xref: /freebsd/usr.sbin/ntp/ntpd/leap-seconds (revision dd41de95a84d979615a2ef11df6850622bf6184e)
1#
2#	In the following text, the symbol '#' introduces
3#	a comment, which continues from that symbol until
4#	the end of the line. A plain comment line has a
5#	whitespace character following the comment indicator.
6#	There are also special comment lines defined below.
7#	A special comment will always have a non-whitespace
8#	character in column 2.
9#
10#	A blank line should be ignored.
11#
12#	The following table shows the corrections that must
13#	be applied to compute International Atomic Time (TAI)
14#	from the Coordinated Universal Time (UTC) values that
15#	are transmitted by almost all time services.
16#
17#	The first column shows an epoch as a number of seconds
18#	since 1 January 1900, 00:00:00 (1900.0 is also used to
19#	indicate the same epoch.) Both of these time stamp formats
20#	ignore the complexities of the time scales that were
21#	used before the current definition of UTC at the start
22#	of 1972. (See note 3 below.)
23#	The second column shows the number of seconds that
24#	must be added to UTC to compute TAI for any timestamp
25#	at or after that epoch. The value on each line is
26#	valid from the indicated initial instant until the
27#	epoch given on the next one or indefinitely into the
28#	future if there is no next line.
29#	(The comment on each line shows the representation of
30#	the corresponding initial epoch in the usual
31#	day-month-year format. The epoch always begins at
32#	00:00:00 UTC on the indicated day. See Note 5 below.)
33#
34#	Important notes:
35#
36#	1. Coordinated Universal Time (UTC) is often referred to
37#	as Greenwich Mean Time (GMT). The GMT time scale is no
38#	longer used, and the use of GMT to designate UTC is
39#	discouraged.
40#
41#	2. The UTC time scale is realized by many national
42#	laboratories and timing centers. Each laboratory
43#	identifies its realization with its name: Thus
44#	UTC(NIST), UTC(USNO), etc. The differences among
45#	these different realizations are typically on the
46#	order of a few nanoseconds (i.e., 0.000 000 00x s)
47#	and can be ignored for many purposes. These differences
48#	are tabulated in Circular T, which is published monthly
49#	by the International Bureau of Weights and Measures
50#	(BIPM). See www.bipm.org for more information.
51#
52#	3. The current definition of the relationship between UTC
53#	and TAI dates from 1 January 1972. A number of different
54#	time scales were in use before that epoch, and it can be
55#	quite difficult to compute precise timestamps and time
56#	intervals in those "prehistoric" days. For more information,
57#	consult:
58#
59#		The Explanatory Supplement to the Astronomical
60#		Ephemeris.
61#	or
62#		Terry Quinn, "The BIPM and the Accurate Measurement
63#		of Time," Proc. of the IEEE, Vol. 79, pp. 894-905,
64#		July, 1991. <http://dx.doi.org/10.1109/5.84965>
65#		reprinted in:
66#		   Christine Hackman and Donald B Sullivan (eds.)
67#		   Time and Frequency Measurement
68#		   American Association of Physics Teachers (1996)
69#		   <http://tf.nist.gov/general/pdf/1168.pdf>, pp. 75-86
70#
71#	4. The decision to insert a leap second into UTC is currently
72#	the responsibility of the International Earth Rotation and
73#	Reference Systems Service. (The name was changed from the
74#	International Earth Rotation Service, but the acronym IERS
75#	is still used.)
76#
77#	Leap seconds are announced by the IERS in its Bulletin C.
78#
79#	See www.iers.org for more details.
80#
81#	Every national laboratory and timing center uses the
82#	data from the BIPM and the IERS to construct UTC(lab),
83#	their local realization of UTC.
84#
85#	Although the definition also includes the possibility
86#	of dropping seconds ("negative" leap seconds), this has
87#	never been done and is unlikely to be necessary in the
88#	foreseeable future.
89#
90#	5. If your system keeps time as the number of seconds since
91#	some epoch (e.g., NTP timestamps), then the algorithm for
92#	assigning a UTC time stamp to an event that happens during a positive
93#	leap second is not well defined. The official name of that leap
94#	second is 23:59:60, but there is no way of representing that time
95#	in these systems.
96#	Many systems of this type effectively stop the system clock for
97#	one second during the leap second and use a time that is equivalent
98#	to 23:59:59 UTC twice. For these systems, the corresponding TAI
99#	timestamp would be obtained by advancing to the next entry in the
100#	following table when the time equivalent to 23:59:59 UTC
101#	is used for the second time. Thus the leap second which
102#	occurred on 30 June 1972 at 23:59:59 UTC would have TAI
103#	timestamps computed as follows:
104#
105#	...
106#	30 June 1972 23:59:59 (2287785599, first time):	TAI= UTC + 10 seconds
107#	30 June 1972 23:59:60 (2287785599,second time):	TAI= UTC + 11 seconds
108#	1  July 1972 00:00:00 (2287785600)		TAI= UTC + 11 seconds
109#	...
110#
111#	If your system realizes the leap second by repeating 00:00:00 UTC twice
112#	(this is possible but not usual), then the advance to the next entry
113#	in the table must occur the second time that a time equivalent to
114#	00:00:00 UTC is used. Thus, using the same example as above:
115#
116#	...
117#       30 June 1972 23:59:59 (2287785599):		TAI= UTC + 10 seconds
118#       30 June 1972 23:59:60 (2287785600, first time):	TAI= UTC + 10 seconds
119#       1  July 1972 00:00:00 (2287785600,second time):	TAI= UTC + 11 seconds
120#	...
