The Viking Mission to Mars James E. Tillman Viking Meteorology Science Team and Director, Viking Computer Facility University of Washington Seattle, Wash., 98195 Keynote Address Prime Computer Users Group National Meeting Orlando, Fla. 1984 1 Introduction The planet Mars has often stimulated the imagination and curiosity of mankind. With the following material, I hope to show some of the intriguing aspects of the planet, as observed from Earth, aspects which generated a strong desire to observe and explore the surface. These Earth bound observers have long known that Mars has seasonal changes, with dust storms and a varying polar cap. As its atmospheric and surface conditions most resemble those of Earth, when compared to the other planets in our solar system, the possibility of life and its relation to terrestrial forms, provided a major portion of the scientific impetus for the Viking Mission to Mars. The Viking Mission to Mars was the most complex scientific exploration of any planet other than Earth conducted by mankind. In the sophistication of the remotely operated instrumentation, it surpasses even the Apollo Lunar Missions. Due to the exceptional effort of its engineering, management, operations and scientific staffs, it exceeded by far the design goals and expected lifetimes of the systems. For example, Lander 1, the Thomas Mutch Memorial Station, ceased operation after 2,245 sols ( 2,306 days ) on Mars, which should be compared with its designed lifetime of approximately 120 sols. In the following segments, some of the highlights of the Viking exploration of Mars will be presented. Some of the interesting aspects of our use of Prime computers for the Viking program will be discussed as well as our acquisition of hardware for and development of the computer driven display, "The Viking View of Mars" for the Smithsonian National Air and Space Museum. This permanent exhibit presents text, graphics and images for the public and should be on display during this meeting. 2 Mars 2.1 Observations of Mars from Earth Several characteristics of Mars have contributed to our interest in it since prehistoric times. It has a reddish color, it is easy to observe, its motions are not a simple circular orbit and, since the invention of the telescope, it has been observed to undergo significant seasonal changes. Seasonal changes of major interest are its development of polar caps, the differences in its albedo, or contrast, on a day to day and year to year basis, and the identification of linear features on the surface as "canale", singular ( Father Secchi ), about 1869 and "canali", plural ( Schiaparelli ) about 1877. Percival Lowell seems to have pushed the linear features to the extremes of artistic endeavor and speculation, in that in 1908 he published a book "Mars as the Abode of Life" wherein he speculated that the canals were the work of "intelligent creatures, alike to us in spirit, though not in form." The possibility of life, but at a far more primitive level than that envisioned by Lowell, was a major aspect of the Viking Mission. Although we have not detected life on the planet, it has the environmental characteristics most similar to Earth and consequently, most likely of the other planets or moons to support life. However, as meteorologists, our observations from the surface of Mars have provided information, conclusions, new phenomena and speculations far beyond my hopes or even dreams at the beginning of the mission. Scintillation in the atmosphere due to temperature variations, limits the resolution of Earth based astronomical observations such that telescopes of a few inches diameter provide as high resolution as can be simply obtained: larger telescopes are constructed to view dimmer objects. However, Earth based observations are valuable in that we do not currently have any active spacecraft at Mars and we have found important year to year differences in this climate of Mars. They supplement our Mariner and Viking observations by providing data on the seasons in which dust storms form and some information as to their extent and intensity. Since Mars is too far from earth for useful observations most of the year, these data can not be very continuous, even if the weather were always cooperative. 2.2 Pre Viking spacecraft observations The first close observations of Mars by NASA spacecraft began with the Mariner 4 fly by in 1965 which took 21 photographs. Each photo contained 240,000 bits of data, transmitted to Earth at 8.33 bits per second. Mariner 4 showed craters and a thin haze in the Carbon Dioxide atmosphere. Two later Mariner flyby's in 1969, mainly revealed only a cratered terrain without major geological features. However, the observations of the Mariner 9 orbiter soon changed the perception of Mars. During approach, a small dust storm in the southern hemisphere developed into an intense global dust storm, injecting dust to heights well above 30 kilometers. At the most intense period, only a small, low contrast feature could be seen. Once the dust cleared, this feature was revealed to be a volcano, Olympus Mons, whose top was 29 kilometers above mean Mars level! Other interesting features were the chain of three other volcanos almost as high as Olympus Mons and a canyon system 5,000 kilometers long, up to 200 kilometers across and up to 7 kilometers deep. Since the diameter of Mars is only 53% that of earth, these large canyons and volcanos seem even more dramatic. At its base, the diameter of Olympus Mons is more than 500 kilometers. In the next section, some of the highlights of the Viking Mission to Mars will be covered.
