Progress on Optical Clock Technology for Operational Timescales

B. Hemingway; J. Whalen; S. Peil; T. G. Akin; W. Tobias

REVIEW 3 major objections 5 minor 13 references

Progress on Optical Clock Technology for Operational Timescales

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read An optical oscillator steered by a rubidium fountain can form a maserless hybrid clock with stability better than $10^{-14}$ at all averaging times.

desk verdict Solid, honest progress report on a maserless hybrid clock; the headline stability claim is projected, not yet demonstrated, but the architecture is credible and worth peer review. read the letter →

arxiv 2412.15403 v1 pith:HBDU6K5X submitted 2024-12-19 physics.atom-ph

S. Peil , W. Tobias , J. Whalen , B. Hemingway , T. G. Akin This is my paper · ORCID

classification physics.atom-ph

keywords opticalclockstimescalesrubidiumfountainoscillatorfrequencycombhydrogenmasercalciumbeamclocklattice

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that an optical oscillator—a cavity-stabilized laser whose output is divided to radio frequencies by a phase-locked frequency comb—can replace the hydrogen maser as the continuously running core of a national timekeeping system when it is steered by a rubidium atomic fountain. The resulting hybrid clock keeps optical-level stability at short times and inherits the fountain's reliable long-term accuracy, with measurements against a portable ytterbium lattice clock giving a one-second instability at or below $5\times10^{-15}$. If the modeling holds, such a clock would maintain better than $10^{-14}$ stability at every averaging time without any maser or steering synthesizer. The same program is developing continuous calcium atomic-beam clocks and a strontium lattice clock so that around-the-clock operations can eventually exceed the performance of rubidium fountains.

What carries the argument

The central object is the hybrid clock built from an optical oscillator—a 1542 nm cavity-stabilized diode laser plus a phase-locked 250 MHz-repetition-rate fiber frequency comb—steered by a rubidium fountain. The comb divides the optical frequency to RF outputs at 10 GHz and 200 MHz while preserving short-term optical stability; feedback to the comb repetition rate makes the fountain the absolute long-term reference. The steering loop averages the frequency difference between oscillator and fountain and corrects the comb's repetition rate, so the output integrates as the fountain after roughly 40 seconds. Supporting machinery includes Ramsey–Bordé spectroscopy on calcium atomic beams (thermal and slowed) for continuous clocks, k-reversal to cancel Doppler shifts from alignment drift, and a strontium lattice clock as the reference of record when available.

What would settle it

Compare the steered clock's 200 MHz or 5 MHz output continuously over days against an independent optical lattice clock; if their frequency difference ever drifts above $10^{-14}$ at any averaging time, or tracks the oscillator's 3 kHz/day temperature-dependent wander, the central stability claim is falsified.

Watch

Extended reading notes

Core claim

The paper's central claim is that a steerable optical oscillator can do the job of a maser in a timescale. A 1542 nm diode laser locked to a high-finesse cavity gives a Hz-level linewidth; a fiber frequency comb phase-locked to it generates 10 GHz and 200 MHz signals that carry the optical stability down to radio frequencies. A rubidium fountain, run continuously for 11 years, measures the optical oscillator's frequency and feeds back to the comb's repetition rate, producing the steered output. A model based on a $6\times10^{-14}$ white-frequency noise floor for the fountain predicts stability better than $10^{-14}$ for all averaging times, with the optical oscillator governing short times until the fountain integration takes over near 40 seconds. First measurements against a portable ytterbium lattice showed one-second instability at or below $5\times10^{-15}$ for the 10 GHz output, supporting the model. The paper additionally reports progress on two clock technologies meant for uninterrupted operation: Ramsey–Bordé spectroscopy on a thermal or laser-slowed calcium beam, and a strontium optical lattice as the ultimate reference.

Load-bearing premise

The entire performance prediction rests on the assumption that the fountain's timing noise stays at the level seen in the test loop once the clock runs continuously; if unmonitored drift from the laser's daily frequency wander enters, the 'better than $10^{-14}$ at all averaging times' promise may not hold.

