From owner-leapsecs@ROM.USNO.NAVY.MIL Wed Feb 12 01:50:27 2003 Received: from [192.5.41.253] (juno.usno.navy.mil [192.5.41.253]) by mail.ucolick.org (8.10.2+Sun/8.10.2) with SMTP id h1C9oMB07631 for ; Wed, 12 Feb 2003 01:50:22 -0800 (PST) Received: from rom.usno.navy.mil by [192.5.41.253] via smtpd (for santo.ucolick.org [128.114.23.204]) with SMTP; 12 Feb 2003 09:56:57 UT Received: from ROM.USNO.NAVY.MIL (rom.usno.navy.mil [10.1.4.27]) by rom.usno.navy.mil (8.12.5/8.12.5) with ESMTP id h1C9oJ7m026131; Wed, 12 Feb 2003 09:50:19 GMT Received: from ROM.USNO.NAVY.MIL by ROM.USNO.NAVY.MIL (LISTSERV-TCP/IP release 1.8e) with spool id 1552 for LEAPSECS@ROM.USNO.NAVY.MIL; Wed, 12 Feb 2003 09:50:19 +0000 Received: from TS-FW.usno.navy.mil (TS-FW.usno.navy.mil [10.1.1.3]) by rom.usno.navy.mil (8.12.5/8.12.5) with SMTP id h1C9oH7k026128 for ; Wed, 12 Feb 2003 09:50:18 GMT Received: from santo.ucolick.org ([128.114.23.204]) by TS-FW.usno.navy.mil via smtpd (for rom.usno.navy.mil [10.1.4.27]) with SMTP; 12 Feb 2003 09:56:52 UT Received: from xocolatl.ucolick.org (xocolatl.ucolick.org [128.114.22.203]) by mail.ucolick.org (8.10.2+Sun/8.10.2) with ESMTP id h1C9o7B07609 for ; Wed, 12 Feb 2003 01:50:07 -0800 (PST) Received: (from sla@localhost) by xocolatl.ucolick.org (8.9.3/8.8.6) id BAA11462 for LEAPSECS@ROM.USNO.NAVY.MIL; Wed, 12 Feb 2003 01:50:07 -0800 Date: Wed, 12 Feb 2003 01:50:06 -0800 From: Steve Allen To: Leap Second Discussion List Subject: UTC vis a vis civil time Message-ID: <20030212095006.GA11451@ucolick.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.4i X-Virus: Clean, Clean Sender: owner-leapsecs@ROM.USNO.NAVY.MIL Precedence: list Status: RO Content-Length: 5115 Lines: 122 A brief tour of time UT0 is angle (but there is more than one angle involved in its values) UT1 is angle (it is the rigorous quantity for earth rotation) UT2 is angle (albeit hardly used anymore) UT is angle (in some fuzzy pre-relativistic way) Sidereal time is angle The distinction between angle and time was not understood until this past century, but by current understanding the above measures are definitely not time. GMT is not legally defined in a way that makes it easily classified as either time or angle. (Its original use tends to classify it as angle equal to UT1, and later use tends to classify it as UTC.) TAI is a practical realization of proper time, as a statistical mean of the participating clocks, reduced to the rotating geoid. (Note that the geoid is not constant over geological time, therefore the rate of TAI may be affected by lack of adequate geophysical modelling.) ET is pre-relativistic time (presumably on the rotating geoid) TT is proper time (TT = TDT). This is an ideal that should theoretically tick in unison with TAI but may not because of defects in TAI. When realized TT is a family of timescales derived from various measures, but for practical purposes TDT = TAI + 32.184 s. TDB is proper time in the frame of the barycenter that ticks in sync with the mean ticks of TT. Technically TDB is ill-posed, obsolete and should be replaced with TCB, but in practice it differs only in rate. TCG is coordinate time in the frame of the geocenter that ticks faster than mean TT because of the gravitational and rotational redshift of clocks on the geoid. TCB is coordinate time in the frame of the barycenter that ticks faster than TCG because of the gravitational and velocity redshift of clocks in the frame of the geocenter. The realization of TDB, TCG, and TCB are families like those of TT that depend on the input measure of TT. Many other forms of coordinate and proper times might be defined for observers with other velocities and depths in gravitational potentials. Recent details of the meaning of all these can be found in papers from. http://www.iers.org/iers/publications/tn/tn29/ Software implementing the gory details can be found at http://www.iau-sofa.rl.ac.uk/ ======================== Okay, so much for the tour of time. Now, to get to the subject of UTC. UTC is, really, angle. Its value has always been designed to indicate angle, although since 1972 it has been constrained to tick with TAI. As seen above, this is consistent with all other forms of "Universal" time, which have always been realized as measures of angle. To change the meaning of UTC into a form of atomic, proper, or coordinate time would create a substantially misleading misnomer capable of causing significant confusion in the context of the other definitions of time. ======================== Who is in charge of the definition of UTC? The CCIR Recommendation 460-4 (1986) says that UTC is maintained by BIPM with assistance from IERS. The CCIR is a predecessor of ITU-R, whence the document can currently be purchased. The BIPM does maintain TAI, but indicates that UTC is within the purview of IERS. IERS says that UTC is defined by CCIR Rec. 460-4 (1986). Is this arrangement an example of a happy cooperative, or tenuous circularity? To their credit, the documents about time from the IAU, IERS, and BIPM are openly available to all. The authorship of the papers and membership of the committees is visible online. The decisions are made with respect to the practicalities of existing systems. In contrast, everything from ITU-R is a closed process. ======================== It is clear that ITU-R is in charge of broadcast time and frequency, and for practical purposes this means civil time. ITU-R could redefine UTC within their document, or they could revise the document to assert that some other timescale than UTC should be broadcast. Could it happen that ITU-R might chose to redefine UTC and IERS choose not? The available documentation is not clear. If disagreement were to happen would the unambiguous meaning of UTC be destroyed? Would various national time broadcasts have to choose which of two versions of UTC to transmit? ======================== If it is deemed that civil time should become a form of atomic time, then I assert that it is more consistent and safer to define a new timescale than to change UTC. The new timescale could be called something like Civil Atomic Time (Temps Atomique Civil = TAC). This is not to say that a new name and character for civil time is the right thing to do, and it certainly does not solve all problems. Whether done by changing the character or changing the name, dropping leap seconds from civil time (and time broadcasts) will still cause grief to many existing systems, especially telescopes. It must not be done without a lead in period of many years. -- Steve Allen UCO/Lick Observatory Santa Cruz, CA 95064 sla@ucolick.org Voice: +1 831 459 3046 http://www.ucolick.org/~sla PGP: 1024/E46978C5 F6 78 D1 10 62 94 8F 2E 49 89 0E FE 26 B4 14 93