[LEAPSECS] some old leap second history

[Tom Van Baak]

Given the question about leap seconds and WWVB, I ran
across some old NIST documents that shed light on the
evolution of broadcast time services.


Dated 1965,
NBS Low-Frequency Station WWVB to Broadcast International Unit of Time
http://tf.boulder.nist.gov/general/pdf/1721.pdf

This (and a number of other documents of the era) mention
the transition from a UT-based time/frequency broadcast to
the modern atomic time/frequency standard to respond to
the needs of their "customers".

Of course this resulted in gradual time drift wrt UT and so
the solution at the time was to shift time by 200 ms about
every 3 months (which is 12000 60 kHz cycles) -- what
could be called a leap fifth second.


Dated 1966,
WWVB Adds Time Code to Broadcast
http://tf.boulder.nist.gov/general/pdf/1731.pdf

Here we see when the time code was first added to WWVB.
Note back in those days the UT2 correction was broadcast to
millisecond resolution and the 3 digit BCD field was updated
daily.

Later those 12 bits were reallocated to today's format where
2 of those 3 BCD digits are used for the year and the UT1
correction gets just 1 BCD digit and a resolution of 100 ms.


Dated 1972,
NBS Frequency and Time Broadcast Services; Radio Stations
WWV, WWVH, WWVB, WWVL
http://tf.boulder.nist.gov/general/pdf/1545.pdf

This larger document has more history and time code details
for all NBS stations. Note the first mention of leap seconds on
page 9 of the PDF, as international time scales switched from
the old rate step model (multiples of 50 parts in 10^10) to the
new time step model (leap seconds).


/tvb