Editorial extensions

If this is right

A timescale front-end can be built with no hydrogen maser and no steering synthesizer, gaining the optical oscillator's short-term stability while keeping the fountain's long-term accuracy.
The hybrid clock's 10 GHz output, measured at $5\times10^{-15}$ for one second and modeled below $10^{-14}$ at all averaging times, would outperform a maser-based timescale whose one-second stability is typically near $10^{-13}$.
Using the optical oscillator to drive the fountain's microwave chain instead of a quartz crystal pushes fountain stability toward its quantum-projection-noise limit, as low as $5\times10^{-14}$.
A continuously operating calcium-beam optical clock with 2.5 kHz Ramsey–Bordé fringes is projected to meet near-future stability goals while avoiding the complexity of trapped-atom systems.
Whatever uptime the strontium lattice achieves, the composite system's average instability improves in proportion to that uptime, and the lattice can be reported by calibrating the fountain rather than a maser.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the central claim holds, national timing labs could simplify their clock ensembles, removing masers and steering synthesizers to cut cost and failure modes—an engineering consequence the paper implies but does not develop.
A sharper test would be to compare the steered 5 MHz output against an independent optical reference over many days; the paper only measures 10 GHz, so division noise on the 5 MHz signal remains uncharacterized.
The k-reversal technique for thermal-beam clocks might eventually produce a passive optical clock simple enough for field use, which would carry the operational-timescale argument beyond laboratory timing.
The same hybrid architecture could work with any reliable reference, not just a fountain, allowing a shared optical flywheel to serve multiple timing sites.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

I read the USNO progress report on optical clock technology for operational timescales. The short version: it is an honest, useful progress report, and the central 'maserless hybrid clock' claim is a projection from a model, not a demonstrated out-of-loop performance. The paper does not overclaim; it says 'can be expected' and 'should be.' So the right read is: solid architecture, promising first data, still unproven at the headline level.

What is actually new: the hybrid clock demonstration, where a rubidium fountain is operated with its microwave chain driven by an optical oscillator, and the resulting 10 GHz output is compared against a NIST portable Yb lattice. The 1 s instability <=5e-15 is a real measurement, even if it is a single short campaign. The calcium beam Ramsey-Borde fringe data, including the k-reversal method for Doppler compensation, is also new and useful. As a community, we do not see many clear descriptions of the operational obstacles (thermal beam, slowed beam, lattice uptime) from a lab that actually runs a timescale. That alone is worth something.

The soft spot is exactly where the stress-test note put it: the better-than-1e-14-for-all-tau curve in Figure 1 is modeled, and the only validation is in-loop fountain data plus one external comparison. The oscillator's 3 kHz/day temperature-correlated drift is a real risk for the steering loop. If the loop cannot track sub-hour temperature fluctuations, the hybrid output will inherit noise not in the model. That is not a flaw in the paper's logic; it is a missing measurement. The paper is clear about the distinction between measured and modeled. My only additional gripe is minor: the Yb comparison is at 10 GHz, while the 5 MHz output is what timing systems actually use. They acknowledge that, but it means the operational performance is even less constrained.

Who this is for: people building timescales, UTC labs, and anyone thinking about maserless optical timing. It deserves a serious referee. The referee should push for a longer out-of-loop intercomparison against a maser ensemble or another optical clock, and for error bars on the Yb data. But the paper is coherent, honest, and technically sound as a progress report. I would accept it to peer review and let the referee set the bar for what 'demonstrated' means. I would probably cite it if I were writing about maserless architectures, though I would not lean on it for the headline performance.

Referee Report

3 major / 5 minor

Summary. The paper reports USNO progress on integrating optical clock technology into operational timescales. Section 2 describes a 1542-nm cavity-stabilized laser divided by a fiber comb to generate 10 GHz and 200 MHz signals, a rubidium fountain that steers the optical oscillator, and a modeled Allan deviation predicting better than 1e-14 stability at all averaging times. A first comparison with a portable Yb lattice from NIST gives a 1-s instability of 5e-15 for the 10 GHz hybrid output. Sections 3 and 4 outline ongoing development of an optical clock based on a calcium thermal/slowed atomic beam and a strontium lattice clock, respectively, with the lattice intended as a gold-standard frequency reference.

