Lunar Laser Ranging at McDonald Observatory:

Beginning the Second Quarter Century

P. J. Shelus, R. L. Ricklefs, J. G. Ries, A. L. Whipple, and J. R. Wiant

McDonald Observatory

University of Texas at Austin

Austin, Texas 78712-1083

X-Y Offset-Guiding Stage

The initial LLR up-grade at the MLRS was the installation and integration of an x- y offset guiding stage. As used for LLR operations at the MLRS, this offset guiding stage allows an observer to guide on a sun-lit, off-axis lunar surface feature, or perhaps a star, while the retroreflector is on the shadow side of the lunar terminator. Not only does this provide for a greater number of observing opportunities during a lunation, ranging to a retroreflector in the dark produces virtually noise free data. The MLRS instrument is a two-axis translation stage that provides for the simultaneous mounting of two electronic cameras, with turning optics to direct the telescope's Cassegrain beam to either of the cameras. The cameras are selectable under computer control or manually. A DC servo motor/encoder combination drives each axis of the stage in a range 0-8 mm/sec and each axis is directly encoded using digital linear encoders. The positioning of each stage is accurate and repeatable to 5 microns and each has a travel of more than 125 mm along each axis, centered on the optical axis. The electronics control package includes a dedicated PC-type computer, a motor controller board inside the PC, plus electronic and computer interfaces and controls. Software provides a closed servo loop between the encoders and the motors and communicates with the external control computer through an ordinary serial port. Routine operation began in the spring of 1993.

Optical Enhancements

The MLRS laser pulse contains 3 x 1017 photons. In lunar mode only a few return to pass through the receive system. A typical lunar return rate is a few signal photons per minute. We use single photo-electron detection devices and as large an optical throughput in the green as possible. With the large number of noise sources, it is impossible to identify a returning photon from the moon without filtering, both physical and mathematical. Range gating provides a temporal filter. A pin-hole aperture provides a spatial filter. A spectral filter allows only the proper color photons to pass through to the receive system. The spectral filter must be as narrow as possible in wavelength to eliminate as many noise photons as possible, transmitting as many proper photons as possible. These requirements change with lunar phase and sky conditions. We specified and purchased additional spectral filters, of intermediate specifications between the filters already being used. These were received in the spring of 1993 and immediately placed it into service. In August 1994, because of aging and laser induced damage over the many years of observing operations, the telescope's #3 mirror, a dichroic, was replaced. The added scheduling flexibility and the extra energy throughput that the new filters and #3 mirror provided, during changing lunar phases and local sky conditions, resulted in a significant increase in the amount of LLR data observed at the MLRS.

Auto-Guiding and Image Enhancement System

The longer the LLR transmitter/receiver system remains on target during laser firing, the more returns will be detected. Our experience with the MLRS is that present pointing and tracking stability requires intensive manual guiding to keep the telescope on target. Providing the operator with guiding tools or even removing the operator completely from manually guiding the telescope will dramatically increase time on target. In the spring of 1993, we specified and contracted for purchase what is now called the MLRS Auto-Guiding and Imaging System (AGIS). AGIS is an integrated hardware and software system that accepts real-time, video signals as input, i.e., a highly magnified image of a small portion of the lunar surface, or a stellar or artificial satellite point-source image. It performs real-time image processing allowing the user to select among various levels and types of image enhancement. Further, it produces tracking error signals, that under user selection, are communicated to the control computer for guiding control. We issued the AGIS construction contract in March 1994. Delivery was made in February 1995. Installation and integration are now underway. We expect that there will be several months of effort from MLRS hardware and software personnel before the AGIS can be totally implemented and integrated into the routine observing operation. We expect a marked increase in the amount of MLRS LLR data.

Avalanche Photodiode Detector

Both the French station at Grasse and the German station at Wettzell have the capability of laser ranging to the Moon. The Wettzell group has been involved in the design and construction of a low light level, avalanche photo-diode (APD) detector. Such an APD detector, already in use at the French CERGA LLR station, has exhibited a significant increase in sensitivity as well as improved accuracy and precision of their lunar data. Another type of avalanche photo-diode device is being used for artificial satellite laser ranging operations in the United Kingdom. In the spring of 1993, we entered into a cooperative effort with our German colleagues to build such an avalanche photo-diode detector for use at the MLRS for LLR operations. Jerry R. Wiant, chief engineer of laser ranging operations at the MLRS, traveled to Wettzell, Germany in late November 1993, and during a several week stay at the Wettzell laser ranging station, he became familiar with the intricacies of the APD detector and participated in the design of a unit that could successfully interface with the MLRS's unique hardware and software systems. He brought a completed unit back to McDonald Observatory for integration within the MLRS lunar operation. Final implementation at the MLRS is being currently being performed. Some technological problems have been experienced in properly cooling our unit. However, we expect to begin using this unit later this year.

