Summary: The Smithsonian Astrophysical Observatory is using lasers to

increase the accuracy of its satellite tracking. The laser system at the Mt.
Hopkins Observatory, Arizona, U.S.A., measures range to within a meter. It can
operate when the satellite is sunlit and when it is in the earth's shadow. It can
also obtain data when the sky is too bright for cameras to photograph the
satellite against a star background. The corrected range values are available
within a few minutes of the observation. The laser has a power output of 500
MW. Its mount is positioned manually, with an accuracy of 1/2 arcmin, to
predicted values of azimuth and elevation. Visual tracking is not required. The
data rate is 1 per minute. The maximum range of the system is sufficient to
track all the satellites that are currently in orbit and that have the
retroreflectors required for laser tracking.
Introduction:
Satellite orbits with meter or decimeter accuracy begin to open up such new
areas of investigation as crustal motions and relativity effects (LUNDQUIST,
1996).
However, these accuracies cannot be obtained by using cameras such as the
Baker-Nunn for satellite tracking. The reason is that the 2- to 3-arcsec
observational error of the cameras limits the results of the orbit computation to
an accuracy of about 10 meters. The situation is different when laser systems
are used to track satellites. Like radars, they measure the distance to the
satellite these rather than its direction in space; and at present they measure
these distances to within a meter or two. Future accuracies of decimeters are
not impossible.
Laser systems have become practical over the last few years six satellites (two
French and four American) now in orbit have been equipped with
retroreflectors.
Each retroreflector is made up of an array of hundreds of corner cubes. These
corner cubes concentrate the returning laser radiation sufficiently to make
laser ranging feasible. In fact, the equipment needed on the ground and on the
satellite turns out to be relatively simple and inexpensive when compared with
other electronic tracking systems. Of the six satellites (with retroreflectors) that
are now in orbit, only two - Geos 1 and Geos 2 - can be ranged on in the
Southern Hemisphere. These satellites are gravitationally stabilized;
consequently, their retroreflectors point toward the earth wherever they are in
their orbit. The other four satellites are stabilized by the earth's magnetic field.
The retroreflectors on these satellites point away from the earth when they are
over the Southern Hemisphere.
Since lasers operate within the visual spectrum, they, like the satellite-tracking
cameras, requires cloudless skies. They should be located at good astronomical

observing sites. In contrast to the cameras, however, they can be used when
increased accuracy is the main reason for using laser systems for satellites
tracking, there is a secondary advantage: The data can be obtained quickly.
The time of flight of the pulse of laser energy from the laser to the satellite and
back again is presented immediately in digital form by a time-interval counter.
The atmospheric correction is small in magnitude and simple in form. This rapid
availability of the data should lead to improved predictions of a satellite's
angular coordinates and, in turn, to an increase in the laser system's maximum
range and to an improvement in its accuracy.
Retroreflectors on satellite:
Retroreflectors consist of arrays of hundreds of corners cubes that direct the
returning energy from the laser beam toward the laser system on earth. There
are three different arrays in current use: the ones on BE-B and BE-C, on D1-C
and D1-D, and on Geos 1 and Geos 2. Each of them was designed to spread the
returning beam enough to compensate for the angular deflection that is due to
velocity aberration.
This spread was introduced because the beam axis of the returning beam
intersects the earth at distances up to about 100 m from the laser system. The
point of intersection varies from one satellite pass to another. If the beam were
not spread, the photoelectric receiver would have to be separated from the
laser transmitter, and receiver would have to be moved to a new position for
each satellite pass.
Principles of operation:
Laser tracking systems are optical radars. A pulse is transmitted to the
satellite; it is reflected and the return is detected. The data consists of the
epoch at which the laser is fired and the time taken for the pulse to travel to
the satellite and return. A correction for the reduced velocity of light in the
atmospheric is made.
The magnitude of this correction is 2 to 6 m. Since its value can be determined
to within an accuracy of about 1% (LERH, 1968a), the maximum atmospheric
error remaining in the corrected range measurements is about 5 cm. When
time is converted to range, the value of c is, the velocity of light in vacuum, is
used. Since the experimental accuracy of c is, at present, only about 1 part in
10^6, a sizable range error is introduced. However, this error turns out to be
unimportant, because ranges expressed in light-seconds are satisfactory for
almost all applications. For proper correspondence with the measured range,
the true epoch is the instant at which the laser pulse reaches the satellite. It is
obtained by adding half the travel time of the laser pulse to the measured
epoch. Usually this correction is not made until the orbits are computed. For