3 The Viking Mission to Mars The Viking Mission was the most ambitious planetary exploration program undertaken by NASA or by any agency. In terms of its scientific goals and complexity, it even surpassed the Apollo Lunar missions, although not in cost or operational complexity. Its success was due to reasonable support in the early and middle stages, to excellent management, and to the dedication of its staff, contractors and vendors. There are numerous instances in my own experience where I requested assistance from a vendor, ( Prime received many of these ), and tasks that normally took weeks were somehow accomplished in hours or days. For example, during our upgrade from a Prime 300 to 400 CPU, we were able to have a government ADP plan approved within several weeks of submission. 3.1 Mission design The initial mission design consisted of a 90 day nominal mission using two identical spacecraft systems. Each spacecraft consisted of a orbiter-lander pair, with the orbiter providing imaging of surface, some science measurements, and high rate communications support for the lander during the nominal mission operations phase. Operations were planned around a nominal 90 day mission where one spacecraft would reach Mars roughly a month prior to the other: both spacecraft would operate simultaneously after the landing. Early in the mission design review process, reviewers from the Mercury, Gemini, Apollo and other projects, favorably commented on the mission design but indicated it would not be possible to operate both spacecraft with the staff and facilities available, even though the mission operations staff was roughly 1,000 full time individuals. A plan was developed to operate the first Lander for approximately 45 sols, ( a sol is a Martian day of 24 hours and 37 minutes ) and then to reduce its activity to a minimum while landing and operating the second lander for a similar period. In this manner, the flight control teams were able to maintain reasonable 10 to 11 hour work days! Launch was scheduled for the summer of 1975, followed by an 11 month cruise to Mars. Landing was scheduled for July 4, 1976, summer in the northern hemisphere and due to the low temperatures on Mars, the system was not required to operate through the winter.
3.2 Early use of Prime in the Viking Mission Since testing by the Viking Meteorology Instrument System, VMIS, was inadequate due to the small wind tunnel constructed to simulate Martian pressures and its CO2 atmosphere, we decided to perform additional testing at NASA Langley Research Center during the summer of 1975. The original plan was to replace the paper tape punch of the TRW VMIS test set with a tape drive and use it for the tests. The proposal to accomplish that and associated work was $200,000 and I proposed that we develop a test set at UW which could both gather and reduce data in real time. After preparation of a lengthy proposal and an extensive evaluation, a Prime 300 was selected during the summer of 1974 to be used for wind tunnel testing of the VMIS. Hardware interfaces to simulate the lander computer were developed, software written and the system checked out between its delivery in September 1974 and its shipment to LRC in June of 1975. I chose RTOS as an operating system, a mistake, since hindsight proved that DOS/VM, ( later Primos III ), would have been a better choice due to our low data rates and its greater software maturity. The system had to function properly, and reliably, from the beginning of testing as some flight qualified hardware was only made available after launch and the NASA is generally scheduled several years in advance. The Transonic Dynamics Tunnel, TDT, has a diameter of 16 feet, generates wind speeds from 10 MPH to transonic and can provide pressures from 1% to 100 % of atmospheric. Another index of its capabilities, is that it is powered by a 20,000 HP motor which draws 1,000,000 watts at our lowest testing speeds. During 1974, a specific two week time slot was assigned to us at no cost, in this facility whose testing cost was $4,000 per hour. The cost of the computer system was less than 10% of the value of the test time. Initial testing was begun in June 1975 in a facility especially designed for us and was interrupted so that we could take part in the launch at Cape Kennedy. After launch, the Flight Spare VMIS was made available for testing and we moved to the TDT. Testing was successful, and as expected, we had to upgrade our software while our instruments were on the way to Mars. This was somewhat unusual, since all other software were essentially in their final form for training exercises at JPL between launch and Mars encounter. The system was moved back to UW in November of 1975. Once it was determined that our complex software would have to undergo significant changes, we found that adequate time would not be available on the JPL Univac 1108 systems ( an extra 1108 was installed for Viking ) and we studied the possibility of providing support at UW on the Prime for users at Martin Marietta, Denver, NCAR, Boulder, TRW, Redondo Beach, JPL, Pasadena and UW. Since it was clear that the P300 would be overloaded, discussions were initiated with Prime as to possible solutions. The architecture of the soon to be completed P400 was discussed with J. W. Poduska as well as its probable performance and availability. A presentation was made to the Viking Project management to purchase a 60 mbyte disk drive and the new CPU as soon as it became available and to provide the software development facility at UW. The proposal was accepted, and we began 24 hour/day operation in November of 1975 which continues at the present with the exception of one week for the installation of air conditioning. In April of 1976, we received delivery of Serial # 2 of the P400 CPU's.
