Vorpal
Neither a dandy nor a clown
- Location
- 76 Totter's Lane, Shoreditch, London, EC1 5EG
In 1983, the metre was redefined in terms of an exact speed of light in vacuum, c = 299792458 m/s. Not only was this a theoretically nice move, but in this case it improves fine length measurements by an almost an order of magnitude, because atomic clocks had much better relative uncertainty.
Since 2007, the International Committee for Weights and Measures (CIPM) started considering various proposals to redefine more units in terms of physical constants, similarly to the speed of light. For example, the current definition of kelvin, the unit of temperature, is based on the triple point of pure water—and there not only difficulties in getting water pure enough to the accuracy modern metrologists want, but this turns out to be inconvenient for accurate measurements below about 20 K or above 1300 K. To redefine temperature, the Boltzmann constant k could be measured in multiple independent ways, to similar uncertainty, a requirement which was met a few months ago.
Deadline for publication of results was 1 July 2017, with new values to be considered in Sep 2017, most likely to be later adopted as exact in Nov 2018, at the 26th General Conference for Weights and Measures
But by far the worst SI unit is the kilogram, defined as the mass of a particular platinum-iridium cylinder since 1889, nicknamed 'le grand K'. This represents a logistical nightmare of copying the artefact accurately, disseminating it to many nations, all the while worrying about whether or not they drift from one another.
Well, it also turned out that pretty much all official copies drift from le grand K, typically on the order of some tens of micrograms, and hence also from each other.
Fortunately, there is an important physical constant, Planck's constant h, that has essentially no reason to exist in physics but for the historical insistence that mass be a base unit, taken as axiomatic by Newton, which is convenient but not strictly necessary. With only one day to the deadline, NIST got it done:
There's a pretty nice youtube video explaining how their method works:
It relies on some very accurate quantum electric effects, so the electron charge is also expected to be fixed as exact.
The other independent way of measuring the kilogram, based on counting atoms, will be used to define the mole as exact instead. So the list of expected substantive changes for New SI are as follows (the values here are from 2014 CODATA; one should expect the adopted to be slightly different based on the newer published measurements):
In the current SI definition, the impedance of free space is exact and the electron charge is uncertain. This will flip things around, so that the electron charge is exact, but the impedance of free space has to be measured. It's a bit inconvenient for people who got used to quantum electrodynamics, since in the current system the (dimensionless) fine structure constant basically corresponds to the strength of an electron's charge, which is intuitive in that system. Well, ultimately both are just a unit choice that doesn't affect anything fundamental.
If one fixes both the speed of light and the impedance of free space (equivalently, magnetic permeability) to an exact value, that would be theoretically very nice, but not very practical, since it would be force one to abandon atomic frequencies as an exact standard for time. That would be unfortunate, because atomic clocks are extremely consistent, reliable, and stable. It's possible that in the future, the time standard will move from caesium-133, though.
Since 2007, the International Committee for Weights and Measures (CIPM) started considering various proposals to redefine more units in terms of physical constants, similarly to the speed of light. For example, the current definition of kelvin, the unit of temperature, is based on the triple point of pure water—and there not only difficulties in getting water pure enough to the accuracy modern metrologists want, but this turns out to be inconvenient for accurate measurements below about 20 K or above 1300 K. To redefine temperature, the Boltzmann constant k could be measured in multiple independent ways, to similar uncertainty, a requirement which was met a few months ago.
Deadline for publication of results was 1 July 2017, with new values to be considered in Sep 2017, most likely to be later adopted as exact in Nov 2018, at the 26th General Conference for Weights and Measures
But by far the worst SI unit is the kilogram, defined as the mass of a particular platinum-iridium cylinder since 1889, nicknamed 'le grand K'. This represents a logistical nightmare of copying the artefact accurately, disseminating it to many nations, all the while worrying about whether or not they drift from one another.
Article: Several years ago, NIST had to reissue certificates for its kilograms because they were 45 micrograms off the French prototype — about the weight of an eyelash. This meant that companies that produce weights based on the NIST standards had to reissue their own weights, and they were not happy about it. Lawmakers were called. NIST was accused of being incompetent. In the end, it turned out that the problem stemmed from le grand K, not NIST.
Well, it also turned out that pretty much all official copies drift from le grand K, typically on the order of some tens of micrograms, and hence also from each other.
Fortunately, there is an important physical constant, Planck's constant h, that has essentially no reason to exist in physics but for the historical insistence that mass be a base unit, taken as axiomatic by Newton, which is convenient but not strictly necessary. With only one day to the deadline, NIST got it done:
Article: The new NIST measurement of Planck's constant is 6.626069934×10⁻³⁴ kg·m²/s, with an uncertainty of only 13 parts per billion. NIST's previous measurement, published in 2016, had an uncertainty of 34 parts per billion.
There's a pretty nice youtube video explaining how their method works:
It relies on some very accurate quantum electric effects, so the electron charge is also expected to be fixed as exact.
The other independent way of measuring the kilogram, based on counting atoms, will be used to define the mole as exact instead. So the list of expected substantive changes for New SI are as follows (the values here are from 2014 CODATA; one should expect the adopted to be slightly different based on the newer published measurements):
- kilogram: by an exact value of Planck's constant, h ≈ 6.626070040×10⁻³⁴ J/Hz,
- mole: by an exact Avogadro's number, NA ≈ 6.022140857×10²³/mol,
- kelvin: by an exact Boltzmann's constant, k ≈ 1.38064852×10⁻²³ J/K,
- ampere:
abrogated as a base unit;redefined through the coulomb with an exact fundamental charge instead, e ≈ 1.6021766208×10⁻¹⁹ C.
In the current SI definition, the impedance of free space is exact and the electron charge is uncertain. This will flip things around, so that the electron charge is exact, but the impedance of free space has to be measured. It's a bit inconvenient for people who got used to quantum electrodynamics, since in the current system the (dimensionless) fine structure constant basically corresponds to the strength of an electron's charge, which is intuitive in that system. Well, ultimately both are just a unit choice that doesn't affect anything fundamental.
If one fixes both the speed of light and the impedance of free space (equivalently, magnetic permeability) to an exact value, that would be theoretically very nice, but not very practical, since it would be force one to abandon atomic frequencies as an exact standard for time. That would be unfortunate, because atomic clocks are extremely consistent, reliable, and stable. It's possible that in the future, the time standard will move from caesium-133, though.
Last edited:
