Slashdot Mirror
GPS Accuracy Could Start Dropping In 2010
adamengst writes "A US Government Accountability Office report raises concerns about the Air Force's ability to modernize and maintain the constellation of satellites necessary to provide GPS services to military and civilian users. TidBITS looks at the situation and possible solutions."
solar event will cause transient events that will recover in a few seconds.
GPS2F was awarded in the early 90s with a launch date of more than 10 years out. This caused parts issues that significantly magnified design issues. Without going into company secrets, let's just say that bean-counters and engineers fought long and hard. I wonder why Boeing lost GPS3...
If LockMart can't deliver as promised, Airforce can always buy more IIF. After 12-or-so builds currently on contract by Boeing, you figure even the incompetent can get their bugs worked out by then (sans part issues)
You don't need military GPS to be that accurate, it can be done with differential phase GPS. See: here. By using a fixed base station at a location with known coordinates, one can expect to see accuracies in the 1 to 2 cm range as long as the receiver is within 10's of km from the base station. There are several manufacturers who make gear that can achieve this level of accuracy, see Leica, Magellan, and Sokkia. I've been using Leica gear at work mostly, and have see ~1cm accuracy under good conditions pretty consistently. A lot of legal surveying in remote areas is done exclusively with GPS, especially in the northern parts of B.C and Alberta. I've done legal surveys with GPS in the Vancouver area, but getting high accuracy in urban areas is more difficult because of multi-path noise and qoor signal quality from obstructions such as buildings. Also people in the city get mad when you cut down trees to get better reception ;)
Even when trying to take a fix on the same point over time, the margin of error is random. Most of the error is due to propagation delays of the signal. The atmosphere is constantly changing density - the jet stream can raise or lower air pressure between you and satellite #5 at this point in time, or between you and satellite #23 at another point in time. The ionosphere can cause delays or reflections. Nearby buildings, or cars, or airplanes can cause ghost signals. Remember your high school physics teacher who always qualified the the speed of light with the phrase "in a vacuum"? It really makes a difference.
Remember, the GPS satellites are flying in lower earth orbits, and are constantly moving over and around you. They are not fixed relative to you the same way a geosynchronous satellite is (like the DirecTV satellites.) So even if the weather patterns could be precisely measured and figured out, they would be meaningless as soon as the satellite moved - and they're always moving very, very fast.
Differential-GPS improves accuracy by having a mounted GPS receiver at a precisely surveyed point, then transmitting (via FM) the "error difference" between what is derived from the GPS system and the surveyed coordinates. If the error at a particular time is 3 meters at the DGPS antenna, then the error at any other GPS receiver in the area is probably pretty close to 3 meters at that same time (assuming the weather patterns are similar between you and the DGPS antenna.)
You can indeed leave a GPS fixed in a point and average the readings to reduce the margin of uncertainty, but you'll never achieve the high accuracy needed for surveying.
John
A friends mom escaped the wreck of a 90ft Fish Packer as it hit the rocks at night in a passage with strong currents due to a problem caused by relying on GPS. It was due to something like how it derived the heading vs the direction of travel or some-such.
Something wrong there. Both LORAN and GPS only give position (GPS gives time too, but that doesn't help here). Direction of travel is determined in both systems by taking the difference in position over a known time interval. GPS can give heading by using the phase difference between receivers on different parts of the vessel, whereas the wavelength of LORAN was probably too long for that to work. Upshot is, a problem involving headings and direction of travel isn't the fault of GPS, and using LORAN would have been no defence. It may have been a problem with the GPS receiver software, but LORAN calculations could go wrong too. Most likely it was a navigator not understanding the systems they had.
Quidnam Latine loqui modo coepi?
The DGPS system at Libby Dam in Montana is accurate to 1-mm vertically:
Performance Monitoring Of Libby Dam With A Differential Global Positioning System UNITED STATES SOCIETY OF DAMS (JUNE 2005)
There's more than just GPS.
Here's a listing of all current/proposed global systems. Regional only systems such as IRNSS or Beidou1 are not listed.
GPS - United States - Fully Operation
Galileo - EU/China/Israel/South Korea/Norway/Etc - 2013
Beidou2/Compass - China - 2012?
GLONASS/ - Russia/India - Complete in 2010