3.3 Mission operations during the Primary Mission Mission Operations at JPL became intensive after launch of the spacecraft. In the normal course of events, the VMIS system would not be operated more than a few times during cruise to check its basic functionality and we would have mainly been involved in training for Mission Operations. However, drift was detected in the Lander 2 temperature system and the system was activated a number of times to determine the magnitude and nature of the drift. Although the cause of the drift was not discovered, we developed corrections for the early mission measurements. In June of 1976, we shipped the system to JPL for Mission Operations. It arrived at 8 A.M. and with the assistance of movers, we installed it and were operational by noon. The system was mainly used for science analysis, using a version of our Mission Operations meteorology software. The project's Mission Operations software was configured to produce a meteorology tape after some preliminary processing for use on Prime, prior to the more extensive data reduction segment of the program. When the first data were relayed from the lander tape recorder, we obtained this tape and produced the first meteorology data from the surface of another planet, Mars, on our system. Since it was summer in the tropics at our landing site, the wind was light and variable, much the same as on earth under similar conditions. The project director was James Martin of NASA Langley Research Center and the success of the mission was due in a large part to the competence and dedication of Jim and his staff. A number of possible landing sites had been selected prior to Mars encounter on the basis of the previous Mariner 9 mapping and Earth radar data. It was the function of the site selection group to choose an acceptable and safe site from the previously selected sites. At least a week in orbit around Mars was planned for this activity and it was desired to land on July 4, 1976, the bicentennial. However, engineering prudence sometimes delays plans and desired timelines. The Viking orbiters were capable of resolving objects larger than about 40 meters and, in the ideal case, one would use observations of roughness elements larger than this and extrapolation to estimate the chance of spacecraft disturbing boulders. However, all cases are not ideal and earth based radar data was required to supplemented the orbiter images. Although the radar can not measure individual small features, they can estimate the size distribution of boulders, etc in large areas which are smaller than can be resolved in the images. The combined data indicated that the previously chosen sites were not safe and the Fourth of July passed as the search went on. The search could not go on indefinitely due to due to the cost of maintaining 1,000 of us at Mission Operations and due to the fact that the second spacecraft was rapidly approaching and would have to occupy the resources of the flight operational staff. Another site was selected on the basis of radar and orbiter data and landing was initiated. Due to the one way propagation time of 18 minutes from Mars to Earth, the landing sequence had to be automated. The entry and landing procedure consists of three phases. First, the lander separates from the orbiter and retro rockets cause it to enter the atmosphere, protected by its heat shield. At about 6,400 meters, the heat shield is jettisoned and the parachute is deployed. Finally at about 1,200 meters, the parachute is jettisoned and the retro rockets are ignited with the descent being under control of the lander's Guidance and Sequencing Computer, GCSC, radar altimeter and inertial reference unit. The first landing was completely successful, witnessed by many thousands of staff and families at JPL and by millions throughout the world. The second landing proceeded 44 sols later with fewer problems in the site selection process. Although Lander 1 was to be placed in a low activity mode for Lander 2's arrival, we managed to schedule continuous meteorological data collection from both landers. To have both landers and orbiters functioning successfully, was the result of good design, management and planning. This is not to imply that there were not a few "cliff hangers" such as the Biology Instrument and the GCSC's plated wire memory, but the final results were more than satisfactory. At the end of nominal mission operations, the Prime system was moved back to UW in November of 1976. 