NavList:

In another thread , Sean wrote:
"So, what are the *benefits* of dropping leap seconds?
I seem to recall hearing once before that this was being considered, but for the life of me, I can't understand why you'd want to do it."

WHY do we have leap seconds in the first place? Many people will tell you that it is because the Earth's rate of rotation is slowing down. But that isn't quite right. The fact that the Earth's rate of rotation is slowing down tells us that we will have MORE leap seconds per year in the future, and eventually, after just a few centuries, leap seconds will be ridiculously frequent.

But why do we have leap seconds now? The answer is that "science" has defined one second in laboratory terms, without reference to the celestial sphere and the Earth's orientation, since 1960. As the Wikipedia article on the second aptly describes it, "the 1960 SI definition abandoned any explicit relationship between the scientific second and the length of a day, as most people understand the term." For a little over ten years, this disconnect between the established definition of the second and the actual length of a single day was small enough to ignore. Seconds in the calendar were allowed to change length very slightly. They were not "true" SI seconds by the definition adopted in 1960. Leap seconds were added to the calendar to "fix" this discrepancy. The Earth rotates at a rate which does not yield exactly 86400 true SI seconds in one rotation (mean solar). So the net count of seconds in the calendar slowly falls behind. And after about two years, it adds up to one second. That's where leap seconds come in. Through the bizarre trick of allowing a minute with 61 seconds, we bring the calendar, as determined by the Earth's orientation, back in line with the total number of counted SI seconds.

So what would happen if we drop leap seconds? First, it would allow us to answer some questions unambiguously. For example, what is the total elapsed time in seconds between 12:00 noon exactly today, 2 September 2013, and some future date, for example 12:00 noon exactly, 2 September 2023 (bot UTC of cour. You can't answer that question today though you can make a good guess within a couple of seconds. If we drop leap seconds, then the answer to this question is simple: count the number of days (3650+2 = ten years plus two leap says) and multiply by 86400 seconds in one day (= 24*60*60). If leap seconds continue as they are, with announcements declared as late as Spring 2023 for a leap second at the end of June, 2023, the actual interval between those two specified instants of the calendar cannot be calculated until roughly six months before the final date. Calculations like these are ubiquitous today. Just Google "Javascript and UTC". You will get millions of hits.

There's plenty more to the issue. I'm just getting the ball rolling...

-FER

----------------------------------------------------------------
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com
----------------------------------------------------------------