I have been checking in detail the specifications of RINEX navigation files version 3.05, and something in the definition of the header of these navigation files caught my attention.

One of the lines that the header can contain is labeled as "TIME SYSTEM CORR". The last field of this line is an integer which, judging from what is described in Table A5 of the linked specifications (page 66), seems to specify the exact UTC type to which the system time correction is performed. There seems to be at least 7 (!) of these: UTC(NIST), UTC(USNO), UTC(SU), UTC(BIPM), UTC(Europe Lab), UTC(CRL) and UTC(NTSC) (BDS).

I am aware that there are different Universal Time (UT) time scales, related but not equivalent, such as UT0, UT1, UT2 and UTC.

However this seems to be a whole new level. Is my interpretation correct in that there are in fact not just one, but multiple UTC definitions/implementations? If so, how do they differ?

  • 1
    $\begingroup$ I am not qualified enough to provide a detailed answer, but I suggest you take a look at the following article gssc.esa.int/navipedia/index.php/…. Especially the description. Hope this helps :) $\endgroup$ Dec 1, 2023 at 9:51

1 Answer 1


I think there's two important ways to think of them:

  • these are the UTC implementations; when you need to know the time now, precisely, these are the current time references you can use. Actual UTC will get worked out later. (Thanks PM 2Ring and Johny Doty)

  • these are subsets of UTC time which track UTC time but which may differ from UTC on the order of nanoseconds. Various groups have to run the various clocks that are ultimately combined to make TAI and UTC. Per BIPM:

    Physical realizations of UTC – named UTC(k) – are maintained in national metrology institutes or observatories, which contribute to UTC by sending their clock data to the BIPM. Values of [UTC – UTC(k)] at five-day intervals are published in the monthly BIPM Circular T, For those institutes participating in the CIPM MRA, their results are also published in the form of a key comparison, CCTF-K001.UTC, resulting in a monthly publication of the degrees of equivalence with respect to the reference time scale UTC.

NIST defines UTC(NIST) thusly:

The National Institute of Standards and Technology (NIST) maintains its own representation of UTC, called UTC(NIST), which serves as the official time standard for the United States. This was codified into law by the America COMPETES Act of 2007 which states that the official time for the United States is Coordinated Universal Time (UTC), as “interpreted or modified by the Secretary of Commerce, in coordination with the Secretary of the Navy.” Because NIST is the division of the Department of Commerce responsible for maintaining the U. S. standards of physical measurement, the passage of this bill formally established NIST as the official U. S. timekeeper.

It goes on to state that UTC(NIST) is maintained by typically 10 to 15 atomic clocks in Boulder, CO, whereas UTC is maintained by roughly 400 clocks around the world.

UTC(USNO) is maintained by the US Naval Observatory with their own set of frequency standards.

UTC(SU) is the Russian national time scale.

UTC(BIPM) is the official UTC time as coordinated by BIPM.

I'm honestly not clear on what "Europe Lab" is--perhaps UTC(ESA) kept by the European Space Operations Centre?

UTC(CRL) is the Japanese time standard.

UTC(NTSC) is maintained by the National Time Service Center (ex Shaanxi Astronomy Observatory) of the Chinese Academy of Sciences, and is also the basis for Beijing Time.

BDS apparently refers to the time standard for the Chinese BeiDou navigation satellite constellation (BDS). COMPASS/BeiDou Coordinate and Time Reference Systems National Academies of Sciences, Engineering, and Medicine. 2012. Global Navigation Satellite Systems: Report of a Joint Workshop of the National Academy of Engineering and the Chinese Academy of Engineering. Washington, DC: The National Academies Press. https://doi.org/10.17226/13292. mentions that it's "linked to the national UTC(k)" which I infer is NTSC, so I guess that's why RINEX 7 is (NTSC)(BDS), even though I think they'd technically be distinct.

  • 2
    $\begingroup$ if this answer looks like it was put together by hastily googling the UTC(k) terms--good eye! please feel free to correct me if I am wrong. $\endgroup$
    – Erin Anne
    Dec 1, 2023 at 10:04
  • 4
    $\begingroup$ In this context, NTSC is China's National Time Service Center. They have an English-language web page at english.ntsc.cas.cn , but it isn't working for me at the moment. $\endgroup$
    – Ryan C
    Dec 1, 2023 at 14:54
  • 4
    $\begingroup$ @Steve Those time labs are actually contributing to International Atomic Time, TAI. The conversion to UTC is simple, since "UTC deviates from TAI by a number of whole seconds". OTOH, you need to do astronomy to decide when a leap second is needed. But you only need to count caesium oscillations to drive a TAI clock. $\endgroup$
    – PM 2Ring
    Dec 1, 2023 at 17:39
  • 12
    $\begingroup$ The crucial point is that the various UTC(k) can be used to timestamp an event right now. If you want a UTC timestamp you have to make a preliminary timestamp using some UTC(k) then wait a few days for Circular T to be published so you know how to convert the UTC(k) value to UTC. $\endgroup$
    – PM 2Ring
    Dec 1, 2023 at 17:45
  • 3
    $\begingroup$ Amateur "time hacker" Tom van Baak has a site dedicated to precise time & frequency measurement. leapsecond.com Some of the links are dead, or out of date, but there's still a ton of fascinating & useful info on the site. $\endgroup$
    – PM 2Ring
    Dec 1, 2023 at 17:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.