Article: 10807 of comp.protocols.time.ntp From: seaman@noao.edu (Rob Seaman) Newsgroups: comp.protocols.time.ntp,comp.std.internat Subject: negative leap seconds (was Re: Proposed New Standard Time Scale UTS) Date: 16 Nov 2000 23:25:05 GMT Organization: National Optical Astronomy Observatories Lines: 48 Distribution: world Message-ID: <8v1qch$23c5$1@noao.edu> References: <8st2li$dg2$1@pegasus.csx.cam.ac.uk> <2000Nov1209.04.55.20446@cr.yp.to> <8uoelq$egj$1@sic.twinsun.com> <3A0FCBA2.D348E274@signature.nl> <3a10b853$1@news.meer.net> NNTP-Posting-Host: bigx.tuc.noao.edu X-Trace: noao.edu 974417105 68997 140.252.1.2 (16 Nov 2000 23:25:05 GMT) X-Complaints-To: abuse@noao.edu NNTP-Posting-Date: 16 Nov 2000 23:25:05 GMT Path: news.meer.net!nntp1.ba.best.com!news2.best.com!news-hog.berkeley.edu!ucberkeley!newsfeed.stanford.edu!arclight.uoregon.edu!news.asu.edu!ennfs.eas.asu.edu!noao!noao.edu!seaman Xref: news.meer.net comp.protocols.time.ntp:10807 comp.std.internat:4704 Jeff Woolsey writes: > Repeat after me: "56... 57... 58... 00... 01..." > > UTC can do _that_, too. Actually - no. The claim is widely made that while all leap seconds to date have been positive that it is always possible that a negative leap second may occur in the future. Negative leap seconds are impossible. Well, negative leap seconds are impossible - short of a Cretaceous level event, perhaps. Consider a plot of UT1-TAI over the last quarter century. There have been about a dozen and a half leap seconds over that period. Why? Is it that the Earth's rotation has slowed by 18 seconds in a quarter century? No. Consider the slope of this roughly linear plot - about -0.7 seconds/year. Where does that slope come from? The fact that there is a leap second every year or year-and-half has nothing to do with the current deceleration of the Earth. There are leap seconds because of the hundred year's worth of deceleration that has already occurred since the 1900 epoch that defined ephemeris time. If the Earth's slowing were to halt, the leap seconds would continue to arrive one after another. The UT1-TAI plot would continue to have a slope. Try to imagine the cataclism necessary to not only reverse this year's tidal braking of the Earth - but necessary to reverse a century's worth of accumulated tidal braking. I seriously question whether even tossing another asteroid at the Yucatan would deliver enough total impulse during the fractional second impact to counterbalance the much larger moon's total effect over a century. What this year's accumulated slowing does effect is future leap seconds. The Earth is spinning down by an amount approximately equivalent to 1 leap second per year per century. In the year 2100, there will be a leap second about every 6 or 8 months. And it would require even more petatonnage to potentially jolt the Earth sufficiently to generate the need for a negative leap second. So yes - there may be a negative leap second in our future - but no, we won't be here to see it. Rob Seaman National Optical Astronomy Observatory. Article: 10816 of comp.protocols.time.ntp From: nmm1@cus.cam.ac.uk (Nick Maclaren) Newsgroups: comp.protocols.time.ntp,comp.std.internat Subject: Re: negative leap seconds (was Re: Proposed New Standard Time Scale UTS) Date: 17 Nov 2000 09:54:27 GMT Organization: University of Cambridge, England Lines: 30 Message-ID: <8v2v8j$5oa$1@pegasus.csx.cam.ac.uk> References: <8st2li$dg2$1@pegasus.csx.cam.ac.uk> <3a10b853$1@news.meer.net> <8v1qch$23c5$1@noao.edu> <3A14DA7C.D0A6F324@hda.hydro.com> NNTP-Posting-Host: libra.cus.cam.ac.uk Path: news.meer.net!nntp1.ba.best.com!news1.best.com!newsfeed.mathworks.com!btnet-peer!btnet!news.vas-net.net!server2.netnews.ja.net!pegasus.csx.cam.ac.uk!nmm1 Xref: news.meer.net comp.protocols.time.ntp:10816 comp.std.internat:4711 In article <3A14DA7C.D0A6F324@hda.hydro.com>, Terje Mathisen wrote: >Rob Seaman wrote: >[snipped good description of why a negative leap second is extremely >unlikely] >> So yes - there may be a negative leap second in our future - but no, >> we won't be here to see it. > >There seems to be just one interesting way to get a negative leap >second: If the distribution of the core material of the Earth could >change, then the rotation rate could also change significantly. Actually, there are several others. It has never been proven that the physical constants are invariant, and some physicists have hypothesised that they may not be. >(I agree absolutely that in the case of a meteor impact which is big >enough to speed up the rotation rate significantly, human beings will >have a tough job simply surviving. Leap second adjustments would >probably not be high on the priority list of any survivors. :-() I think that the same applies in the example I mentioned :-) Regards, Nick Maclaren, University of Cambridge Computing Service, New Museums Site, Pembroke Street, Cambridge CB2 3QG, England. Email: nmm1@cam.ac.uk Tel.: +44 1223 334761 Fax: +44 1223 334679 Article: 10814 of comp.protocols.time.ntp From: Paul Keinanen Newsgroups: comp.protocols.time.ntp,comp.std.internat Subject: Re: negative leap seconds (was Re: Proposed New Standard Time Scale UTS) Date: Fri, 17 Nov 2000 10:42:22 +0200 Organization: SAUNALAHDEN asiakas Lines: 39 Message-ID: References: <8st2li$dg2$1@pegasus.csx.cam.ac.uk> <2000Nov1209.04.55.20446@cr.yp.to> <8uoelq$egj$1@sic.twinsun.com> <3A0FCBA2.D348E274@signature.nl> <3a10b853$1@news.meer.net> <8v1qch$23c5$1@noao.edu> NNTP-Posting-Host: dccl.hdyn.saunalahti.fi Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: tron.sci.fi 974450316 13848 195.74.25.150 (17 Nov 2000 08:38:36 GMT) X-Complaints-To: newsmaster@saunalahti.fi NNTP-Posting-Date: 17 Nov 2000 08:38:36 GMT X-Newsreader: Forte Agent 1.7/32.534 Path: news.meer.net!nntp1.ba.best.com!news1.best.com!nntp.primenet.com!nntp.gblx.net!newsfeed.icl.net!news.algonet.se!newsfeed1.telenordia.se!algonet!newsfeed1.funet.fi!newsfeeds.funet.fi!news.cc.tut.fi!uutiset.saunalahti.fi!not-for-mail Xref: news.meer.net comp.protocols.time.ntp:10814 comp.std.internat:4710 On 16 Nov 2000 23:25:05 GMT, seaman@noao.edu (Rob Seaman) wrote: >The claim is widely made that while all leap seconds to date have been >positive that it is always possible that a negative leap second may >occur in the future. > >Negative leap seconds are impossible. Well, negative leap seconds are >impossible - short of a Cretaceous level event, perhaps. You should look at a longer perspective. According to Meeus: Astronomical Algorithms (2nd ed. p. 79), the TDT-UT difference was +121 s in the year 1650, falling to a local minimum of +7 s in 1700, corresponding to a slope of -2.3 s/year. There is a local maximum of +16 s around 1785, thus a slope of +0.1 s/year. A minimum of -6.5 s in 1894, thus a slope of -0.2 s/year. In 1998 the difference was 63.0 s, thus an average slope of +0.67 s/year. Since the definition of the atomic second was originally derived from the length of the year 1899, which is close to the 1894 minimum, so there is not a great wonder that all the leap seconds have been positive so far. Defining the atomic second based on some other year e.g. in the 1860 .. 1910 era would have produced quite different leap second events. >Try to imagine the cataclism necessary to not only reverse this year's >tidal braking of the Earth - but necessary to reverse a century's >worth of accumulated tidal braking. I seriously question whether >even tossing another asteroid at the Yucatan would deliver enough total >impulse during the fractional second impact to counterbalance the much >larger moon's total effect over a century. The Earth is not a completely solid body as the Moon is. The solid core is floating around below our feet, causing all kinds of wobbling phenomenon. Article: 10823 of comp.protocols.time.ntp From: seaman@noao.edu (Rob Seaman) Newsgroups: comp.protocols.time.ntp,comp.std.internat Subject: Re: negative leap seconds (was Re: Proposed New Standard Time Scale UTS) Date: 17 Nov 2000 18:54:57 GMT Organization: National Optical Astronomy Observatories Lines: 106 Distribution: world Message-ID: <8v3uu1$1j81$1@noao.edu> References: <8st2li$dg2$1@pegasus.csx.cam.ac.uk> <2000Nov1209.04.55.20446@cr.yp.to> <8uoelq$egj$1@sic.twinsun.com> <3A0FCBA2.D348E274@signature.nl> <3a10b853$1@news.meer.net> <8v1qch$23c5$1@noao.edu> NNTP-Posting-Host: bigx.tuc.noao.edu X-Trace: noao.edu 974487297 52481 140.252.1.2 (17 Nov 2000 18:54:57 GMT) X-Complaints-To: abuse@noao.edu NNTP-Posting-Date: 17 Nov 2000 18:54:57 GMT Path: news.meer.net!nntp1.ba.best.com!news1.best.com!newsfeed.mathworks.com!arclight.uoregon.edu!news.asu.edu!ennfs.eas.asu.edu!noao!noao.edu!seaman Xref: news.meer.net comp.protocols.time.ntp:10823 comp.std.internat:4716 Nick Maclaren comments on Terje Mathisen's reply to my assertion: >>> So yes - there may be a negative leap second in our future - but no, >>> we won't be here to see it. >> There seems to be just one interesting way to get a negative leap >> second: If the distribution of the core material of the Earth could >> change, then the rotation rate could also change significantly. > Actually, there are several others. It has never been proven that > the physical constants are invariant, and some physicists have > hypothesised that they may not be. The most likely effect