Viking & The First Seismometers on Mars

Mention NASA’s Viking mission to Mars in 1976 and its first-of-its-kind in situ search for life on another planet immediately comes to mind. While astrobiological experiments certainly dominated the science objectives of the Viking landers, other investigations were also performed in NASA’s first Mars landing mission such as meteorology and determining the composition of the soil. Almost completely forgotten today is another first-of-its-kind investigation: seismology on the surface of Mars.

Diagram showing the placement of instruments on NASA’s Viking lander. Click on image to enlarge. (NASA/JPL)

The Viking Seismometers

The Principle Investigator for Viking’s seismology investigation since its inception in 1969 was American geophysicist, Don L. Anderson (1933-2014), of the California Institute of Technology – the parent organization of the Jet Propulsion Laboratory (JPL) which was responsible for Viking mission operations. While a highly sensitive seismometer placed directly on the Martian surface of the sort flown to the Moon during the Apollo program would have been ideal for detecting “Marsquakes” and begin the investigation of Mars’ internal structure, restrictions on instrument mass, power and data downlink bandwidth (not to mention budget!) made this impractical. Instead, Viking’s seismometers were mounted on top of the lander’s equipment bay adjacent to Leg #1 where, unfortunately, it would be affected by noise caused by the wind and activity by the lander’s equipment. Since this decreased the sensitivity of any instrument by three orders of magnitude, the Viking seismometers would only be able to detect comparatively strong events and limit the seismology investigation to performing an initial reconnaissance of Martian seismic activity. The objectives of the investigation were 1) characterize the seismic noise environment at the landing sites, 2) detect local events in the vicinity of the lander and 3) detect large seismic events at greater distances from the lander.

An explode view of the Viking landers’ seismometer assembly. Click on image to enlarge. (NASA/JPL)

The seismometer assembly included on each of the pair of Viking landers was enclosed in a 12 by 15 by 12 centimeter box. The instrument had a mass of 2.2 kilograms and a nominal power consumption of 3.5 watts. At the heart of the instrument was a trio of seismic sensors mounted at right angles to each other (two horizontally and one vertically) with each occupying a volume of 7.5 by 3.8 by 3.8 centimeters. The detectors were sensitive to ground movements in all three dimensions in the 0.1 to 10 hertz frequency range and were capable of resolving motions of its 16-gram test masses to 10 and 2 nanometers at frequencies of 1 and 3 hertz, respectively. The seismometer would be capable of detecting a magnitude 3 quake within 200 kilometers of the lander. In order to protect the instrument from the shocks and vibrations associated with launch, separation and landing, each test mass was held firmly against a stop by a spring-loaded plunger secured in place by a palladium-aluminum fuse wire. The instrument’s sensors would be released or “uncaged” after landing by passing a current through the fuse wire to cause it to fail nonexplosively by electrical heating.

A schematic of a Viking seismometer detector. Click on image to enlarge. (NASA/JPL)

Because of the limits placed on the amount of data which could be downlinked to Earth, the Viking seismometers had three different modes of operation employing a variety of data compression schemes. In the “high data rate mode”, the data from each sensor was digitally filtered and each 7-bit plus sign word was stored in one of two temporary 2048-bit recirculating memories at a rate of 20.2 samples each second for each channel for a total data rate of 1.77X10⁶ bits per hour. As the memory filled, its data would be transferred and stored on the lander’s data tape recorder. This mode of operation was capable of recording all of the motions of the test masses but was sparingly used because to the large volume of data it produced.

The “normal mode” offered the lowest data rate and was employed for long term monitoring. In this mode, the data were filtered in such a way that only the amplitude envelope of the detected waves averaged over 12.7 seconds were recorded at a rate of 4.04 samples per minute per channel instead of recording the motion of each oscillation of the ground (as was done in the high data rate mode or is seen in conventional terrestrial seismometers). While the details of the ground oscillations were lost in this mode, it allowed monitoring of seismic activity for longer periods of time with only 0.35% of the data volume of the high data rate mode.