satellite tracking, optical radar has the following advantages over microwave
radar:
A. The inherently short time duration of the pulse favors high accuracy.
B. The narrow beamwidth improves the efficiency.
C. The low wavelength simplifies the atmospheric correction.

The laser system on Mt. Hopkins:

The laser system on Mt. Hopkins was designed to be a prototype of one that
could be installed at each of the 12 SAO stations. Its features reflect the
particular capabilities and needs of the Smithsonian program of precise orbit
computation. The system has a static pointing mount rather than a tracking
mount. It uses predictions of a satellite's azimuth and elevation instead of
relying on visual tracking. The accuracy of the predictions is several arcmin, a
value that should be well within the maximum 6 mrad (20 arcmin) beamwidths
of the laser and the receiving telescope. The static pointing mount was
selected because it permits the laser system to operate when the satellite is in
the earth's shadow and it may permit operation during daylight hours. Because
the mount is manually operated, the pulse repetition rate is low. But this rate
should be adequate for the computation of precise orbits when the network of
stations is complete. The operational simplicity of the system may be desirable
in some of the network's more remote locations.
The integration of the system is outlined in figure 9.1
The static pointing pedestal

The static pointing pedestal (see Fig 9.4) is a rotatable mount for pointing two
instrument packages, a photoreceiver, and a laser transmitter, to an overall
positional accuracy of better than +- 1/2 arcmin. It employs a T-type elevationover-azimuth configuration and is operated manually.
The azimuth-axle housing is driven by means of a worm drive assembly located
at the top right side. A Veeder-Root counter is used for coarse readout, and a
goniometer for fine readout.
Another hand-driven worm drive assembly is located near the left end of the
elevation-axle housing. The goniometer for dine readout of elevation setting is
mounted on the opposite end of the housing.
As seen when facing the pedestal from the operator's position, the left end of
the elevation axle is flat to accommodate the photoreceiver mounting plate.
This plate provides for a +- 2" elevation adjustment of the photoreceiver. By
means of opposing adjustable screws, which between them accept a lug on the
tube saddle, the worm gear for elevation drive is also on the left end of the axle
within the housing.
The right end of the elevation axle is machined concave to accept a matching
spherical convex portion of the laser mounting plate. This ball joint provides a
+-2" azimuth adjustment of the laser. The right end of the axle also carries the
elevation-setting circle, which is adjustable for aligment with true horizontal.
The laser transmitter
The laser transmitter system (outlined in Fig. 9.5) provides a pulse rate of 1 per
min. The transmitter package contains both an oscillator and an amplifier rubylaser head, a Pockels-cell Q-switch, a Brewster stack polarizer, a rear reflector,
optics for coupling the oscillator and amplifier rods, a set of beam-forming
optics with facilities for boresighting, and an output monitor.
All these units are mounted on an aluminium I beam. The ruby rods and
flashlamps are water cooled, and the cavities are purged with dry nitrogen
during operation of the laser. The entire unit is protected by a dustproof cover.
The oscillator head contains two 6.5-inch xenon flash lamps and a ruby rod 3/8
inch in diameter and 6-5/8 inches long. The lamps are located at the focal
points of an intersecting elliptical cross-section silver-coated cavity; the rod is
located at the common focal point. Both flashlamps and the ruby rod are
contained within individual quart water jackets sealed with "O" rings. The
amplifier cavity is arranged in the same way as the oscillator cavity, but
utilizies a 5/8-inch-diameter, 7-inch-long ruby rod.
A Pockels cell containing a potassium deuterium phosphate crystal is used as
the Q-switch. The Pockels cell is held at high potential to prevent firing of the