3.4 Science An idea of the complexity of the scientific instrumentation can be gained by considering the experiments. The major individual ones are: 1) three biology experiments, 2) organic analysis, 3) a gas chromatograph-mass spectrometer serving both organic and atmospheric analysis, 4) stereo black and white and color fascimilie imaging system 5) seismology 6) meteorology 7) other supporting components and experiments such as the soil sampler arm used to feed the experiments and to determine soil properties. To appreciate the stringent engineering requirements, one should also be aware that the complete lander operated on an average power of 50 watts and had to be sterilized at 130 degrees Celsius for two 24 hour periods. These simultaneous requirements, mandated innovative and careful design as well as somewhat higher than normal development and component cost. Another indication of the success is that the proceedings from only two of the special Viking conferences weigh 8 pounds and contain 1368 pages: this is only a small portion of the Viking generated research publications. As to specific results, the consensus is that no evidence for life was found even though the biology experiments reacted in a strongly positive way. The reason for the reaction is that the Martian sols contain compounds that liberate oxygen in the presence of water. The atmosphere is composed of 95% CO2, small amounts of nitrogen and oxygen, as well as traces of Argon, hydrogen and other gases. In the field of meteorology, we found that fronts on Mars were more similar to those on Earth than we expected. However, in one instance at Lander 2, ( Lander 1 was at 22 degrees north while Lander 2 was at 48 degrees north ) six or seven fronts passed by at almost identical time intervals and strengths, as measured from the pressure data: on earth, we rarely see such regularity. Since we were able to continue operation for more than three Martian years, we were able to observe year to year similarities in many meteorological phenomena as well as differences. Major atmospheric, and over long time scales, geological, process are the global dust storms which decrease the daily average temperature on the order of 14 degrees Celsius or approximately 25 degrees Fahrenheit. As the dust remains in the atmosphere for many tens of sols, the effect was thought to be similar to that discussed in Nuclear Winter: Global Consequences of Multiple Nuclear Explosions, _______________________________________________________________________ R.P. Turco, O.B. Toon, T.P. Ackerman, J.B. Pollack and Carl Sagan, _______________________________________________________________________ Science, 23 Dec. 1983, pp 1283-1292. _____________________________________ An interesting, and as yet unexplained characteristic of the dust storms, is that some years have major storms while others do not. For example, Year 1 had two, years two and three had none and the beginning of year four had one. We do not know about the rest of year four due to the landers' failure during this fourth year storm, the most intense observed from the surface of Mars. Another unusual feature is the discovery of transient dust storms, which seem to repeat at the same time of year, and which indicate a mode of global oscillation in the atmosphere of Mars. Many other interesting, and important, atmospheric processes have been studied on Mars. One is the year to year similarity in the formation and sublimation of Mars' polar caps. Mars has an inclined axis of rotation of 25 degrees Celsius, similar to Earths, which produces a large annual temperature range. Due to its low temperatures and carbon dioxide atmosphere, around 20% of the atmosphere condenses, in the form of "dry ice" each year with surprisingly precise repeatability each year despite the variation in dust storm intensity and number from year to year. A difference is the erosion of soil from year to year. During the third year, between sols 1720 and 1756, ( there are 669 sols per Martian year ) piles of dust placed by the soil sampler moved as well as small pebbles. These were probably accompanied by local dust storms but why did they not move in previous years and why was there no global dust storm this year? We hope that analysis of the data that we presently have will provide some insight into these questions in addition to numerical modeling of the processes. Martian meteorology is important, as well as interesting, in that its atmosphere resembles that of Earth more closely than any of the other planets, or moons, in our solar system which have atmospheres. By testing our physical and numeric models of atmospheric motion, climate, etc. on Martian observations as well as terrestrial ones, we can refine them to better explain and predict the effects of changes or differences in the variable parameters. For example, the effect of major amounts of dust on the atmosphere are sometimes easier to understand Mars, than on Earth, since it swamps the other meteorological variables at times. However, there are no funds if FY85 for Mars Data analysis!