Disciplined Oscillator

The present MLRS station clock is a commercial Hewlett Packard cesium-beam standard. These timing units are extremely expensive and must undergo considerable preventive maintenance on a time-critical basis. With the presence of a Global Positioning System (GPS) timing receiver on site at the MLRS, we are planning a move to a more cost-effective manner, with much lower maintenance responsibility, to obtain more precise station timing. A so-called disciplined (or steered) oscillator can provide excellent short-term timing stability, while a suitable GPS receiver is used as a fly-wheel to provide good long-term timing stability. A significant cost and time savings in station operation and maintenance could result, with a reasonable improvement in timing accuracy and precision. Several units have been identified for further study.

MLRS Control Computer

We are now in the final stages of replacing the MLRS's 15 year-old Data General NOVA control computer system with a LynxOS based, X-windows, real-time UNIX system running on PC hardware. Completing a more than three-year effort, this is being coordinated with similar upgrades at other NASA-based laser ranging systems. This effort has resulted in the design and implementation of a system with compatible hardware and software architecture in an open-systems (POSIX) environment. This approach allows a maximum amount of software portability and sharing. Data from more than 75 artificial satellite passes has already been acquired with the new system in shake-down mode. Lunar data was taken with the system in March. We expect complete conversion to the new system within the next month.


To fully understand the advances accomplished under the MLRS up-grade, one must recognize how the MLRS now compares with the French LLR effort, the flagship of the lunar network. The French lunar station is essentially a dedicated lunar facility with little or no artificial satellite laser (SLR) ranging commitments. On the other hand, the MLRS is one of the premiere SLR stations in the world. The MLRS has only 25% of the collecting area and only 50% of the laser power of the CERGA station. Several ways of comparing the MLRS to the French LLR station appear below.

MLRS/CERGA Comparison (1992-April 1995)
Lunar Laser Ranging Data Throughput

	1992	1993	1994	Apr 1995	1993	1994	Apr 1995	1993	1994	Apr 1995
LLR	Apollo 11	-	3	25	11	53	55	26	6%	45%	42%
Normal Points	Apollo 14	-	8	17	10	53	44	27	15%	39%	37%
	Apollo 15	58	151	160	118	433	499	213	35%	32%	55%
	Lunakhod 2	-	1	3	2	12	17	4	8%	18%	50%

	Total	58	163	205	141	551	615	270	30%	33%	52%
Minutes	Apollo 11	-	28	382	141	561	484	240	5%	79%	59%
of	Apollo 14	-	141	242	151	485	383	239	29%	63%	63%
LLR Data	Apollo 15	757	1,953	2,122	1,560	4,156	4,603	2,038	47%	46%	77%
	Lunakhod 2	-	3	40	16	105	142	33	3%	28%	48%

	Total	757	2,125	2,776	1,868	5,307	5,612	2,550	40%	49%	73%
Nights when	Apollo 11	-	2	16	8	27	23	12	7%	70%	67%
LLR Data was	Apollo 14	-	5	12	7	26	17	14	19%	71%	50%
Taken	Apollo 15	23	56	61	34	75	71	33	75%	86%	103%
	Lunakhod 2	-	1	3	1	11	9	4	9%	33%	25%
UT0 points		5	26	30	21	49	44	29	53%	70%	72%

The up-graded MLRS appears excellent in its own right and compares very well with the French station. While being one of the most prolific producers of SLR data in the world, the MLRS, beginning in 1993, also is producing more than 1/3 as much lunar data as the lunar dedicated CERGA station. Also, the MLRS gathered lunar data on up to almost 90% as many nights as CERGA. The multi-faceted MLRS has accomplished a great deal with the LLR upgrade, while still maintaining all of its regular SLR observing responsibilities. And still, two vital up- grades remain for final MLRS implementation, i.e., a more sensitive and noise free photon detection device and a sophisticated lunar surface automatic guiding system.