The last mode of operation was the event or triggered mode which could be automatically enabled upon the detection of a seismic event above a set threshold. Once again, the data were filtered to record the amplitude envelope (in a 7-bit word) as well as the number of zero crossings (in a five-bit word) so that the number of individual oscillations could be detected. Operating at a rate of 1.01 samples per second per channel, this mode of operation provided a more detailed record of a seismic event but with only 8.3% of the data volume of the high data rate mode. While its design was affected by many engineering tradeoffs, the Viking seismology investigation would provide the first look at the seismic activity of another planet using an instrument with a sensitivity comparable to terrestrial counterparts.

The Viking Mission & Findings

Viking 1 and 2 were successfully launched on separate Titan IIIE-Centaur rockets on August 20 and September 9, 1975, respectively. Viking 1 was the first to arrive in orbit around Mars on June 19, 1976 but the planned landing on July 4 to commemorate the American bicentennial was postponed when initial high-resolution imagery from the Viking 1 orbiter showed the original landing zone to be rougher than expected. After numerous delays, the Viking 1 lander made the first fully successful landing on Mars at 11:53 GMT on July 20 at 22.27° N, 47.95° W on Chryse Planitia.

A view of the Viking 1 seismometer (the small box near the center) as it appeared on July 23, 1976 (VL-1 Sol 3). Unfortunately, the detectors failed to uncage and no usable data were returned. (NASA/JPL)

The command to uncage the seismometer’s detectors on Viking 1 was originally sent on July 25, 1976 (VL-1 Sol 5) as part of a series of instructions to various lander systems but subsequent telemetry showed no change in status. A second set of commands sent specifically to uncage the seismometer was transmitted the following day but to no effect. With the seismometer detectors still caged, only the largest Marsquakes could be detected by the instrument. An investigation into the triply redundant uncaging circuit pinned the source of the failure to the non-redundant interface of the instrument with the lander. Although measurements from the Viking 1 seismometer was monitored for a time, no usable data were returned resulting in the only major (and very disappointing) instrument failure of the Viking mission.

This is a Viking photomosaic of the Viking 2 landing site (yellow dot) about 200 km west of the crater Mie in Utopia Planitia. (NASA/LPI)

Viking 2 entered orbit around Mars on August 7, 1976 with the lander successfully touching down at 22:38 GMT on September 3 at 47.64°N, 225.71° W on Utopia Planitia. The command to uncage the seismometer detectors was initiated 2¼ hours after landing and, much to the relief of Anderson and his team, the instrument responded this time. The first working seismometer on another planet was now ready to get to work. With the initial data returned by the Viking 2 seismometer showing that the ambient seismic activity was low, on Sol 4 the instrument was commanded to its maximum gain setting. While the instrument was acquiring data in its normal mode during lander activities, on Sols 9, 19 and 29 the seismometer was commanded to use its event mode during these times to characterize the signatures of those operations as well as the vibrational properties of the lander – data vital to help Anderson and his team differentiate an actual Marsquake from lander-generated noise. With high data rate downlink from the lander available during the initial phase of the mission, the seismology team took every opportunity to gather data during various parts of the day in the event and high data rate modes.

A color view of Viking 2 at Utopia Planitia acquired on September 24, 1976 (VL-2 Sol 21). The seismometer is the small box between the color calibration targets left of center on the top of the lander’s equipment bay. (NASA/JPL)

By November 1, 1976 (by which time data up to Sol 59 had been analyzed), the Viking 2 seismometer had returned 982 hours of data in the event mode and 99 hours of data in the high data rate mode. As expected, wind-generated signals were a major source of background noise with the seismometer readings strongly correlated with the wind speed independently recorded by Viking’s meteorology package. The winds seemed to be at a minimum in late afternoon a couple of hours before sunset and again in the early morning a couple of hours before sunrise at which time the wind-generated noise was essentially undetectable. A total of 450 hours of data acquired at these quiet times at the maximum gain setting did not record a single seismic event. If Mars had the same level of seismic activity per unit area as the Earth, one or two Marsquakes should have been expected suggesting that Mars was seismically quieter than the Earth, but more data were still needed.

