[ntp:questions] DUT1 frolics

David L. Mills mills at udel.edu
Tue Jan 9 15:54:30 UTC 2007


DUT1 is the UTC correction in units of .1 second determined by the 
International Earth Rotation Service (IERTS) and disseminated to the 
public by radio, satellite and telephone. It decrements every few months 
or so. When DUT1 hunts the outskirts of -5 it increments one second and 
a leap second is declared in UTC.

While tweaking the audio reference clocks I discovered a mystery. The 
CHU broadcast signals show +0 for the DUT1 correction, as does the NIST 
modem time. However, the WWV signal last I checked showed +1. Intrigued, 
I checked the NIST web 
http://tf.nist.gov/timefreq/pubs/bulletin/leapsecond.htm and found it 
said the correction has been +1 since 28 September 2006. A polite note 
from the CHU operators said the +0 is correct, but they received notice 
only hours before it went into effect. Well, according to IERTS Bulletin 
D http://hpiers.obspm.fr/eop-pc/ dated 15 December 2006, the DUT1 was 
changed to +0 on 22 December, just in time for Christmas. Then, I 
checked WWV again and found DUT1 +0. Apparently, I must have last 
checked before 22 December. Life moves fast.

Curious, I looked for evidence why the IERTS declared when it did. 
Looking at the daily data, the predicted DUT1 crossed .05 s on the way 
down on 23 December midway between 0.1 and 0.0, so the criteria must be 
round to the nearest decisecond. The scary part is the increased rate 
DUT1 is dropping.


More information about the questions mailing list