SoylentNews is people
SoylentNews
Mainly due to the slowing down of earth rotation, it is about time to add another (the 35th) leap second to UTC in order to keep its time of day close to the mean solar time. NASA features an explanation of why the leap seconds have to be inserted, how earth rotation can be measured precisely and why it is impossible to give precise predictions on when the next one will happen.
The length of day is influenced by many factors, mainly the atmosphere over periods less than a year. Our seasonal and daily weather variations can affect the length of day by a few milliseconds over a year. Other contributors to this variation include dynamics of the Earth's inner core (over long time periods), variations in the atmosphere and oceans, groundwater, and ice storage (over time periods of months to decades), and oceanic and atmospheric tides. Atmospheric variations due to El Niño can cause Earth's rotation to slow down, increasing the length of day by as much as 1 millisecond, or a thousandth of a second.
VLBI [Very Long Baseline Interferometry] tracks these short- and long-term variations by using global networks of stations to observe astronomical objects called quasars. The quasars serve as reference points that are essentially motionless because they are located billions of light years from Earth. Because the observing stations are spread out across the globe, the signal from a quasar will take longer to reach some stations than others. Scientists can use the small differences in arrival time to determine detailed information about the exact positions of the observing stations, Earth's rotation rate, and our planet's orientation in space.
Original Submission
(Score: 4, Informative) by Anonymous Coward on Tuesday June 30 2015, @12:17AM
You are basically correct. You add them for the same reason you add leap days. However, because our lives are now affected by time measurement on the second or sub-second level (e.g., GPS), these seconds do add up. Many things operate on GPS time, which does not include leap seconds, so to convert GPS measurements to UTC, you need to figure in the corresponding amount of leap seconds (I think it is now up to 17 with this). A spacecraft moves quite a bit in orbit in 17 seconds, so if you grabbed imagery off of the ISS and didn't figure in your leap seconds, you'll see a different part of the Earth.
However, the ITU has been soliciting comments from people who are affected by these leap seconds. A nice summary page of the pros and cons is here [wordpress.com].
(Score: 2) by kaszz on Tuesday June 30 2015, @12:27AM
So have one clock time which is UTC that includes leap seconds and other leap stuff. Then another clock time that is strictly time based and that have no synchronization to the Earth specific movements.
(Score: 2) by takyon on Tuesday June 30 2015, @12:36AM
Time is relative! Muahahahaha!
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 1) by kazzie on Tuesday June 30 2015, @06:24AM
Lunchtime doubly so.
(Score: 3, Interesting) by bob_super on Tuesday June 30 2015, @12:38AM
https://en.wikipedia.org/wiki/Coordinated_Universal_Time [wikipedia.org]
https://en.wikipedia.org/wiki/Universal_Time [wikipedia.org]
UT1 is continuous and drifts from true noon as the earth slows down. UTC isn't continuous but tracks the earth to less than a second.
UTC is a royal pain when doing broadcast TV, especially in the US.
Don't get me started on pre-1972 seconds...
(Score: 2) by bob_super on Tuesday June 30 2015, @12:43AM
ARGH!
Wrong brain, Sorry!
https://en.wikipedia.org/wiki/International_Atomic_Time [wikipedia.org]
TAI is continuous, based on "seconds since"
UT1 is earth rotation
UTC is based on TAI, such that |UT1-UTC| 1s (leap seconds needed)
(Score: 2) by kaszz on Tuesday June 30 2015, @12:50AM
So people should use TAI for GPS and broadcast etc. And UTC for alarm clocks etc. And everybody will be happy?
(Score: 2) by bob_super on Tuesday June 30 2015, @12:57AM
In a nutshell, but this perfect world is screwed because humans who only occasionally care for 27 seconds, use machines who always do.
And instead of going TAI, humans have designed machines who seem to all have a different way to handle UTC leap seconds...