Significance. If the hybrid clock's modeled stability is confirmed out-of-loop, the architecture would replace hydrogen masers as timescale flywheels with an optical front-end, a significant practical advance for timing laboratories. The paper's strengths are its use of realistic measured parameters (fountain white-frequency floor of 6e-14, oscillator drift of 3 kHz/day), an explicit falsifiable prediction in Figure 1, and a first interlaboratory comparison with a NIST portable Yb lattice. The 6-month continuous operation of the optical oscillator and the in-loop fountain data are useful progress milestones. However, the paper does not provide a full archival demonstration; its central quantitative claim is a modeled expectation, not yet an out-of-loop verified measurement.

major comments (3)

[Section 2, Figure 1] The central claim of the abstract, that the hybrid clock has 'optical-level stability at short times and a reliable long-term reference' and obviates the need for a steered maser, rests on the modeled Allan deviation shown in Figure 1. The diamonds are explicitly in-loop fountain measurements; they validate the fountain's white-frequency floor when referenced to the unsteered optical oscillator, but they do not validate the steered hybrid output. In particular, the assumption that the hybrid integrates as a fountain with 6e-14 white-frequency noise after ~40 s is never tested out-of-loop. Without a measurement of the steered output against an independent reference over at least 10^3-10^4 s, the 'better than 1e-14 for all averaging times' statement is a prediction, not a demonstrated result. Please provide such an out-of-loop record or explicitly revise the abstract and Section 2 to state that this is a modeled expectation.
[Section 2, optical oscillator drift paragraph] The oscillator is described as having 'an average frequency drift of 3 kHz/day, with significant nonlinear variations strongly correlated with temperature over shorter times.' The steering loop must suppress this drift to achieve the modeled floor. The manuscript reports neither the loop's time constant nor the residual drift of the steered output. If sub-hour temperature fluctuations are not tracked by the fountain's measurement cycle, the hybrid output will exhibit an unmodeled bump in the Allan deviation. Please characterize the steering loop's bandwidth and present a continuous record of the steered output (or the steering corrections) to show that the 3 kHz/day drift is actually removed.
[Section 2, first test with portable Yb lattice] The statement 'Multiple runs gave frequency records indicating 1 s instability of our hybrid clock at or below 5e-15, and instability reaching 10^-14 before integrating as white-frequency noise' lacks supporting details. The number of runs, the duration of each record, the confidence interval on the 5e-15 value, and a plot of the Allan deviation should be provided. The Yb lattice is an independent out-of-loop reference, but the comparison appears limited to short times; without longer averaging data, it does not validate the 'better than 1e-14 for all averaging times' claim.

minor comments (5)

[References] Reference [11] has a duplicated year in the citation '20202020'; please correct it to '2020'.
[Figures 2 and 3] The horizontal axes of Figures 2 and 3 are labeled 'detuning' without units; adding units (e.g., kHz) would make the stated fringe widths of 2.5 kHz and 2 kHz directly verifiable.
[Section 2, first test paragraph] The phrase 'maiden journey' is informal; consider replacing it with 'first deployment' or 'first interlaboratory comparison.'
[Section 3.1] The stability budget for the thermal-beam optical clock is cited as [11] but not summarized; a sentence stating the dominant noise contributions and the projected stability floor would help readers assess the claim that a 2.5 kHz fringe width meets the short-term stability goals.
[Throughout] The symbol '10^-14' and similar expressions appear in running text; in a formal article these should be typeset with a superscript exponent (10^{-14}) for consistency with the displayed equations.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the hybrid-clock stability prediction is a forward model from measured noise floors and an independent Yb comparison; self-citations are supporting, not definitional.