This figure gives another view of the effects of the MLRS’s recent improvements. The total McDonald LLR data set comprises observations spanning more than two-and-a-half decades. For the first 16 years, the 2.7-m system was used; since then, the MLRS, with several months of overlap in 1985. Further, the data has been obtained successively at three different locations: 1) the 2.7-m system on Mt. Locke; 2) the MLRS in the saddle between Mt. Locke and Mt. Fowlkes; and 3) the MLRS on Mt. Fowlkes. Summarized, on an annual basis, is the number of McDonald normal points from all retroreflectors. Also included is the annual weighted root-mean-square of the post-fit residuals from our own analyses. To assure a large enough sample, to be statistically significant, these rms’s are calculated from Apollo 15 observations only. The 2.7-m system produced more than 200 normal points/year, with a weighted rms of 9-15 cm. When the MLRS first came on line, scheduling logistics caused the number of 2.7-m system normal points to drop. When the MLRS attained full lunar capability and the 2.7-m lunar system was de-commissioned, the McDonald LLR data throughput dropped to less than 100 normal points per year. However, the weighted rms of that data also dropped to something like 3-5 cm.

Two different ranging systems and two different observing philosophies have led to very different quantities and qualities of McDonald LLR data. The 2.7-m system was a lunar-only station built around a Korad 3 joule/pulse, 3 nanosecond pulse-length ruby laser, firing at 1/3 hertz. The epoch timing system resolution was approximately 125 picoseconds. The MLRS is a joint lunar/artificial satellite station built around a Quantel 120 millijoule/pulse, 200 picosecond pulse-length Neodymium-YAG laser, firing at 10 hertz. Its epoch timing system precision is about 25 picoseconds. Further, the 2.7-m system had a much stronger lunar return signal with its 2.7-m receive aperture, as opposed to the MLRS’s 0.76-m receive aperture. Moving laser operations from the 2.7-m system to the MLRS should have produced a reduction in the volume of lunar data. The trade-off, of course, that made the transition acceptable, lay in the accuracy of the MLRS data. Laser ranging data accuracy, to first order and with everything else held constant, scales inversely with laser pulse length. The much shorter pulse length of the MLRS’s laser led to almost a factor four improvement in the accuracy of the MLRS ranges. Since the total weight of a set of observations scales linearly with data accuracy, but only as the inverse square of the number of observations, the MLRS data is much stronger than the 2.7-m system, in spite of its lower volume. The significant point is that now, with the recent MLRS LLR-related up-grades, not only has the accuracy and precision of the LLR observations been mightily improved over the 2.7-m system, the LLR data volume of the MLRS is now approaching that produced with the 2.7-m system.

Lunar laser ranging (LLR) has turned the Earth-Moon system into a laboratory for a broad range of scientific investigations. The unique contributions already accomplished include a three orders-of-magnitude improvement in the accuracy of the lunar ephemeris, a several orders-of-magnitude improvement in the measurement of the variations in the Moon’s rotation, and the verification of the principal of equivalence for massive bodies with unprecedented accuracy. Complementing and supplementing other observational disciplines, the LLR analysis has also provided measurements of the Earth’s orbit precession, the 18.6 year nutation, the Moon’s tidal acceleration, lunar rotational energy dissipation and the Moon’s free librations, Earth orientation, and the determination of the obliquity and the equinox. LLR also contributes to the determination of station locations and motions, the Earth-Moon ratio, lunar gravity field harmonics and Love numbers, and retro-reflector coordinates. An itemization of the results obtained appear in the table.

For future analysis tasks, it is natural to divide the LLR-derived parameter solutions into short-term and long-term effects. Because of the tremendous improvement in observational accuracy brought about by the transition from the 2.7- m system to the MLRS, it is understandable that parameters with short characteristic signatures already have good accuracy since they are determined from the best data. Parameters having signatures with longer time scales still must depend on the 15 cm data accuracy of the 1970’s and early 1980’s. The need for more and better data is twofold. A practical reason is that, when extrapolated outside of the time span of the observations, the lunar orbit and librations (both forced and free) will suffer run- off. More vital is the need for a longer observational time base for investigating long term effects. Tidal acceleration goes as the square of the time and diurnal and semidiurnal tides split from an 18.6 yr term. While the lunar theory arguments l, L, and F enter analytical equations with monthly periods, L and F take 18.6 yr to separate; L and l take 9 years. The lunar libration in longitude, with a period of only 2.9 yr, has a 69 yr beat with another term of similar period; a third mode of that libration has a 24 yr period with respect to the node; the dissipation term and the lunar obliquity have an 18.6 yr period. The precession of the Earth’s axis is secular and its principle nutation term has an 18.6 yr. period. The “Chandler wobble” mode has a period of 74 yr. Several of the general relativistic effects, including G-dot, give rise to secular effects in t and t-squared.

