Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock (
) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
Summary
Atomic clocks are key to a wide array of technologies and scientific experiments, including tests of fundamental physics. SED staff accounted for the effects of excess scatter in frequency ratio measurements and rigorously evaluated the measurement uncertainties in a common framework by developing a comprehensive Bayesian model for each ratio.
Description
Atomic clocks are key to a wide array of technologies and scientific experiments, including tests of fundamental physics. Clocks operating at optical frequencies have now demonstrated fractional stability and reproducibility at 10-18, two orders of magnitude beyond their microwave predecessors. Frequency ratio measurements between optical clocks are the basis for many of the applications that leverage this extreme precision. However, the highest reported accuracy for frequency ratio measurements has remained nearly static at 5×10-17 for more than a decade. In this project, we operate a network of optical clocks based on 27Al+, 87Sr, and 171Yb, and measure their respective frequency ratios with fractional uncertainties at or below 8×10-18. The optical clock network utilizes not just optical fiber, but also a 1.5 km free-space link. This key advance in frequency ratio measurements lays the groundwork for future networks of mobile, airborne and remote optical clocks that will be used in new ways to test physical laws, perform relativistic geodesy, and transform international timekeeping.
For the three sets of ratio measurements from these clocks, scientists observed various degrees of between-day variability that exceeded what was expected from the statistical uncertainties alone. SED staff helped the scientists account for the effects of this excess scatter and evaluate the measurement uncertainties rigorously in a common framework by developing a comprehensive Bayesian model for each ratio. The model incorporates uncertainty due to the known statistical and systematic effects but also allows for unknown effects that may vary between days. Such an approach provides not only an estimate of the ratio uncertainty but also an estimate of the uncertainties of other model outputs such as an additional between-day variability not included in the clocks’ standard uncertainty evaluations. This work resulted in a publication in Nature.