A Press Release: Initially Launched For Orbits Over Lands
on Earth, the Global Positioning System may be moving and with it Selective Availability is explained and moving to Mars.
Falcon AFB, Colorado. April 1, 1998.
After more than a decade of testing, it was announced today that the Mars Gravity Model (MGM) is now ready for operational use. The MGM will be used to control a re-deployed Global Positioning System (GPS) in orbit around the planet Mars by the year 2000.
Beginning on August 21, 1999 the current GPS satellites (with the exception of the single Block IIR satellite now in orbit) will be "re-deployed" to act as a positioning system for use in the exploration of Mars. The new system will be called the Mars Positioning System (MPS). The current Global Positioning System will be maintained by combining the Russian GLONASS system satellites (functionally the same as GPS space vehicles - having been designed from the same plans) with the Block IIR satellites now in production and planned for launch over the next few years. The GLONASS satellites will switch from their current frequency select mode to the GPS code select mode at the midnight August 21/22 crossover in 1999. Barring unforeseen difficulties the GPS will experience no decrease in functionality during the period of transition. When all Block IIR satellites are in orbit, the remaining GLONASS satellites will themselves be re-deployed in orbit around the Moon where they will be switched back to the frequency select mode so that they will not interfere with the GPS signals.
Because the "red planet" has no facilities to act in the role of Control Segment, the satellites will operate in "autonomous mode" while in orbit around Mars. In order to accomplish this it has been necessary to precisely model the gravity field (geoid) of Mars. Data from all the preceding NASA Mars missions have been incorporated (using an extended 3rd order convoluted Christini filter) into a Mars Gravity Model that has been tested extensively on Earth for a decade during which it has been refined to its final form.
The Mars gravity model has already been incorporated into ranging signals that will be broadcast to the surface of Mars from the MPS satellites. The Department of Defense (DoD) has been using the MGM (in its pre-operational form) as the basis for the so-called "intentional degradation" of the satellite pseudo-ranging signals that have been applied to the GPS navigation signals for the last ten years. The current Selective Availability (SA) program, which has applied the MGM to the GPS ranging signals, has been considered something of a nuisance by GPS users for years.
Degrading the accuracy of GPS from a potential of about forty-five meters to one-hundred meters, SA has never been a real problem for professional users of GPS who have always used differential corrections (DGPS) to achieve 2-3 meter accuracy when needed. While the reasons for it are unclear, recreational users have long wished to know their position to the nearest second of latitude and longitude (about 30 meters) rather than the nearest three seconds (about 100 meters).
This GPS-MPS plan has long been known to the GPS "insiders" who have had to remain silent while the complaints of recreational GPS users and the speculations of the popular press have grown louder and more strident.
Questions have included:
"Why have SA when the Russian system does not?"
"What is the point of SA when our tax dollars are being spent on DGPS?"
"Why can't we use GLONASS and GPS together?"
Recently the press and others have commented on what they thought was a sure sign that GPS design was flawed, when in fact it was designed as a feature of the new MPS. The GPS Block IIa satellites were all designed as potential Mars orbiting space vehicles. They transmit calendar information in the form of a ten-bit binary word known as the Week Number. The GPS count of weeks began on January 5, 1980. The ten bits will "roll over" to begin a count of zero after 1024 weeks on August 22, 1999, just in time for the re-deployment to Mars and the beginning of the new MPS epoch.
While receiver software designers have known about the re-deployment for years, most anticipated the possibility of delays in the Mars system and have used simple heuristics such as the leap second count also transmitted by the satellites to insure that the "roll-over" would have no effect on date calculations. Recreational users and the press however have speculated that this "GPS Week Number" issue is a problem of the same magnitude as the "Year 2000" problem, when financial institutions still using COBOL software from the 1960s will finally have to replace them with FORTRAN or BASIC versions that do not suffer from the "two-digit date field" problem (elevator controllers that were written in hard to understand Assembly Language code will not have to be replaced but it may be necessary to push the "up" or "down" button several times at midnight at the start of the "millennium").
When examined over a period of twenty four hours the MGM (or SA) appears to be a slowly varying bias error in the transmitted signal from each satellite as it orbits the Earth. When plotted with respect to the Mars "day" of 24 hours and 37 minutes, and adjusted for the difference in semi-major axis (Earth = 6,378,137 meters, Mars = 3,403,763 m), the SA variations can be seen to exactly model the gravity field of Mars, allowing the constellation of MPS satellites to operate without monitor stations and the daily uploading of orbital information as required on Earth.
The new MPS will operate with respect to a Mars Centered, Mars Fixed (MCMF) coordinate frame based on the new Mars Mapping and Imagery Agency (MMIA) MGS-99 geodetic datum based on the Taranaki Ellipsoid (semi-major axis=3403763.0 meters, flattening = 1/154.4163) with the Prime Meridian located at the Airy crater at the familiar Sinus Meridiani at 5.2 degrees south latitude. Because the concept of a week is merely a cultural aspect of some Earth calendars, the Week Number will simply be used to count periods of 620,340 seconds (the Martian "day"), and MPS time will be steered to within one microsecond of Airy Mean Time traceable to Universal Time Coordinated (UTC) which has operated independently of the rotation of the Earth for many years. The UTC computation will be facilitated by the extra bits incorporated into the GPS Day Number value in the UTC parameters sent by the satellites. Eight bits were designed into the data frame (page 18) so that re-deployment to any planet (in our solar system) could be accomplished. Currently only three bits are used for the seven days of the Earth week. The period of revolution for Mars is about 686 days (1.88 * 365 days), so dividing by the normal 52 weeks per Earth year gives 13.1 days per week requiring 14 DN values. For Mars re-deployment this only means using bit four of the eight available bits in the DN.
The Mars Positioning System is expected to facilitate exploration, mapping, and exploitation of the planet over the next twenty years (aided by the water resources mapped by GLONASS on the moon). MPS receivers will provide all users with better than 45 meter accuracy on the surface of Mars (due to the difficulty of exactly modeling the gravity field without monitor stations). While this is about the same accuracy available on Earth without SA (due to be discontinued after August 22, 1999), because of the difference in size of the reference ellipsoid used as the basis for MGS-99, this will result in accuracies still only to the nearest three seconds of latitude and longitude on Mars, no doubt a disappointment to recreational users of the new MPS.