[time-nuts] Fast than light neutrino

Tom Van Baak tvb at LeapSecond.com
Sun Sep 25 13:17:35 UTC 2011

> What about rotation of the earth?


Correct, there is the Sagnac effect to account for when you travel
easterly or westerly with a portable cesium clock (or equivalently,
when you send E/M timing signals over wire, fiber, or up/down to
a satellite).

Typically the correction is built into the software timing that labs use
to perform one-way, common view, or two-way sat synchronization.
For example, the 1pps from a GPS timing receiver already has this
effect included, so no one needs to worry about it anymore.

Still, let's say they forgot. You can estimate how much the Sagnac
effect would be between CERN (46N 06E) and GSNL (42N 14E)
and the answer is about 2.4 ns. Using sagnac.exe (src sagnac.c)
from www.leapsecond.com/tools/ here's the rough estimate:

0.2074 us sagnac effect (full round-trip at equator)
0.1073 us sagnac effect (full round-trip at latitude 44)
0.0024 us sagnac effect (8 degree longitude trip at latitude 44)

So earth rotation is not where their missing nanoseconds are.

The other way to calculate: at 44 degrees latitude, the earth spins
about 335 m/s (750 mph). The neutrinos travel east from CERN
to GSNL at c for about 2.43 milliseconds. During that short trip,
the earth moves only 0.8 meters.


>> We did not forget. The two GPS calibration campaigns (zero baseline and
>> portable receiver) were done with antenna and antenna cable included.

>> You mean between CERN and Gran Sasso? No, but that's certainly something we
>> could explore for the future.
>> Cheers,
>> Javier

More information about the time-nuts mailing list