ddhr.org - Timekeeping

Date: 2006 | 01 | 28

Timekeeping (1)

Jan 28, 2006

On December 31, 2005, a leap second was added to the Coordinated Universal Time (atomic time) so that it would be in sync with Greenwich Mean Time (astronomical time).

This article talks about how certain scientists are proposing to get rid of the leap second because it could cause some major computer problems down the road. The reason the leap second needs to be added in the first place is because astronomical time is based on the movements of the Earth, and the Earth isn't perfect: Its days and years vary in length. Atomic time is based on the vibrations of the Cesium-133 atom, which doesn't change. So in order to keep the two methods of timekeeping in sync, the International Telecommunication Union decided that UTC couldn't differ from GMT by more than 9 tenths of a second. So every once in a while, a leap second is added so that everything stays in sync.

(via Boing Boing) #technology

Comments

Rich

Jan 30, 2006

I had read about this a while ago, and once again it just strikes me that we are worrying about fly turds in the pepper. First of all, the basic premise here that astronomical time (by which I mean the time that astronomers use) and standard time (by which I mean the time that everyone else uses (Coordinated Universal whatever...)) need to be somehow correlated, and that we need intermittant leap-seconds to stay within 0.9 seconds of each other. Um... why? As it is, siderial time goes on its merry way, completely separate from standard time. That tends to happen when your day has 23 hours 56 minutes instead of 24. It seems kinda wacky, but it makes a lot of sense for astronomy, since siderial time is defined by the (apparent) movement of the stars, and can be used very simply in straightfoward calculations to determine where various celestial bodies will appear. It doesn't need to have anything to do with the alarm clocks in every American's home. Who came up with 0.9 seconds anyway? Its arbitrary. We could have said 1E-12 seconds, and require leap femtoseconds on the order of every day. OR. More reasonably, we could say that since the correlation doesn't really matter, but we don't want midnight in the middle of the day in the year 5000, we make a bigger correction every couple decades or so. It's a matter of mixing practicality with technicality. Midnight in the middle of the day? Yea, that would mess people up. Electronics could pretty much care less, though. And for the guy who said that all these leap-seconds would mess up all our computers and bring on a cyber apocalypse, I'll just cite the same example he used: y2k. Some poorly designed software caused some inattentive people some trouble, but in general was there a problem? No. As long as there is some rule about how leap-seconds are applied, and it is implemented without cutting corners, there won't be a problem.
Oh, and for the record. Some pulsars are so regular that they could provide precision an order of magnitude better than cesium 133.
I also like how they mentioned gps before they mentioned atomic clocks. Again, for the record, GPS is only feasible because of the precision offered by atomic clocks. That's a story for another day, but really, it's kinda amazing it works at all. The principle is simple, and anyone who has had high school geometry can do the math, but you won't get even close to the right answer unless you have amazingly precise numbers to work with. These clocks are accurate to 1 second in 20 million years, but they still can only really place you to within 3 feet on the earth's surface. (Get cell towers in on the action, and you can zoom in to an inch, but that's another story alltogether.)
There I go again. I think my rebuttal might have been longer than the article. If you bothered to read it all, good on you.