humanity can have on the Earth's rotation rate is through the high latitude impoundment of water in reservoirs or the polar ice caps. A naive analysis suggests that global warming will increase the Earth's rotational braking (deeper equatorial oceans for the Moon to grab onto). The slowing trend should be accelerated until we cool down again. Paul Keinanen cautions: > You should look at a longer perspective. > > According to Meeus: Astronomical Algorithms (2nd ed. p. 79), the > TDT-UT difference was +121 s in the year 1650, falling to a local > minimum of +7 s in 1700, corresponding to a slope of -2.3 s/year. > There is a local maximum of +16 s around 1785, thus a slope of +0.1 > s/year. A minimum of -6.5 s in 1894, thus a slope of -0.2 s/year. In > 1998 the difference was 63.0 s, thus an average slope of +0.67 s/year. A little Meeus goes a long way. I might suggest consulting the original table in the Astronomical Almanac - pages K8-9 in the 2001 edition (if only for the improved typography :-) A longer perspective is good. Dennis McCarthy and William Klepczynski's article "GPS and Leap Seconds - Time to Change?" in the November 1999 "GPS World" includes a nice graph (with error bars) of the delta-T data from 1600 to the present. This includes an upward sweeping parabolic fit and nicely shows the local minimum in the late 19th century. Yes - I did indeed simplify the situation. It is handy that the ET epoch happens to lie near the base of a local minimum - this is not necessary, however, to produce the trend of positive leap seconds. Currently the length of the day is longer than the length of the day at the reference epoch. That's all that's needed. In terms of an absolute physical effect this amounts to the length of the day currently being about 0.2 milliseconds longer than in 1900. That is - every 500 days or so we accumulate enough of an additional "time bias" to issue another leap second. Note that McCarthy and Klepczynski's article also suggests that the leap second mechanism requires an overhaul. I hope Posix and future computer instantiations of civil time take this into effect. The suggested overhaul that has received the most discussion is to drop leap seconds entirely and allow UTC and UT1 to diverge. (Needless to say I find this suggestion absurd.) The mailing list that was formed to discuss this topic is currently offline: http://clockdev.usno.navy.mil/archives/leapsecs.html The USNO/IERS predictions for the next decade or so are available from: ftp://maia.usno.navy.mil/ser7/deltat.preds The trend continues. The USNO also offers its own discussion parallel to my arguments: "Confusion sometimes arises over the misconception that the regular insertion of leap seconds every few years indicates that the Earth should stop rotating within a few millennia. The confusion arises because some mistake leap seconds as a measure of the rate at which the Earth is slowing. The one-second increments are, however, indications of the accumulated difference in time between the two systems. As an example, the situation is similar to what would happen if a person owned a watch that lost two seconds per day. If it were set to a perfect clock today, the watch would be found to be slow by two seconds tomorrow. At the end of a month, the watch will be roughly a minute in error (thirty days of the two second error accumulated each day). The person would then find it convenient to reset the watch by one minute to have the correct time again." (From http://maia.usno.navy.mil/eo/leapsec.html.) I presume that they regularly get questions along the lines of: "If we're losing a leap second per year, won't we run out of time in 86,400 years?" :-) (Answer - no, but in 43,200 years it would be midnight at noon without leap seconds.) > Since the definition of the atomic second was originally derived from > the length of the year 1899, which is close to the 1894 minimum, so > there is not a great wonder that all the leap seconds have been positive > so far. Defining the atomic second based on some other year e.g. in the > 1860 .. 1910 era would have produced quite different leap second events. Meeus's table is not directly connected to the length of the year and his use of time is rather ad hoc. The TT-UT differences that are the current standard are more explicitly patched together with prior standards such as TDT and ET in the Almanac. Yes, there are wiggles superimposed on the general slowing trend, but these are wiggles in the zero point alignment of two time scales, not in the absolute duration of the year or the day - that was the entire point. Rob Seaman National Optical Astronomy Observatory