NIST and Metrology

June 11, 2012  - By
Image: GPS World

I must govern the clock, not be governed by it.
— Golda Meir

The Question

A few months ago at a speaking engagement, I took questions from the audience after my presentation. The audience was made up of GPS enthusiasts, GPS equipment vendors, and evidently GPS neophytes as well, because the last question was asked by a young lady, from a large well known government user segment, who was intrigued by but obviously knew little about the inner workings of GPS. Her question so stunned me and the audience that it brought the questions to an abrupt end. Thank goodness no one actually laughed out loud, and frankly I was so incredulous that I almost gave an impertinent answer that would not have served any purpose other than to embarrass the young lady and expose some insensitivity on my part. No, fortunately, after recovering from the initial shock due to the naiveté of the question, I answered her with a straight face, because it seemed to be an honest and sincere question.

Allow me to set the stage. My talk was on the Perfect Handheld GPS Transceiver and how the PHGPST could be aided by a Symmetricom Chip Scale Atomic Clock (CSAC).

The young lady’s comment and eventual question was exactly as follows: “I came here this afternoon because I was intrigued by your columns in GPS World and I wanted to know more about the Perfect Handheld GPS Transceiver and possibly learn how I could even purchase one or more for my organization. However, most of your comments have been focused on the benefits of atomic clocks. Frankly, I am a bit disappointed. So where can I buy a PHGPST and what do atomic clocks have to do with GPS anyway?”


I have had a few months to think about that episode, and although all ended well, with no one being overly embarrassed, despite some good-natured ribbing, it was a little unsettling. It also takes me back to a previous theme in several of my columns concerning educating users about the Global Positioning System. Not just what GPS can do for you — certainly that is well covered in GPS World, other publications, and on the Internet. Indeed, just type the acronym “GPS” into any search engine and you will be rewarded with the rather daunting number of 1,670,000,000 hits — yes that’s 1 billion, 670 million hits.

A Daunting Perspective

Considering that the average person today, who has reached my advanced age, my will probably live to be approximately 80 years of age or more, the obvious question is does anyone actually have the time to peruse ~1.6B websites on GPS?

You don’t have to be a professional metrologist or an expert mathematician to determine the logical answer, but if you are really concerned about time it might help. Consider the following answers — yes, plural — to the question posed, which assumes that a person would spend one minute or 60 seconds at each website — which begs the question, just how many minutes are there in an 80-year life span anyway? Of course, this answer assumes the hopefully unlikely event that one would come out of the womb Googling “GPS.”


By the nominal quartz clock on the wall, and using the Gregorian calendar, and not considering leap years, 80 human years equates to 42,048,000 minutes.  If we utilize the Julian calendar and add leap years, it equates to 42,076,800 minutes. By a standard years definition it equates to 42,075,936 minutes. By SI, international system of units, or true metrology standards, 80 years equates to, 42,075,901.3 minutes. The differences have to do with metrology and atomic reference systems versus the nominal unaided quartz clock. And although we say GPS runs on atomic clocks, the true answer is GPS runs on highly stable (accuracy is not a relevant term to be used here) atomic reference systems — noble gases and all that.

Now stay with me and allow me to explain the 80-years-in-seconds answers (and you naively thought there was only one answer) in terms a metrologist (the guys and gals who really care about time and frequency) at NIST, the National Institute of Standards and Technology in Boulder, Colorado, would use:

80 Years by the Calendar and Clock

For an entire block of 80 years, containing 20 leap years, the number of minutes would be the same as in 80 Julian calendar years of 365.25 days.

The number of minutes as calculated by calendar and clock is 42,076,800 — 80 years x 365.25 days/year x 24 hours/day x 60 minutes/hour = 42,076,800.

Or 80 times the number of minutes in a year, which is calculated as 525,960.

SI or Leap Years

Now, when we approach the question from an SI perspective, the answer is slightly but significantly different. The definition of a year is 31,556,926 standard seconds, while the standard leap year calculation is equal to 365 days, 5 hours, 49 minutes, and 12 seconds (31,556,952 seconds). Instead of 525,960 clock minutes, you have 525,949.2 (standard) minutes. For 80 years, the results equal 42,075,936 minutes.

The variation between clock/calendar minutes and the measured length of the year only becomes important to those unfortunate enough to be born on the 29th of February. That’s when the two calculations and calendars diverge by enough to subtract an entire day from the normal leap year system, hence the varying length of February, on the Gregorian calendar, every four years. However, to metrologist and GPS experts, who define an SI second as the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the Caesium 133 atom, the differences are astronomical (pun intended).

Caesium, Rubidium, Hydrogen…

At this point many GPS/PNT cognoscenti are probably saying, wait a minute, there are also Rubidium and hydrogen maser clocks on orbit today, so why use the Caesium standard for the SI second?