Finally, LLR is sensitive to the relative orientation of the equator, ecliptic and lunar orbit planes, presently at the 1.5 mas level, and improving with time, due to our being on a spinning Earth while the lunar orbit precesses along the ecliptic. Thus, an LLR solution implicitly includes the obliquity and dynamical equinox. The Earth’s orbit plane and the equatorial plane are linked together to form a dynamical reference frame.

Lunar Laser Ranging (to must be continued)

In summary, the science that LLR addresses is multi-disciplinary and includes, among others, astronomy, celestial mechanics, gravitational theory, relativity, and Earth physics. Even after more than 25 years of continuous activity, it remains a substantial technological challenge. However, with care, attention, and dedicated effort, as evidenced by past and present operations at the Observatoire de Cote d'Azur in France, the LURE Observatory on the island of Maui, the German station at Wettzell, and the McDonald Observatory in Texas, it has been shown that LLR observations can be regularly and routinely obtained. In spite of hard times, the LLR community continues to "push the envelope" to improve the precision, accuracy, and volume of this important data type. CERGA, already a very high volume station for LLR data, is striving with new detector, multi-color, and timing technologies to reach millimeter precision and accuracy levels. MLRS, with a significant number of hardware and software upgrades, has increased its LLR data volume by a factor of 3-5 in the past 24 months. The MLRS is also beginning to use Avalanche Photo Diode detectors and modern auto-guiding technologies to not only further increase data volume, but to increase accuracy and precision levels as well. Efforts at Wettzell in Germany and Orroral Valley in Australia continue to push for LLR as a viable observational technique at those stations. A new laser ranging station with LLR capability is currently being constructed to replace the present Matera artificial satellite only station in Italy. Other groups around the world are being encouraged to investigate whether LLR can be established at their own laser ranging stations. Consultation and encouragement, both in hardware and software areas, can be provided, when and where requested.

It must be clearly understood that the MLRS and the CERGA stations constitute only a bare minimum network for the best use of the LLR data type. More stations are needed. As a minimum, additional stations will improve observational coverage, will tend to eliminate correlated weather data drop-outs, and will allow the easier determination of systematic errors. There are also many important scientifically related reasons for more and better multiple station LLR data.

It is noteworthy that we have recently celebrated the 25th anniversary of the first placement of a retroreflector on the Moon. LLR is the only active Apollo experiment that is still obtaining new data. And, it is still marching at the forefront of science. During these times of tight budgets, it is important to point to examples of efficient and cost-effective research. LLR, and the science it is able to accomplish, should be a source of pride to the scientific community in general. At McDonald Observatory, we intend that the MLRS will continue to improve to provide a constant stream of more and better LLR data into that scientific community.

Selected references . . .

"Lunar Laser Ranging: Technology and Scientific Results" (with J. O. Dickey et al), Invited Review Article, Science, Vol. 265, pp. 482-490, 1994.

"Lunar Laser Ranging at McDonald Observatory: 1969 to the present" (with R. L. Ricklefs et al), AGU Geodynamics Series, Vol. 25, pp. 183-187, 1993.

"UT0 Determination from Lunar Laser Ranging Observations for 1969-1992" (with A. L. Whipple et al) In NEOS, Ann. Rpt. for 1992, pp. 41-42, USNO/NOAA, 1993.

"A Computer-Controlled x-y Offset Guiding Stage for the MLRS" (with A. L. Whipple et al), Proc. of the 8th International Workshop on Laser Ranging Instrumentation, May 18-22, 1992, Annapolis, MD, ed. J. J. Degnan, NASA Conf. Publ. 3214, 1993.

"To the Moon ... and Back", Discovery (Research and Scholarship), Vol. 10, p. 33-37, 1987.

"MLRS: A Lunar/Artificial Satellite Laser Ranging Facility at McDonald Observatory", IEEE Trans. Geosci. and Remote Sensing, Vol. GE-23, No. 4, p. 357-359 (1985).

Report for October 1993 McDonald Laser Ranging Operations (NAS5-30942) McDonald Laser Ranging Operations (NAS5-29404) Report for January 1987 Report for October 1993 McDonald Laser Ranging Operations (NAS5-30942) McDonald Laser Ranging Operations (NAS5-29404) Report for January 1987