121#
122#	in both cases the use of timestamps based on TAI produces a smooth
123#	time scale with no discontinuity in the time interval. However,
124#	although the long-term behavior of the time scale is correct in both
125#	methods, the second method is technically not correct because it adds
126#	the extra second to the wrong day.
127#
128#	This complexity would not be needed for negative leap seconds (if they
129#	are ever used). The UTC time would skip 23:59:59 and advance from
130#	23:59:58 to 00:00:00 in that case. The TAI offset would decrease by
131#	1 second at the same instant. This is a much easier situation to deal
132#	with, since the difficulty of unambiguously representing the epoch
133#	during the leap second does not arise.
134#
135#	Some systems implement leap seconds by amortizing the leap second
136#	over the last few minutes of the day. The frequency of the local
137#	clock is decreased (or increased) to realize the positive (or
138#	negative) leap second. This method removes the time step described
139#	above. Although the long-term behavior of the time scale is correct
140#	in this case, this method introduces an error during the adjustment
141#	period both in time and in frequency with respect to the official
142#	definition of UTC.
143#
144#	Questions or comments to:
145#		Judah Levine
146#		Time and Frequency Division
147#		NIST
148#		Boulder, Colorado
149#		Judah.Levine@nist.gov
150#
151#	Last Update of leap second values:   8 July 2016
152#
153#	The following line shows this last update date in NTP timestamp
154#	format. This is the date on which the most recent change to
155#	the leap second data was added to the file. This line can
156#	be identified by the unique pair of characters in the first two
157#	columns as shown below.
158#
159#$	 3676924800
160#
161#	The NTP timestamps are in units of seconds since the NTP epoch,
162#	which is 1 January 1900, 00:00:00. The Modified Julian Day number
163#	corresponding to the NTP time stamp, X, can be computed as
164#
165#	X/86400 + 15020
166#
167#	where the first term converts seconds to days and the second
168#	term adds the MJD corresponding to the time origin defined above.
169#	The integer portion of the result is the integer MJD for that
170#	day, and any remainder is the time of day, expressed as the
171#	fraction of the day since 0 hours UTC. The conversion from day
172#	fraction to seconds or to hours, minutes, and seconds may involve
173#	rounding or truncation, depending on the method used in the
174#	computation.
175#
176#	The data in this file will be updated periodically as new leap
177#	seconds are announced. In addition to being entered on the line
178#	above, the update time (in NTP format) will be added to the basic
179#	file name leap-seconds to form the name leap-seconds.<NTP TIME>.
180#	In addition, the generic name leap-seconds.list will always point to
181#	the most recent version of the file.
182#
183#	This update procedure will be performed only when a new leap second
184#	is announced.
185#
186#	The following entry specifies the expiration date of the data
187#	in this file in units of seconds since the origin at the instant
188#	1 January 1900, 00:00:00. This expiration date will be changed
189#	at least twice per year whether or not a new leap second is
190#	announced. These semi-annual changes will be made no later
191#	than 1 June and 1 December of each year to indicate what
192#	action (if any) is to be taken on 30 June and 31 December,
193#	respectively. (These are the customary effective dates for new
194#	leap seconds.) This expiration date will be identified by a
195#	unique pair of characters in columns 1 and 2 as shown below.
196#	In the unlikely event that a leap second is announced with an
197#	effective date other than 30 June or 31 December, then this
198#	file will be edited to include that leap second as soon as it is
199#	announced or at least one month before the effective date
200#	(whichever is later).
201#	If an announcement by the IERS specifies that no leap second is
202#	scheduled, then only the expiration date of the file will
203#	be advanced to show that the information in the file is still
204#	current -- the update time stamp, the data and the name of the file
205#	will not change.
206#
207#	Updated through IERS Bulletin C60
208#	File expires on:  28 June 2021
209#
210#@	3833827200
211#
2122272060800	10	# 1 Jan 1972
2132287785600	11	# 1 Jul 1972
2142303683200	12	# 1 Jan 1973
2152335219200	13	# 1 Jan 1974
2162366755200	14	# 1 Jan 1975
2172398291200	15	# 1 Jan 1976
2182429913600	16	# 1 Jan 1977
2192461449600	17	# 1 Jan 1978
2202492985600	18	# 1 Jan 1979
2212524521600	19	# 1 Jan 1980
2222571782400	20	# 1 Jul 1981
2232603318400	21	# 1 Jul 1982
2242634854400	22	# 1 Jul 1983
2252698012800	23	# 1 Jul 1985
2262776982400	24	# 1 Jan 1988
2272840140800	25	# 1 Jan 1990
2282871676800	26	# 1 Jan 1991
2292918937600	27	# 1 Jul 1992
2302950473600	28	# 1 Jul 1993
2312982009600	29	# 1 Jul 1994
2323029443200	30	# 1 Jan 1996
2333076704000	31	# 1 Jul 1997
2343124137600	32	# 1 Jan 1999
2353345062400	33	# 1 Jan 2006
2363439756800	34	# 1 Jan 2009
2373550089600	35	# 1 Jul 2012
2383644697600	36	# 1 Jul 2015
2393692217600	37	# 1 Jan 2017
240#
241#	the following special comment contains the
242#	hash value of the data in this file computed
243#	use the secure hash algorithm as specified
244#	by FIPS 180-1. See the files in ~/pub/sha for
245#	the details of how this hash value is
246#	computed. Note that the hash computation
247#	ignores comments and whitespace characters
248#	in data lines. It includes the NTP values
249#	of both the last modification time and the
250#	expiration time of the file, but not the
251#	white space on those lines.
252#	the hash line is also ignored in the
253#	computation.
254#
255#h 	064356a8 39268b92 76e4d5ef 3e22fae1 0cca529c
256