The Standard Tale

The Caesium standard for the SI second was established in 1960 when it was decided that it was time to abandon the astronomical or ephemeris, revolution of the Earth around the Sun, basis for the second. Indeed Louis Essen from the National Physical Laboratory (NPL in Teddington, England) and William Markowitz of the United States Naval Observatory (USNO in Washington, D.C.) determined the relationship between the hyperfine transition frequency of the Caesium atom and the ephemeris second. Using a common-view measurement method based on the received signals from radio station WWV  (operated by NIST and broadcast continuously from Boulder, Colorado) they determined the orbital motion (ephemeris) of the Moon about the Earth. From this data they inferred the apparent motion of the Sun, in terms of time as measured by an atomic reference system, in this case Caesium. They determined a second of ephemeris time (ET) to be equal to 9,192,631,770 ± 20 cycles of the atomic Caesium frequency. Consequently, in 1967 the 13th General Conference on Weights and Measures formally defined a second of atomic time in the International System of Units as the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the Caesium-133 atom.

So, since 1967 the Caesium atom has determined and supplied the atomic time reference for the globe. Caesium clocks have certainly undergone numerous refinements, and companies like Symmetricom in the U.S. still build Caesium reference systems that are used globally to determine the SI second and hence coordinate both International Atomic Time (IAT) and UTC or Coordinated Universal Time. And just how accurate is Caesium as a standard? The Swiss have built a continuous cold Caesium fountain atomic clock in Switzerland that started operating in 2004 at an uncertainty of one SI second in 30 million years.

Numbers Can Be Deceiving

Now while that sounds and is very impressive, the fact of the matter is one SI second is a long duration when it comes to atomic clocks — notice I avoided saying a “long time.” Indeed, today metrologists routinely speak in terms of 18 orders of magnitude less than an SI second. That is 18 places to the right of the decimal place, a very small number, destined to become even smaller.




yoctosecond 10−24 s
zeptosecond 10−21 s Future of optical reference systems
attosecond 10−18 s Shortest time uncertaintyin present measurements
femtosecond 10−15 s Pulse time of ultrafast lasers(100 as = 0.1 fs) – Hydrogen Maser
picosecond 10−12 s
nanosecond 10−9 s Time for molecules to fluoresce
microsecond 10−6 s Think millionths of a second
millisecond 0.001 s, 10E-3 Think thousandths of a second
second 1.0 s SI base unit

We are speaking of incredibly short measures of duration (time) and they keep getting smaller.

So What?

Now many of you may be thinking, this is all very well and good, possibly even interesting, but so what? OK, here is the “so what.” Asking what GPS and all the incredible technology it enables has to do with atomic clocks is like asking what gas or electricity, as the case may be, has to do with my automobile. Atomic clocks, and the increasingly minute measurements of time they define, fuel our global economy today. And if you doubt the veracity of that statement, think about all that GPS/PNT enables in around the globe. All of our critical national infrastructure depends on a coordinated time, and the primary distribution system for that time for the last 20+ years has been GPS.

While there are academically numerous definitions of time and several so-called standards, primarily today the world runs on UTC (Universal Coordinated Time) distributed by GPS and other PNT systems with the requisite corrections.









Mean Solar Time


UTC = UT1 – DUT1

TT = UT1 + 32.184 s + LS – DUT1

TAI = UT1 – DUT1 + LS

GPS = UT1 – DUT1 + LS – 19 s


Civil Time

UT1 = UTC + DUT1


TT = UTC + 32.184 s + LS


GPS = UTC + LS – 19 s


Terrestrial (Ephemeris) Time

UT1 = TT – 32.184 s – LS + DUT1

UTC = TT – 32.184 s – LS


TAI = TT – 32.184 s

GPS = TT – 51.184 s


Atomic Time

UT1 = TAI + DUT1 – LS


TT = TAI + 32.184 s


GPS = TAI – 19 s


GPS Time

UT1 = GPS + DUT1 – LS + 19 s

UTC = GPS – LS + 19 s

TT = GPS + 51.184 s

TAI = GPS + 19 s



Note: In June 2012, GPS will add another leap second.


NIST and Metrology

What then is time? If no one asks me,
I know what it is. If I wish to explain

it to him who asks, I do not know.
— Saint Augustine
Now, just how do you learn about atomic time and frequency, GPS, UTC, clock errors, ephemeris errors, and all the other multitude of errors that affect the GPS time we all, meaning the world, depend on every day?

Obviously there are many routes to that knowledge, but for me, and I have been involved with GPS since 1978, the best resource I have experienced to date was the annual week-long Time and Frequency Metrology Seminar that has been held at NIST in Boulder, Colorado, for the last 37 years.

I was kindly invited by Dr. David Howe, the seminar chair and group leader of the Time and Frequency Metrology Physics Measurements Laboratory, Time and Frequency Division at NIST, to attend this year’s seminar, and it was an eye-opening experience. Not only for the knowledge gained, but for the people met and the networking opportunities as well.