4 Real time operations at the Viking Computer Facility One of the main problems with the Viking Mission was its success! Prior to landing, I suspect that most of us would readily have traded the chance of a multi-year mission for a guaranteed 90 day mission without hesitation. It the end of the nominal mission we were faced with four healthy spacecraft and an excellent flight operations team. During the next year or so, we at UW were content to continue receiving data from JPL in processed science form even though we had operational software to obtain meteorology data from the meteorology Front End Processor, FEP, tapes produced by the JPL UNIVAC 1108 system. However, if we were to be able to obtain meteorology data in the future, it was clear that we would have to implement some of the operational software which handled the raw spacecraft data stream from the Deep Space Network, DSN, since the flight operations IBM 360/75's were certain to be decommissioned soon. A minimal set of this software was implemented on the VCF's Prime for use in obtaining meteorology results from the meteorology science data format in the raw data stream and pressure from the engineering data format. While implementing this capability, changes were made in the data format and block lengths by NASA, which we included as options. In January of 1981, we began processing data on a weekly basis, including a comprehensive set of engineering parameters for the operations team at JPL. Data were provided to JPL by dial up access or by mail. In the next few months, we expanded the engineering processing, including plotting selected, and then all, engineering parameters. Prior to our conversion of the software, data were plotted by hand at JPL if at all. Around that time, it became clear that the lander might continue to function for many years and dedicated, part time, Viking staff members initiated a program to recondition the two dead nickel-cadmium batteries of the four on Lander 1. To assist in the rapid turnaround required for this effort to be successful, I proposed that we implement a direct, real time link between JPL and the VCF since the previous method was to mail tapes from JPL to UW. This was accepted and after several iterations, the communications configuration of Figure 3-8, Telecommunications and Data Acquisition System Support for the Viking 1975 Mission to Mars, The Viking Lander Monitor Mission May 1980 to March 1983, D. G. Mudgway, JPL publication 82-107 was implemented. The synchronous NASA communications codes were decoded and input to an AMLC port on the VCF using an Intel system. First testing of the system was accomplished on May 14, 1982 and our permanent installation was implemented in October of 1982. The only special provisions made on the Prime end were to increase the size of the AMLC buffer to around 1,000 bytes to preclude the possibility of buffer overflow. Since the data rate from Mars to Earth was 1,000 bits/sec, there was no need to assign a high priority to this task. With our real time capability, we were able to provide immediate analysis of the battery conditioning results for planning the following sequences in a timely manner.
5 Development of the Smithsonian National Air and Space Exhibit Prior to the establishment of the real time link, we had been displaying lander images on our AED 512 image processor. Once the planning for the link was initiated, Dr Farouk El-Baz of the Smithsonian National Air and Space Museum was consulted as to whether the Museum would like to have a weekly picture "Live from Mars", provided we could obtain the donation of an image processor. After a commitment of AED to donate an image processor I requested that Prime donate a used system for the Museum. After long negotiations with the various parties, and Stan Kent of the Viking Fund providing maintenance funds, a system was donated including two Prime microcoded MPC 4 controllers, to allow high speed DMA communications between Prime and the image processor. Microcode for the MPC 4 was written by Noel Cheney of the Atmospheric Sciences Department at UW, while we continued to operate 24 hours/day. To minimize impact to users and disruption of our continuous weather data collection and processing, initially the testing of MPC 4 microcode was done between 7 and 8:30 AM. To make the development of this code possible under such constraints, Dr. Harold Edmon of the VCF wrote a debugger for microcode development as well as a routine to down load microcode without cold or warm starting the system. User interfaces for the MPC 4 and image processing software were written by James Synge of the VCF. William Guest processed and generated the meteorology graphics that are included in the display and Neal Johnson provided in Viking data processing. The sequence of text, graphics and images that are presented at the Museum, were developed by Dr. Ted Maxwell of the museum, Rachel Tillman and me. Finally, I designed the overall system as well as convincing the interested parties to donate the