UTC might be redefined without Leap Seconds

After over a decade of discussions the ITU-R and other groups have not managed to come to a consensus on whether or not radio broadcast time signals should continue to have leap seconds. As seen in the following linked pages, a problem is that simply omitting leap seconds would redefine the meaning of the word "day" so that it is not related to the sun in the sky, nor connected with the rotation of the earth.

In 1970 the CCIR (predecessor of the ITU-R) decided to disconnect clocks from the rotation of the earth, but they kept the calendar connected to the rotation of the earth. That decision was implemented starting in 1972, and since then the leap seconds have maintained the connection.

In 2015 the ITU-R decided not to decide whether the calendar will also become disconnected from the rotation of the earth. There is currently no open question about leap seconds at the ITU-R, so they will take no further action until 2023. The request from the ITU-R at WRC-15 was for a bunch of external international agencies (none of which is involved in setting standards for computing systems) to try to forge some agreement which the ITU-R might be able to approve in 2023. At least until that time leap seconds will continue.

There are basically two options available:

if UTC retains leap seconds

one calendar day counts one turn of the earth on its axis with respect to the sun

if UTC abandons leap seconds

one calendar day counts 794 243 384 928 000 hyperfine oscillations of cesium-133

Three pictures that show the reasons for leap seconds

A brief and pictorial page showing how our clocks are affected by leap seconds, and why there is no schedule for them.

time scales have a choice between three goals

A time scale can pick two of three.
But which two? A look at where the decision processes went wrong.

timing systems that always refused to use leap seconds

One person who suggested the idea of leap seconds also pointed out that leap seconds would cause problems for automated systems. Before the first leap second happened he explained that USNO time systems would not use leap seconds.

When the second became unrelated to the day

These documents contain transcriptions from the 1955 meeting where one second was redefined to have a duration which is unrelated to the duration of one day. From this point on time became split between the traditional meaning that has always been in the calendar and the technical meaning needed for precision systems.

Elapsed time

A javascript program showing the international problem that POSIX is not currently capable of providing an interface that properly describes time, and that international agreements confound the process of fixing that problem.

correct precision handling of leap seconds using code already on POSIX systems

A prescription that shows how existing deployed and tested code could allow computing and telecommunications systems to handle leap seconds without difficulty, but only if the ITU-R changes the name of the broadcast time scale.

Differences between time scales over 4000 years

Plots showing how time keeping got the way it is, and suggesting a way to reduce the issues associated with leap seconds which is compatible with human history as well as modern systems for navigation, telecomms, and POSIX.

Differences between time scales at year 2100

Plots with estimates of how far off the calendar and clock might drift by the year 2100 if leap seconds are abandoned in the year 2020.

calendar of events and gathering of references

This document gives a picture of how the ITU-R has spent over ten years on redefining UTC without coming to a conclusion. It includes hundreds of links to almost all on-line material pertaining to the future of leap seconds. This is the document which was originally visible at this URL.

Figures and tables showing the need for leap seconds

These include plots of how the rotation of the earth slows down and speeds up, and other material on the proposed changes.

A brief history of time scales

During the past century or so the concept of time has changed considerably. The results of struggling with ever more refined notions of time have been some two dozen new ways of expressing time.

Two kinds of time are necessary

In 1954 D.H. Sadler described the need and implication of two kinds of time. His monograph is clearer than any other writeup for the next 60 years.

What could happen if civil time becomes atomic time?

This question was treated at the meeting Requirements for UTC and Civil Timekeeping on Earth held during 2013 May in Charlottesville VA USA. The presentations at that meeting resulted in 350 pages of proceedings (also available on CDROM).

Previously this question was treated at the meeting Decoupling Civil Timekeeping from Earth Rotation held during 2011 October in Exton PA USA. The presentations at that meeting resulted in 400 pages of proceedings (also available on CDROM).