laser until the proper time; then the potential is dropped to 0V in about 30 ns. A
giant pulse is formed up to 200 ns later. The beam-forming telescope consists
of a Galilean lens and a 6-inch-diameter double convex lens. Both lenses are
coated for maximum transmission at 694 nm. The Galilean lens is fastened to a
movable platform and can be moved by a micrometer attached to a flexible
cable to vary output-beam divergence from 2 to 20 arcmin.
The output pulse monitor is a ITT biplanar photodiode with an S-20 cathode
driven at -1000VDC. The monitor is arranged to look at laser light reflected
from the rear surface of the 6-inch lens of the beam-forming telescope. The
pulse from this monitor starts the time-interval counter.
An input to the oscillator cavity of 3000 joules and to the amplifier cavity of
6000 joules produces an output of 500 MW in a 15-ns pulse.
The telescope photoreceiver
The telescope photoreceiver, designed primarily to be used with a photomultiplier tube (PMT), detects returning laser pulses. Equipped with an
eyepiece, it permits visual checking of satellite positions, collimation of the
receiver with the transmitter, and alignment of the readout devices on the
pedestal with the receiver optical axis.
The receiver employs a coaxially folded, primefocus optical system and has
separate optical subsystems for the eyepiece and the PMT. The primary
subsystem is a 20.5-inch parabolic mirror with a 6-inch hole in the center. The
other is a 5-inch optical flat supported through a cored hole in the center of the
mirror.
The eyepiece optical system (Fig. 9.6) operates, in essence, as any standard
viewing telescope, with the eyepiece focused on the prime-focus plane. The
only departure from a conventional system is the addition of a pair of relay
lenses that transfer the image from the prime-focus plane to the side of the
telescope tube and the eyepiece. A flip mirror deflects the light through the
relay lenses to the eyepiece.

The optical system for the PMT (Fig. 9.6) gives an image of the primary mirror
at the face of the RCA 7265 PMT. Thus, instead of the PMT seeing a point image
of the returning laser light, the detection of which might depend on where the
image fell on the photocathode, the PMT looks at a donut-shaped image of the
uniformly illuminated primary mirror. In this way, the laser return is evenly
distributed across the PMT cathode.
In order to provide the image of the primary mirror at the PMT face, a positive
meniscus field lens with a 2-inch focal length is placed in the primary focal
plane. The field lens forms an image of the primary mirror at a point slightly
more than 2 inches behind the field lens. This image is then transferred
through the hole in the primary mirror to the PMT face by a pair of relay lenses,
a collimating lens, and an imaging lends. The collimating lens also collimating
the light bundle so that the angle of incident of the rays passing through the
interference filter, set just in front of the imaging lens, does not exceed 1".

The field stops are located at the filed lens positions. The stops are holed disks
that permit a field range from 2 to 20 arcmin in 2-arcmin increments.
An 0.6 nm passband interference filter is mounted into a tiltable holder inside
the tube between the collimating lens and the imaging lens. The filter can be
tuned by a micrometer (located at the back of the telescope), which adjust the
position of the tiltable holder.
The optical components in the system are coated to provide optimum
performance in the spectral region in which they are used. The lenses in the
eyepiece optical system are standard AR coated to provide good response in
the visual region. The lenses in the PMT optical system have a dielectric "V"
coating to allow less than 0.3% reflectance at each air-to-glass surface at 694
nm, the ruby-laser wavelenght. The primary, secondary, flip, and diagonal
mirrors are surfaced with aluminium overcoated with silicon monoxide. The
efficiency of the PMT optical system, including the primary and secondary
mirror, is about 43%.