required resources. Subsequent to our development of the display, which requires only a special cable and our microcode to drive the AED, we have developed an expansion capability which permits multiple AED's to be driven by one port with a small interface box, which supports up to three local image processors and provides line drivers to another similar interface. The next interface can be located thousands of feet away and in turn can support several image processors. While this does violate our "commercial hardware only" policy it does allow high speed and significant flexibility with minimal additional hardware. I hope that such features will be provided as off the shelf items in the near future by Prime or others. Unfortunately, schedules and other complications delayed the opening of the exhibit until after our loss of the lander. However, during one of our routine real time downlinks from Mars, we processed and transmitted the weekly image to the Smithsonian system via Primenet within an hour of receipt of the data by our facility as a simple demonstration which required no software changes on our part. Primenet was routinely used between NASM and UW during our joint development and debugging of the display. While the loss of the "Live from Mars" aspect, lessens the excitement, the display was never planned to include more than a small amount of the latest data, due to the once a week transmission. Also, the fact that the data are transmitted at Mars noon coupled with the 37 minute difference in day length, guarantee that the majority of the downlinks would not have been live during public hours. I would recommend that you take advantage of the opportunity to view the exhibit, at the Museum due to the other historical items associated with the display and during this meeting. In the Smithsonian, it is a permanent exhibit buried upstairs in the "Exploring the Planets" gallery, which requires diligence to find. The Viking Mission to Mars was the most sophisticated and interesting planetary missions ever executed. The enclosed table summarizes some of the operational parameters and it should be remembered that the landers both lasted through winters with temperatures of -118 degrees Celsius and Lander 1, the Thomas Mutch Memorial Station, lasted 2,245 sols, roughly 20 times its design life. Such performance is a proper tribute to the spacecraft designers and operators, as well as the Project management! As a follow on, I suggest that we begin planning for a manned mission to Mars, inviting and encouraging all nations to take part.Copyright, J. Tillman:
Viking Lander "As Built" Performance Capabilities. NAS1-9000, June 1976. Martin Marietta Corp., Denver Division, Denver, CO 80201
Figure 2 Mission Operations Communications including Real Time. NASA Deep Space Network >> Jet Propulsion Laboratory >> University of Washington >> Smithsonian National Air & Space Museum, Washington DC. From "Telecommunications and Data Acquisition System Support for the Viking 1975 Mission to Mars, The Viking Lander Monitor Mission May 1980 to March 1983", D. G. Mudgway, JPL publication 82-107. ______________________________________________________________________ ______________________________________________________________________ Parameter V0-1 V0-2 ______________________________________________________________________ Number of days from launch to end of mission 1813.98 1049.52 Number of orbits of Mars 1488.0 706.1 Number of pictures recorded in orbit 36,622 16,041 BILLION data bits played back from 357.7 161.3 the tape recorders (including lander relay data) Tape travel across recorder heads, km 2955 1397 Number of commands sent by the Network 269,500 Number of tracking passes supported by the Network 7,380 Hours of tracking time provided by the Network, 56,500 _________________________________ Table 1. General Viking Orbiter Statistics ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________ Event Viking 1 Viking 2 ______________________________________________________________________ Launch Aug. 20, 1975 Sept. 9, 1975 Arrival June 19, 1976 Aug. 7, 1976 Landing July 20, 1976 Sept. 3, 1976 Site Chryse Planitia Utopia Planitia Coordinates 22.3 N, 48.0 47.7 N, 225.8 Orbiter in orbit 1,509.9 days 718.8 days Lander active on surface 2,245 days 1,316.1 days End lander operations Nov. 13, 1982 April 11, 1980 End orbiter operations Aug. 7, 1980 July 25, 1978 Orbiter photos 51,539 Lander photos More than 4,500 Photo coverage 97% of planet with resolution of 300 m (1,000 ft) or better. 25% of planet with resolution of 25 m (82 ft) or better. Lander weather reports: more than 1 million Orbiter infrared observations: more than 1 million Orbiter weight: 2,325 kg Lander weight: 571 kg Orbiters built by Jet Propulsion Laboratory Lander built by Martin Marietta Aerospace Project managed by NASA Langley Research Center ____________________ TABLE 2 ______________________________________________________________________ ______________________________________________________________________