full rationale

The paper's central claim is that an optical oscillator steered to a rubidium fountain forms a hybrid clock with stability better than 1e-14 at all averaging times. This is presented in Section 2 and Figure 1 as a modeled expectation, not as a quantity fitted to itself. The model inputs are the measured fountain white-frequency floor of 6e-14 (validated by the diamond data points, which are in-loop measurements made with the unsteered optical oscillator) and the optical oscillator's measured 3 kHz/day drift; the 'better than 1e-14' conclusion follows by integrating these measured noise contributions. The separate 1 s instability at or below 5e-15 result was obtained against the NIST portable Yb lattice, an external reference independent of the model. Citations [2] and [11] are self-citations, but they support operational reliability of the USNO fountains and the thermal-beam stability budget, respectively; neither citation defines the hybrid-clock stability prediction, and neither is used to forbid alternatives. No equation in the paper defines the predicted result in terms of itself, and no fitted parameter is relabeled as a prediction. The in-loop validation caveat noted by reviewers is a measurement-evidence concern, not a circularity. Accordingly, no circular step meets the standard of quoting a specific reduction, and the paper is self-contained against external benchmarks for its main demonstration.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper does not introduce free parameters or invented entities. However, its headline performance estimate depends on assumed noise floors and on the validity of in-loop validation, listed as axioms.

assumptions (4)

domain assumption The rubidium fountain is reliable enough to serve as the continuous long-term reference for steering the optical oscillator.
Section 2 states the USNO fountains have demonstrated 11 years of continuous operation and UTC contribution. The hybrid clock's long-term performance inherits this reliability.
domain assumption In-loop measurements of the fountain when run with the unsteered optical oscillator capture the true noise of the steered output.
Figure 1's data points are in-loop fountain performance; the text calls this validation, but it does not constitute an out-of-loop verification of the composite clock.
domain assumption The modeled Allan deviation uses assumed noise floors that are realistic inputs, specifically a fountain white-frequency noise of 6e-14 and a drift-noise crossing at about 40 seconds.
Section 2, Figure 1 is a model, not a measurement; if the assumed crossing time or noise floors are wrong, the 'better than 1e-14 at all averaging times' claim weakens.
domain assumption k-reversal suppresses alignment-induced Doppler shifts in the thermal atomic beam.
Section 3.1 describes this as an open question being investigated; it is not yet demonstrated, and the atomic-beam clock's long-term stability depends on it.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Progress on Optical Clock Technology for Operational Timescales." pith.science (2026). https://pith.science/paper/HBDU6K5X

@misc{pith2026241215403,
  author       = {Pith},
  title        = {Pith review of: Progress on Optical Clock Technology for Operational Timescales},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HBDU6K5X}},
  note         = {Machine review of arXiv:2412.15403}
}

read the original abstract

While optical clock technology has advanced rapidly in recent years, incorporating the technology into operational timescales has progressed more slowly. The highest accuracy frequency standards for groundbreaking measurements do not easily translate to critical timing where continuous, uninterrupted operation over many months and years is required. For example, intermittent steering of a hydrogen maser with an optical standard fails to harness all of the dramatic improvements possible with optical technology. Here we present progress on development and integration of optical-clock technology for operational timescales. An optical oscillator steered to an atomic fountain comprises a hybrid clock with optical-level stability at short times and a reliable long-term reference, and obviates the need for a steered maser. Atomic beam optical clocks are being developed to support 24/7 operations at a level that improves upon the performance of the U.S. Naval Observatory's rubidium fountains. An optical lattice is being developed as a gold-standard frequency reference, complementing the role of the atomic beam clocks.

Figures

Figures reproduced from arXiv: 2412.15403 by the authors.

**Figure 1.** Modeled behavior of hybrid optical-microwave clock. Solid and dashed curves labeled in graph legend show modeled Allan deviation for USNO rubidium fountain using optical oscillator for microwave generation (red, dashed), 10 GHz signal derived from optical oscillator (blue, solid), and lattice clock with 2 × 10−15 white-frequency noise level (black, dotted). The thick grey band illustrates the expected performance of… view at source ↗

**Figure 2.** Plot of Ramsey-Bord´e signal for each direction of laser propagation. The 2.5 kHz fringe width enables clear resolution of the two recoil resonances separated by 22 kHz. The horizontal “detuning” axis is roughly centered between these two resonances [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 12 canonical work pages

[1]

Wynands R and Weyers S 2005 Metrologia 42 S64

work page 2005
[2]

Peil S, Hanssen J, Swanson T B, Taylor J and Ekstrom C R 2016 Journal of Physics: Conference Series 723 012004

work page 2016
[3]

