation), this effect of the earth's shape is not of direct concern. The interesting and important effect is a corollary of the polar [5] perturbation, namely, a significant regression of the nodes when the satellite orbit is oblique. This regression is similar to the precession of a gyroscope caused by externally applied torques.

801

Further, the regression period of the satellite orbit will vary, depending on the orbital altitude and obliquity. After the method of Ref. 14, the orbital regression periods relative to the sunny side of the earth and to celestial space are plotted as functions of altitude and orbital angle in Fig. 3. For useful reconnaissance orbits, 45° to 60° obliquity and 350 to 500 mi altitude, the change in period relative to the earth is not great.

298. Velocities less than that required for a circular orbit obviously prevent the vehicle from establishing the prescribed orbit; hence the satellite will either fall to earth or assume an elliptical orbit which will cause marked altitude variations. Velocities greater than required yield less disastrous, but also undesirable, elliptical paths. 299. In this report an orbit will be designated by the degrees of an angle between it and the equator. An alternative but equivalent description is the maximum latitude to which the orbit is tangent. Thus a 0° orbit is equatorial, a 90° orbit is polar, and a 56° orbit is 56° oblique to the equator and tangent at 56° latitude.

300. An exception here is an orbit around the equator where this seasonal change is irrelevant. 301. Regression of the nodes may be visualized as a westerly rotation of the line of intersection (nodal line) between the satellite's orbital plane and the earth's equatorial plane (see Figs. 1, 4, and 5).

302. Regression period, as used here, is the time required for the intersection line (see footnote above) to make one complete revolution relative either to the sunny side of the earth or to celestial space, as applicable (see Fig. 4).

For illustrative purposes only, a 56° oblique orbit, approximately the latitude of Moscow, will be studied for most of the balance of this discussion. Figure 4 depicts the nodal regression for a vehicle on such a path. It may be seen in this illustration that the position of the orbit relative to the sunny side of the earth changes not with the earth's seasons, but much more rapidly; for this particular orbit, the period relative to the earth is 70 days rather than a year. Under these conditions, the satellite can see a given target in the daytime only during alternate 35-day intervals regardless of whether the satellite circles the earth once a day or a thousand times; Fig. 5 amplifies this point. Thus a single satellite cannot [6] give a continuous record of daytime viewing of a particular target, but only during alternate 35-day periods. If continuous chronological daytime coverage is desired for longer periods, a minimum of two vehicles would be required. Further, if contrast requirements exclude twilight intervals, then three satellites operating on 8-hr shifts, with paths as shown in Fig. 6, are necessary.

Altitude, Velocity, and Duration

So far, discussion has been centered on the path of the satellite in its orbit. Its speed and altitude will now be considered. Figure 7 gives a plot of the required satellite velocity as a function of altitude above the earth for a nearly circular orbit. Since this velocity is independent of the earth's rotation, a satellite launched eastward gains by the component of the earth's peripheral speed in that direction. Figure 7 also shows the number of satellite revolutions per day as affected by orbital altitude.

The duration of an orbiting vehicle depends on the amount of atmosphere tending to slow it down. This in turn means that the higher the altitude, the longer the satellite [7] can stay up.

303

Figure 8, taken from Ref. 3, gives anticipated duration as a function of altitude. At a 100-mi altitude the vehicle will be pulled to earth in less than one revolution because of the atmospheric drag. At 350 mi the duration is about 2 years. At 500 mi the satellite will stay up around 50 years; at 600 mi, several centuries. From this standpoint alone, it is desirable to use as high an altitude as possible. Also, the range of line-of-sight radio transmission increases with altitude. Counterbalancing these factors is the greater size of the satellite required to put a given payload on an orbit at higher altitudes (e.g., 10 to 20 per cent higher gross weight is required to increase altitude from 350 to 500 mi; see Fig. 40, page 77). Another deterrent factor is the increased size and weight of camera equipment necessary to scan the earth from higher altitudes, which requires higher resolving power for an equivalent picture. Therefore, the desirable altitude will represent a compromise between these opposing features but will probably lie between 350 and 500 mi. For purposes of consistency, a 350-mi altitude will be used in the remainder of this report, except where altitude is considered as a variable.