The Metrology Seminar

The seminar is billed as a course on understanding clocks, oscillators, atomic frequency standards, RF (radio frequency) and optical synchronization, optical oscillators, quantum information, optical cooling and heating (think lasers); making precise frequency, time, phase-noise, and jitter measurements; and establishing measurement accuracy and traceability. I am convinced the four-day course is the most comprehensive available anywhere today.

The June 2012 seminar included extremely informative presentations by Judah Levine, David Howe and David Allan (Ph.D.s all) — David Allan is the author of the Allan Variance — and 23 other presenters in subjects as diverse as direct-digital PM (phase modulation) noise measurements, how to specify frequency uncertainty, oscillator needs for new radars and surveillance systems, GPS vs. other global navigation satellite systems, photonic (laser-based) oscillators, chip-scale atomic clocks, femtosecond laser dividers, active PM-noise reduction techniques in oscillators, millimeter-wave applications and noise measurements, and ultra-low noise amplifier design techniques.

As the seminar descriptor says it is comprehensive in nature and there is no wasted time. One day the presentations lasted from 8 a.m. until 9 p.m. that evening. So it is an intensive metrology seminar filled with incredible learning opportunities. This year’s activities included an enjoyable and educational evening with Symmetricom, a forward-looking company that I sincerely view as the last true major commercial atomic reference builder in the United States. Symmetricom produces Hydrogen, Cesium (the ubiquitous 5071A), and Rubidium atomic reference standards, as well as high-end quartz oscillators with superior spectral purity and short-term frequency stability that support more military communications, satellite ground stations, and test and measurement applications than any other precision frequency references in the world today. Plus the incredible Symmetricom Chip-Scale Atomic Clock (CSAC), the world’s smallest atomic reference, has achieved historic breakthroughs in size, weight and power consumption — you can and I have balanced a CSAC on the end of my little finger. It is tiny yet powerful, and supports reference requirements as small as 1x10E-12 for specified periods of time. Indeed, this is the device that has the potential to revolutionize the Perfect Handheld GPS Transciever (PHGPST), but that is the subject for another column. As are the interviews I conducted with Judah Levine, who I refer to as “Father Time,” a Fellow at NIST, and David Allan, the originator of the Allan Variance and Allan Deviation, which allows the GPS Kalman Filter to work wonders and assures a GPS position accuracy of less than 1 meter possible for us all, with the right equipment of course.

The Right Equipment

Which leads me to my closing comments for this month’s column. While it is true that you can now routinely utilize GPS for real-time centimeter accuracy, since GPS does have the best atomic reference systems of any PNT system in orbit today, without monitoring systems and excellent GPS receivers you can’t depend on those parameters — the integrity is just not guaranteed without adequate signal monitoring, corrections, and augmentations.

Consequently, to ensure the best possible results, the receivers in the GPS global monitoring stations, of which there are hundreds more than those sanctioned by the U.S. government, tend to be high end and of stellar quality. And a majority of those receivers, according to my sources are, NovAtel receivers. Certainly NIST has a good number of NovAtel reference receivers, even though they are not allowed to specify or recommend a reference-grade receiver. I have also noticed large numbers of NovAtel receivers at FAA facilities around the U.S., and NovAtel’s reference receivers are at the core of national aviation ground networks around the world.

Recently NovAtel announced the development of a new receiver card known as the OEM625S Selective Availability Anti-Spoofing Module (SAASM) Global Navigation Satellite System (GNSS) receiver, which is actually a collaborative effort between NovAtel and L-3 Interstate Electronics Corporation (IEC). Since this is designated an OEM card, it is primarily for system integrators. However, since many users worldwide have come to rely on the centimeter-level positioning accuracy of real-time kinematic (RTK) GPS receivers, this may be a card you want to specify in your next high-end GPS receiver purchase or upgrade.

PPS Accuracy

Since defense users routinely rely on access to the Precise Positioning Service (PPS) for single-point positioning and this is certainly available with the OEM625S, it should be of special interest to those users. The card combines a commercial dual-frequency NovAtel GNSS receiver with an L-3 IEC XFACTOR SAASM in a single card solution, reducing overall size and power requirements for user applications. The new card maintains the OEMV-2 form factor, meaning it should be a drop-in replacement with backward compatibility for existing users.

NovAtel has promised to ship me a card to review in depth at the end of the third quarter of this year, so more on the new NovAtel/L3 card at that time. It promises real-time centimeter level accuracy and I can’t wait to see how it performs.

Make Your Reservation

Meanwhile, give Dr. David Howe a call at NIST and sign-up for the 38th Annual Time and Frequency Metrology Seminar — you will be glad you did.

I hope to see you all at ION JNC (Institute of Navigation – Joint Navigation Conference) in Colorado Springs, June 12-15, 2012! The classified Warrior Panel promises to be the hit of the show. Don’t miss it!

Until next time, Happy navigating.


This article is tagged with , and posted in Defense, Opinions