Plots of wind speed (top) and amplitude of the seismometer signals in all three dimensions (bottom three panels) illustrating the correlation between wind and seismic background noise for an event recorded on Sol 36. Click on image to enlarge. (Anderson et al. 1976)

The Viking 2 seismometer continued taking data during its primary mission on a 6-day command cycle. During solar conjunction from Sol 63 to 108 when communications were not possible, the seismometer continued operating at a reduced level. Regular operations for the extended mission began on Sol 109 with two-week command cycles. By early 1977, a total of 171 Sols of data had be acquired with the only gaps being on Sols 103, 104 and 141 to 146. Of this total, 87 Sols of data in the normal mode had been acquired, 75 Sols in the event mode and a total of 63 hours in the high data rate mode. The Viking seismology team had the additional data they needed to come to some firmer conclusions. They continued to find that the amplitude of the seismometer signals associated with wind events varied with the square of wind velocity just as expected from turbulent flow of the thin Martian atmosphere over the lander. A Marsquake candidate was detected on Sol 53 but independent measurements of the winds demonstrated that it was instead the result of a 10 meter per second wind gust.

On Sol 80 during the solar conjunction period (November 25, 1976 back here on Earth), a single seismic event was recorded which bore the signature expected for a genuine Marsquake. The apparent 13-second difference in the arrival times of the P and S waves indicated that the candidate event took place about 110 kilometers from the lander with a magnitude of 2.8 on the Richter scale, assuming that the properties of the Martian crust are roughly similar to Earth’s. A pair of spikes in the record suggested that they might have come from reflections from the base of the crust at a depth of about 15 kilometers, although later analysis of Viking orbiter gravity data (as well as that from subsequent orbiters) are consistent with the crust being about triple this thickness. While the high frequency of the event is in line with a quake, unfortunately, there were no wind data taken at the time of the Sol 80 event making it impossible to definitively identify it as a genuine Marsquake.

Plots for the amplitude and frequency over time in all three dimensions for the possible Marsquake event recorded by Viking 2 on Sol 80. The arrival times of the P and S waves are indicated with the peaks labelled 2 and 3 being possible reflections from the bottom of the crust. Click on image to enlarge. (Anderson et al. 1977)

Continued analysis of the incoming seismometer data returned by Viking 2 failed to show a single definitive Marsquake. With over 1500 hours of low-noise data by the beginning of 1978, Anderson and his team were able to demonstrate with 95% confidence that Mars was seismically less active than the Earth although it would not be possible to determine if the Red Planet was less active than even the Moon using the Viking instrument. With an extended data set, Anderson was able to show how the winds at Utopia Planitia varied with the time of day and season more accurately than Viking’s wind sensors but there was no hint of genuine Marsquake activity through to the loss of contact with the Viking 2 lander on April 12, 1980 (Sol 1280).

An artist’s depiction of NASA’s InSight on the surface of Mars. Its deployable seismometer is the small dome-shaped object on the left in the foreground (NASA/JPL-Caltech)

After nearly a four decades long wait, it will now be up to NASA’s new InSight Mars lander to search for the illusive Marsquakes which will allow scientists to probe the interior of Mars. With its sensitive seismometer deployed on the Martian surface beneath a protective cover to shield it from the winds and the extremes in surface temperatures, InSight will be able to make the most sensitive search for Martian seismic activity to date.

Follow Drew Ex Machina on Facebook.

General References

Don L. Anderson et al., “The Viking Seismic Experiment”, Science, Vol. 194, No. 4271, pp. 1318-1321, December 17, 1976

Don L. Anderson et al., “Seismology on Mars”, Journal of Geophysical Research, Vol. 82, No. 28, pp. 4524-4546, September 30, 1977

D. L. Anderson et al., “Viking Martian Seismology: A Summary of Current Status”, Lunar and Planetary Science IX, pp. 13-14, 1978

D. L. Anderson and Y. Nakamura, “Martian wind activity detected by a seismometer at Viking lander 2 site”, Geophysical Research Letters, Vol. 6, No. 6, pp. 499-502, June 1979

John Longhi, Elise Knittle, John R. Halloway and Heinrich Wanke, “The Bulk Composition, Mineralogy and Internal Structure of Mars”, in Mars (H.H. Kieffer, B.M. Jakosky, C.W. Snyder and M.S. Matthew ed.), The University of Arizona Press, 1992

Andrew Wilson, Solar System Log, Jane’s Publishing, 1987

The Viking Mission to Mars, NASA SP-334, 1976