tdscanuck From Canada, joined Jan 2006, 12710 posts, RR: 80
Reply 1, posted (3 years 2 months 6 days 5 hours ago) and read 3826 times:
Quoting wardialer (Thread starter): As the leap second approaches, will this be a big concern for GPS nav systems on aircrafts as they must calculate or compute all times very accurately to fly over a certain waypoint?
Most aircraft get their time signal from the GPS constellation anyway, so as soon as the satellites shift their clock the airplane will too.
However, it's not a very big deal...nobody in the ATC system cares if you hit a waypoint to the second. +/- perhaps 20 seconds is just fine in even busy environments for everything except approach, at which point you're under radar control or RNP anyway.
There's not really any risk of position error on the aircraft side either; the signal from the GPS satellites is basically where the satellite was in physical space at a particular time according to the satellite and all position calculations are done against the satellite's time. So even if the aircraft's clock is off, the position calculated by the GPS system should still be right.
This all assumes that the GPS constellation updates properly; however, they've done this many times before so I'm guessing they know what to do.
I don't think they actually stop their internal clocks, I think they change the offset between their clocks and UTC.
Quoting timz (Reply 2): UTC does, but I'd have thought there was no reason for the satellites to pause.
They have to insert some correction because they send UTC down in the signal. They can either shift their own clock or insert an offset between their clock and UTC. I rather suspect the latter but I don't know how they actually implement it.
Quoting timz (Reply 2): They're not trying to stay close to mean solar time, are they?
I don't think there's any reason to keep the internal clocks in sync with that, but there's no reason they couldn't. The satellites orbital planes are fixed relative to the distant stars and they're setup to do two orbits every sidereal day so they're close to solar time but I don't think it's exact.
It seems like it would be simpler to just let each satellite's clock run (counting up from initialization?) and just periodically update the offset between clock time and UTC.
rwessel From United States of America, joined Jan 2007, 2695 posts, RR: 2
Reply 5, posted (3 years 2 months 5 days 16 hours ago) and read 3653 times:
The GPS system itself basically ignores leap seconds, and has been slowly drifting out of sync with solar time. They were 15 seconds fast before the change, and are now 16 seconds fast.
The GPS signal includes the current offset between GPS time and UTC. That number (only) just went from 15 to 16.
You don't actually need to know the correct time for GPS to work, you just need to have the same time as the satellites do. In fact, almost no GPS receivers actually know that time (it would require a rather expensive and large atomic clock to stay sufficiently close over any reasonable period of time), rather the GPS receiver solves for the time as part of the process of determining the position. In a very simplified sense, the GPS has receiver has four unknown variables (its X, Y and Z coordinates and the time), and four equations (basically it can write an equation for the distance to each of the four satellites its tracking). So four equations and four unknowns, and you can solve for your three dimensional position and the current time. If a GPS receiver actually knew the current time with sufficient accuracy, it would need only three satellites to produce a three dimensional fix. Obviously real GPSs can make use of more than four satellites to get a better position.
But the GPS/UTC time offset is not really used, it's mainly just passed out to the systems using the output of the GPS receiver (for example, you'd expect that to be used by the time display on the GPS). Obviously any systems that mishandle the change in the offset may have problems, but the GPS system (and the receivers themselves - at least the parts that determine the current position and time), are not going to be affected.
nomadd22 From United States of America, joined Feb 2008, 2031 posts, RR: 0
Reply 6, posted (3 years 2 months 5 days 9 hours ago) and read 3578 times:
To put it in a little simpler terms, the time that your GPS equipment displays is controlled by an offset sent from the system. The leap second won't have any effect on the operation of the system or the recievers. The offset will just change by one second to keep the displayed time correct.
BigSaabowski From United States of America, joined Jul 2008, 181 posts, RR: 0
Reply 10, posted (3 years 2 months 2 days 6 hours ago) and read 3040 times:
Quoting rwessel (Reply 8): That's correct. The GPS signal includes the current offset between GPS time and UTC for the convenience of the end users.
Also, many non GPS users rely on the time being sent over their air from the satellites to be exact. For example, most TV and radio stations have a GPS receiver for the purpose of keeping the exact time for obvious reasons.