Previously, the ITU-R convened a Colloquium on the UTC Time Scale in Torino Italy during 2003 May. At the conclusion of that meeting the invited experts suggested that a solution to the problems of leap seconds in the radio broadcast time scale would be to create a new atomic timescale named TI in about the year 2022. Subsequent to that meeting the delegates to the ITU-R seem to have rejected or ignored the advice given to them by the experts who attended their meeting.

Everything below here is old information.
Please focus on the links at the top of the page.
The information below remains here largely for the sake of URL permanence.

Optical and Infrared Observatories

Study of the bibliography above indicates that no formal survey has ever been performed specifically to ask optical and infrared observatories to consider the schedule and cost implications of discontinuing leap seconds from UTC. The initial URSI survey was published very broadly, and attempts were made to contact observatories. Nevertheless, I believe that this survey was inadequate for several reasons:

• The survey was performed on behalf of URSI Commission J, which is Radio Astronomy.
• The survey was performed before the nature of the cost implications of discontinuing leap seconds had been considered.
• The survey was performed before it had become clear that those who were suggesting changes to UTC were either serious or capable of doing it.

Some older optical and infrared telescopes were built or designed before Stoyko had confirmed that the length of day varied seasonally. Others were not designed to require pointing accuracy of 1 second of time. Often this was because they are manually pointed by a full-time telescope technician -- not by the observer, not by software. Telescopes such as these have never required their control systems to distinguish between UT1 and UTC. Discontinuing leap seconds requires a fundamental change in their operational procedures.

Astronomers should note in particular that the agenda for the meeting in Torino considered financial aspects (costs and opportunities) for several disciplines, but not for astronomy.

A survey of schedule and cost consequences for observatories

A description of costs and consequences

This is an introduction to the survey below.

The survey about schedule and cost

This is an attempt to gather some estimates of the scheduled time and costs which observatories will have to expend if leap seconds are discontinued.

The results of the survey

Here are the data gathered in response to the above survey. The results are sparse. This may be because it is difficult to answer the questions without expending most of the effort required to identify the deficiencies, and this is amplified by the absence of a clear indication of the detailed nature of a change.

Various postings that I have made to the LEAPSECS mailing list

UTC vis-a-vis civil time (2003-02-12) local text/plain, Mail Archive, LEAPSECS subscribers

UTC without leap seconds would deviate from all other forms of "universal" time.

legal time (2003-04-10) local text/plain , LEAPSECS subscribers, (ideas included here)

If UTC switches to leap hours, how far in the future will those occur?
Under the current legal time for the US and UK (which is, effectively, UT2) will lawsuits begin to happen when DUT1 exceeds a few seconds?

UTC is doomed (2003-04-22) local text/plain , LEAPSECS subscribers, (ideas included here)

But it is not doomed anytime soon.

UT and the IAU (2003-05-10) local text/plain , LEAPSECS subscribers, (ideas included here, here, and here)

Universal Time was defined and approved by action of the IAU

will history repeat itself? (2003-05-10) local text/plain , LEAPSECS subscribers, (ideas included here)

In 1970 the CCIR (predecessor of the ITU) was in a similar position with respect to UTC. Rather than create an international and interdisciplinary committee to choose a solution, the CCIR acted unilaterally and imposed a solution on the world with less than 2 years notice.

Polyglot search engine bait

English

leap second, leap seconds

German

Schaltsekunde, Schaltsekunden

French

seconde intercalaire, secondes intercalaires

Spanish

segundo intercalar, segundos intercalares, segundo intercalado, segundos intercalados, segundo adicional, segundos adicionales, segundo extra (tal vez segundo bisiesto, segundos bisiestos)

Italian

secondo intercalare, secondi intercalari

Portuguese

segundo intercalado, segundos intercalados

Swedish

skottsekund, skottsekunder

Icelandic

hlaupasekundum, hlaupasekundur

Danish

skudsekund, skudsekunder

Dutch

schrikkelseconde, schrikkelseconden

Russian

дополнительная секунда, дополнительные секунды