SI definition of second: what day/year is the reference year

[Patrick Aberdeen]

The SI definition of a second was originally based on 1/86 400 of a day - but the Earth's rotation is a) unsteady and b) changing so it was changed to a defined number of cycles of an atomic clock. Was there a specific day for which the atomic clock was calibrated? Put another way, why were 9 192 631 770 cycles of a Caesium atomic clock chosen as the number to define a second?

 

[scottdave]

I haven't found an exact answer for you, but it looks like around 1950 or so. https://en.wikipedia.org/wiki/Atomic_clock has a chart of how accuracy of atomic clocks have improved since 1950.

 

[jbriggs444]

The SI definition of a second was originally based on 1/86 400 of a day - but the Earth's rotation is a) unsteady and b) changing so it was changed to a defined number of cycles of an atomic clock. Was there a specific day for which the atomic clock was calibrated? Put another way, why were 9 192 631 770 cycles of a Caesium atomic clock chosen as the number to define a second?

My recollection is that the second was first redefined in terms of the tropical year and then redefined in terms of an atomic clock.
https://en.wikipedia.org/wiki/Second#History_of_definition

 

[Patrick Aberdeen]

Thanks scottdave and jbriggs - I think I found it. The day was the beginning of 1900 (technically Dec 31, 1899) - although the calculations were made in 1952. https://en.wikipedia.org/wiki/Ephemeris_time.

Now I just want to know what was the rotational velocity of Earth on that day :)