[LEAPSECS] Crunching Bulletin B numbers

[Mark Calabretta]

On Thu 2011/02/17 10:52:26 +1100, Mark Calabretta wrote
in a message to: Leap Second Discussion List <leapsecs at leapsecond.com>


>That said, if leap second insertions were simply deferred for 10

>years, DUT1 would probably grow to no more than about 6s (even

>including deceleration), which seems much preferable to letting it

>grow without limit.


Assuming 6 leap seconds per decade as the long-term average rate
since 1972, the simple expedient of deferring leap second insertion
by 10 years would have constrained DUT1 to the range -3 to +5 as
shown below.

It shouldn't be hard to do better than this, even if only by adjusting
the decadal rate to accord with the obvious decadal trend (column N).
The decadal rate would be applied by predicting the required number of
leap seconds within the next decade.

Regards,
Mark Calabretta


    Date
 ----------
 1972/01/01
 1972/07/01
 1973/01/01
 1974/01/01
 1975/01/01
 1976/01/01
 1977/01/01
 1978/01/01
 1979/01/01
 1980/01/01    Date+10    N   N-6
 1981/07/01  ----------  --   ---
             1982/01/01  11    5
 1982/07/01  1982/07/01  10    4
             1983/01/01  10    4
 1983/07/01
             1984/01/01  10    4
             1985/01/01   9    3
 1985/07/01
             1986/01/01   9    3
             1987/01/01   8    2
 1988/01/01  1988/01/01   8    2
             1989/01/01   7    1
 1990/01/01  1990/01/01   7    1
 1991/01/01
             1991/07/01   7    1
 1992/07/01  1992/07/01   7    1
 1993/07/01  1993/07/01   7    1
 1994/07/01
             1995/07/01   7    1
 1996/01/01
 1997/07/01
             1998/01/01   8    2
 1999/01/01
             2000/01/01   8    2
             2001/01/01   7    1
             2002/07/01   6    0
             2003/07/01   5   -1
             2004/07/01   4   -2
 2006/01/01  2006/01/01   4   -2
             2007/07/01   3   -3
 2009/01/01  2009/01/01   3   -3
             2016/01/01   ?    ?
             2019/01/01   ?    ?

- Date = historical date of leap second insertion
- Date+10 = Date if leap second deferred by 10 years
- N = count of leap seconds in the preceding 10 years (inclusive)
- Last column is N - 6, where 6 is the long term average decadal
  count of leap seconds.