The most fun part of the dialog when broadcasters were trying to standardize time, is "how does a machine which just rebooted know what the time is". Trust me, it's a whole lot more complex than it sounds.
(Score: 2) by TheLink on Tuesday June 30 2015, @08:00AM
Most programs can handle that better than they can handle leap seconds - to them it's just like suddenly the CPU got faster - most modern programs should be able to deal with CPUs running faster and slower. Most programs and their use cases can also handle ten different seconds throughout the day each taking 10% longer than it should normally. Or just have 10000 seconds that each take 0.01% longer.
Leap seconds are mostly a stupid idea. Who wants their clock to ever show something like "23:59:60" ?
(Score: 3, Insightful) by iwoloschin on Tuesday June 30 2015, @11:07AM
Are you actually going to notice it showing the extra second? I mean, literally, look away for a second and you miss it!
(Score: 2) by bob_super on Tuesday June 30 2015, @03:23PM
> Who wants their clock to ever show something like "23:59:60" ?
Someone who likes the idea that the time is right?
The Incas had the whole planetary rotation figured out hundreds of years ago. By introducing amazingly better precision in the measurement of time, should we have to give up on the basics?
(Score: 2) by tangomargarine on Tuesday June 30 2015, @05:33PM
how does a machine which just rebooted know what the time is". Trust me, it's a whole lot more complex than it sounds.
CMOS.
"Is that really true?" "I just spent the last hour telling you to think for yourself! Didn't you hear anything I said?"
(Score: 2) by Non Sequor on Tuesday June 30 2015, @01:52AM
The argument that breaking UTC from UT1 would require changing laws, regulations, and technical specs seems like self inflicted brain damage. All of these things are already overruled by enforcement of a de facto standard through time synchronization mechanisms and they all have to tolerate some drift due to variations in synchronization quality among various parties (or else they must already have some mechanism for policing drift).
Doesn't the GPS network effectively define its own context specific time scale which should be sufficient relative to the accuracy of its distance coordinates? Skew in time signal between satellites should be independent of the relationship between the satellite timescale and terrestrial timescales, so long as the satellites stay in synch in terms of the time they let treat as elapsed.
Write your congressman. Tell him he sucks.
(Score: 0) by Anonymous Coward on Tuesday June 30 2015, @06:51PM
You are correct that GPS keeps its own time. It is a continuous counting clock that started in 1980 [nps.edu], and since it uses atomic clocks, it is a pretty nice clock indeed. Because of the ubiquity of GPS, it gives us all a very nice clock in our watches, phones, cars, etc. Anything with a GPS receiver in it automatically becomes a nice clock. However, to be useful, one needs to convert GPS time to UTC, which is where the leap-seconds come in to play.
Now let's suppose you have a few billion burning a hole in your pocket and you want to go into the space imaging business. You put up some cubesats. They take pictures of the Earth, and time-stamp their images with GPS time, which makes the most sense. Now you need to know where on the Earth you're pointing. You get that from your orbit, which is specified as a Two Line Element (TLE), which is defined at a specific point in time, in GMT. That inexpensive GPS receiver you used on the cubesat had the firmware to convert GPS time properly, but it can't be updated and it will be off by a second the next time a leap second gets added, so you'll need to handle that in ground processing. Or, you do away with the leap seconds so your cubesat now always keeps continuous time, but now you need to work your orbital calculation side. Leap seconds are not really an issue when you're talking about your computer, or phone, or if your car GPS gives you time that is off by a second or two. However, when it comes to things that depend upon precision synchronization, like orbits, it does matter.
(Score: 2) by sudo rm -rf on Tuesday June 30 2015, @05:02AM
Funny, how inconsistent sources are on the total number of leap seconds. If I understand correctly, a total of 35 seconds have already been added, because the first leap second in 1972 was in fact 10 seconds. Ah, Wiki [wikipedia.org] tends to agree.
At 7.66 km/s the ISS would already be 268 km behind (or ahead?)