Roslund J D, Cing¨ oz A, Lunden W D, Partridge G B, Kowligy A S, Roller F, Sheredy D B, Skulason G E, Song J P, Abo-Shaeer J R and Boyd M M 2023 Optical clocks at sea ( Preprint 2308.12457)

work page Pith review arXiv 2023
[4]

Martin K W, Phelps G, Lemke N D, Bigelow M S, Stuhl B, Wojcik M, Holt M, Coddington I, Bishop M W and Burke J H 2018 Phys. Rev. Appl. 9(1) 014019

work page 2018
[5]

Grebing C, Al-Masoudi A, D¨ orscher S, H¨ afner S, Gerginov V, Weyers S, Lipphardt B, Riehle F, Sterr U and Lisdat C 2016 Optica 3 563–569

2016
[6]

Hachisu H, Nakagawa F, Hanado Y and Ido T 2018 Scientific Reports 8(1) 4243

work page 2018
[7]

Yao J, Sherman J A, Fortier T, Leopardi H, Parker T, McGrew W, Zhang X, Nicolodi D, Fasano R, Sch¨ affer S, Beloy K, Savory J, Romisch S, Oates C, Diddams S, Ludlow A and Levine J 2019 Phys. Rev. Appl. 12(4) 044069

work page 2019
[8]

Hati A, Nelson C W, Barnes C, Lirette D, Fortier T, Quinlan F, Desalvo J A, Ludlow A, Diddams S A and Howe D A 2013 IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 60 1796–1803

work page 2013

Show all 13 references

[9]

Mehlst¨ aubler T E, Grosche G, Lisdat C, Schmidt P O and Denker H 2018 Reports on Progress in Physics 81 064401

2018
[10]

Olson J, Fox R W, Fortier T M, Sheerin T F, Brown R C, Leopardi H, Stoner R E, Oates C W and Ludlow A D 2019 Phys. Rev. Lett. 123(7) 073202

2019
[11]

Hemingway B, Akin T G, Taylor J and Peil S 20202020 Joint Conference of the IEEE International Frequency Control Symposium and International Symposium on Applications of Ferroelectrics (IFCS-ISAF) pp 1–4
[12]

Taylor J, Hemingway B, Hanssen J, Swanson T B and Peil S 2018 J. Opt. Soc. Am. B 35 1557–1562

2018
[13]

Ito N, Ishikawa J and Morinaga A 1994 Optics Communications 109 414–421 ISSN 0030-4018

1994

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Wynands R and Weyers S 2005 Metrologia 42 S64

work page 2005

[2] [2]

Peil S, Hanssen J, Swanson T B, Taylor J and Ekstrom C R 2016 Journal of Physics: Conference Series 723 012004

work page 2016

[3] [3]

Roslund J D, Cing¨ oz A, Lunden W D, Partridge G B, Kowligy A S, Roller F, Sheredy D B, Skulason G E, Song J P, Abo-Shaeer J R and Boyd M M 2023 Optical clocks at sea ( Preprint 2308.12457)

work page Pith review arXiv 2023

[4] [4]

Martin K W, Phelps G, Lemke N D, Bigelow M S, Stuhl B, Wojcik M, Holt M, Coddington I, Bishop M W and Burke J H 2018 Phys. Rev. Appl. 9(1) 014019

work page 2018

[5] [5]

Grebing C, Al-Masoudi A, D¨ orscher S, H¨ afner S, Gerginov V, Weyers S, Lipphardt B, Riehle F, Sterr U and Lisdat C 2016 Optica 3 563–569

2016

[6] [6]

Hachisu H, Nakagawa F, Hanado Y and Ido T 2018 Scientific Reports 8(1) 4243

work page 2018

[7] [7]

Yao J, Sherman J A, Fortier T, Leopardi H, Parker T, McGrew W, Zhang X, Nicolodi D, Fasano R, Sch¨ affer S, Beloy K, Savory J, Romisch S, Oates C, Diddams S, Ludlow A and Levine J 2019 Phys. Rev. Appl. 12(4) 044069

work page 2019

[8] [8]

Hati A, Nelson C W, Barnes C, Lirette D, Fortier T, Quinlan F, Desalvo J A, Ludlow A, Diddams S A and Howe D A 2013 IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 60 1796–1803

work page 2013