[LEAPSECS] Ghosts of Leap-seconds past and future

Rob Seaman seaman at noao.edu
Tue Dec 28 13:04:28 EST 2010

On Dec 28, 2010, at 5:56 AM, Richard B. Langley wrote:

> I have been asked to remind list members of the presentation by Ron Beard, the chairman of ITU-R Working Party 7A, at ION GNSS 2010.

Thanks for forwarding this.

> Here are the conclusions, summary, and actions from the presentation:



These are chockfull of logical fallacies.

> Major scientific and GNSS organizations have not taken issue with the subject

> There has been ample opportunity and encouragement to contribute

This is simply not true. What is "the subject"? Prior discussion, as minimal as it has been, has been in reference to extremely unfocused talking points. The ITU proposal as it stands today is not widely available and has not been debated on its merits by these organizations.

More fundamentally, the stakeholders for any change to civil timekeeping are far broader than just "major scientific and GNSS organizations".

Due diligence has been completely lacking.

> The lack of response has been interpreted as having no concern and thus no established opinion

Absence of evidence is not evidence of absence.

> Little information on quantitative costs has been provided

> The few estimates offered seem to be guesses at best

"Millions and millions of dollars" is a pretty safe guess for the astronomical community.

> Few observers noted there are costs associated with maintaining the status quo that may or may not be mitigated

By all means include an accounting of costs for all options. But it is insipid to use this bullet as an argument to ignore accounting for both alternatives:


> Most experts in time metrology agree on the necessity for the change and offer technical support

It doesn't matter how many experts are dancing on the head of this pin:


There has been no evidence to date of such "technical support". For instance, what mechanism is proposed for disseminating DUT1 (or equivalent) in the future after the current mechanism is deprecated? Wouldn't it be helpful to attach such a technical white paper to the proposal?

> The Consultative Committee on Time and Frequency strongly recommends proceeding with a decision so enough time is available for any necessary software and systems modifications

Oh please! We're rushing to a decision precisely to provide "enough time" to deal with the fallout from this unseemly and unnecessary haste? By contrast, writing a coherent and complete system engineering plan allows sorting these issues out before a final decision is made.



> Documents demonstrate a clear misunderstanding of the definitions and applications of time scales and system times for internal synchronization

"Internal synchronization" is not the only use case. The lack of a coherent system engineering plan demonstrates ignorance as well as arrogance.

> o Indications that users have the choice between UTC, TAI, UT1, GPS Time for their applications is incorrect

"Indications"? There are many timescales, not just these listed. For civil timekeeping, these divide generally into two groups: Earth orientation and interval timescales. The entire point of the original definition of UTC was to recognize this fact.

Including UT1 in the above list is remarkable chutzpah. So the utility of UTC for astronomers is to be destroyed, but UT1 is recognized as not being an acceptable substitute?

Steve Allen has already addressed their apparent antipathy to both TAI and GPS.

> o UTC is the only international standard time scale, represented by local approximations in time laboratories, that should be used for worldwide time coordination and measurement traceability

UTC from the beginning has been what it says, "Universal Time". Universal Time is a modern equivalent of Greenwich Mean Time. Mean Solar Time is the natural cadence of the synodic (solar) day on Earth. Our civilization is pierced through and through with dependencies on the solar length-of-day. Inconvenient? Perhaps. A requirement? Yes.

> o TAI is not an option for applications needing a continuous reference as it has no means of dissemination, and it is not physically represented by clocks

This is a straw man argument - set it up, knock it down.

> o GPS time is not a reference time scale, it is an internal time for GPS system synchronization, as other GNSS system times would be

The synodic day is the reference time scale for civil timekeeping on Earth, as it has proven to be for operations of various missions on the Moon and Mars.

Calling GPS "an internal time" is disingenuous given the millions of devices that helpfully reveal GPS time to end users. This operational success appears to anger the various special interest committee members.

> o A variety of continuous internal system time scales have proliferated to provide a solution to the problems associated with discontinuities in UTC


> The existence of multiple time scales creates potential problems in operational use as well as conceptual confusion on the proper definition and roles of time references

This precipitous action to completely change the definition and character of UTC would ensure the proliferation of additional timescales as new realizations of universal time are necessarily reinvented.



> Working Party 7A exhausted technical considerations and studies

Not even close!

> Consensus not reached on other than technical grounds

No consensus, so by all means ram through your naive and self-serving proposal...

> Submitted to Study Group 7 for resolution

We wait with bated breath...

Rob Seaman
National Optical Astronomy Observatory

More information about the LEAPSECS mailing list