Effect of Orbital Altitude on Ground Coverage and Related Problems

At orbital altitudes of 350 to 500 mi, the satellite circles the earth fifteen to fourteen times a day (see Fig. 7). The satellite tracks cross the equator at intervals of 24° to [8] 25° longitude or, roughly, there are 1700 mi (measured east-west at the equator) between tracks for the 350-mi altitude.

At 56° latitude, for example, this interval is about 800 mi; near the tangent latitude the tracks recross each other several times. Figure 9 indicates the tracks for a satellite at an orbital altitude of 350 mi and at an orbital angle of 56°. Also shown is the average daytime coverage during the daylight "season" with a 400-mi optical scan to either side of the satel

303. Only line-of-sight transmission can be used because high-frequency waves are necessary for television equipment. Also, long radio wavelengths will be adversely affected by the ionosphere; for instance, reflection by the Heaviside layer will prevent such wavelengths from reaching the earth rather than to increase their range.

lite (800-mi optical-scanning band); the light-green area shows targets covered once a day; medium green, those covered twice; and dark green, those covered three or more times. White areas (below the tangent latitude) are those viewed less than once a day; as indicated in the figure, for the assumed satellite orbit, coverage in any one day is not complete below 30° N. latitude.

The 800-mi optical-scanning band at 350 mi altitude represents approximately a 94° included scanning angle, i.e., a 47° scan to either side of the vertical. The included angle of the horizon is 135°, but the value of pictures taken beyond 45° on either side of vertical is questionable. This point is shown schematically in Fig. 10, which also gives a plot of horizon angle as a function of altitude. A discussion of the effects of scanning angle, as well as those of the orbital inclination, upon the minimum resolvable surface dimension is presented in Appendix I.

Proper initial selection of the orbital altitude would enable the satellite to make an integral number of revolutions for one revolution of the earth relative to the orbital plane (not necessarily per 24-hr day, since the orbital plane regresses). Integral [9] numbers of satellite revolutions every other (24.3-hr) day, every third day, etc., are also possible. Such orbital conditions, however, cannot be made accurately enough with present control equipment to afford the same trace on the earth's surface day after day. Thus a drift can be expected so that the satellite will come within a few miles of its track on the previous day (or the previous alternate day, etc.). The significant fact is that by adjusting an orbital period so that it is nearly integral on alternate days, one can obtain, the following day, a picture in the center of the camera scan of a target which was on the periphery the day before (see Fig. 9), except, of course, near the tangent latitude, in which region still amounts of overlap are obtained.

greater

As mentioned earlier, one factor indicating the desirability of a 500-mi altitude is the need to receive the satellite's television broadcasts by stations sited either in friendly territories or on ships. Figure 11 shows the area of reconnaissance interest which would be covered by transmission ranges of 1396 and 1743 mi with 5 stations and 2000 mi [10] with 4 stations (see Fig. 22, page 30, for range as a function of altitude and elevation angle). Transmission must be "line-of-sight" because of the required radiation frequencies. It is estimated that the maximum range for acceptable transmission 305 from a 350-mi altitude is about 1400 mi. At this range, 5 stations would be required to pick up Asiatic observations, but about 15 per cent of the USSR, a significant portion near 105°E longitude, would be left out. Increasing the satellite's altitude to 500 mi affords (on the same basis) a range of approximately 1750 mi. With this range and the same 5 stations, the unobserved area is reduced to a small amount.

With a 2000-mi range (not shown), the unobserved area would be eliminated. However, by accepting a small unobserved area near 95°E longitude, 4 sea-borne stations could be employed. At this latter range, the orbital altitude required for equivalent clarity of the transmission exceeds 600 mi (see Fig. 22, page 30); it may be possible to attain a 2000-mi range from a 500-mi altitude, although some uncertainty and signal distortion would occur in the 100- to 250-mi extremity.

The possibility of eliminating so-called unobservable areas by using delayed broadcasting becomes apparent. It is well to note, however, that the number of frames to be filed would cause the transmitting device to be so bulky and complex that this method does not appear to warrant further investigation at the present time.

[11] The effect of different altitudes upon target viewing, as well as upon television camera resolution and contrast, is discussed further in the next section.

304. The period for a 350-mi altitude, 56° orbit is 24.31 hr, which is termed a day throughout the remainder of this discussion.

305. It is assumed that a minimum elevation angle (above the horizon) of 5° be employed for completely acceptable signal reception.