[LEAPSECS] Leap smear

[Warner Losh]

On Sep 20, 2011, at 10:13 AM, Greg Hennessy wrote:


>> Yes, indeed, if we allowed leap seconds to build up at the current

>> rate, there would be an issue with an additional hour being required

>> in the range of timezones in approximately five and a half thousand

>> years' time (one leap second per eighteen months, times 3600 seconds

>> per hour).  

> 

> Of course leap seconds are expected to build up much more rapidly than

> that, they are expected to increase quadradically.


Of course.  You can see the progress in http://www.ucolick.org/~sla/leapsecs/dutc.html for various acceleration rates.

Numbers turn out to be closer to 500-700 years for a 30 minute offset and 800-1000 for a 1 hour offset.  5k years out would be a ~8-12 hour delta.

The reason for the ranges is that we don't known which numbers to use for long term projects.  Recent history shows a much faster rate of change than the long term average from 1620 when good records first become available.  The long-term records from events like eclipses, though less accurate, show a lower rate of change. (25.6s/century^2 vs 42s/century^2).

To put things in perspective, ~700AD to the present has accumulated 1 hour.

Warner