[time-nuts] The forbidden question

[Dana Whitlow]

In answer to the question about radio astronomers, consider VLBI (Very Long
Baseline Interferometry).

VLBI is a mapping (imaging) process in which signals are *simultaneously*
received from
a small sky region of interest by a collection of radio telescopes
scattered about the world.
In order for the process to work, ultimately it is necessary for all the
signals to be lined up
in time (think phase) to a small fraction of a period of the microwave
frequency being used.
Suppose the frequency is around 5 GHz, then the period is just 200 psec,
and a small
enough fraction thereof is maybe 10 psec or less.

Now present day clocks and time transfer schemes are not good enough to
accomplish
that feat "open loop" across a collection of observatories spanning
thousands of miles.
Fortunately, radio astronomers have developed a set of magical techniques
collectively
called "self calibration" by which, with extreme care, the job can still be
done provided
that the data sets from each observatory are initially lined up to within
around 100 nsec.

However, another requirement is that the LOs for the different
observatories' receivers maintain
that constant phase relationship I mentioned above during the course of an
observation.  In
practice this condition can typically only be held for a few minutes at a
time, and the only
atomic clock in commercial production that can do that is the hydrogen
maser.  Cs beam
clocks are simply not stable enough in that time frame to get the job
done.  But I'm quite sure
that much more stable clocks yet would relieve the overall problem of
making much longer
continuous observations.  Not solve all aspects of the problem by itself,
but it would sure
help.

Another use for such observations is earthbound geodesy, used to track
continental drift
and the like.  Here the preferred radio source is a bright point source
such as a quasar,
and the point is monitoring the position of each observatory.   In this
application it's
important to accurately know the position of the selected source, and one
of the applications
of VLBI is also to measure source locations.  So if you're an
astrophysicist you want to know
the precise locations of all the telescopes, and uncertainty in this is
"noise".  But the person
wanting geodesic information want to measure drift in telescope positions,
and needs to
know where the source is located in the sky.  So one man's signal is the
other man's noise,
and vice versa.  In any case, anything that can be done to improve phase
coherence between
all the receivers in the network can only help.

Dana

Previous "Keeper of The Clock" at the Arecibo Observatory in Puerto Rico

On Tue, Jun 4, 2019 at 6:01 PM Poul-Henning Kamp <phk at phk.freebsd.dk> wrote:

> --------
> In message <CANy2iXowYBPreNnbrjnU7_XLz=NJ5ZVaGVMT0PRmUD0=
> 7BLf9A at mail.gmail.com>
> , "William H. Fite" writes:
>
> >What I am asking is not the validity of the quest for better timing
> >but rather its tangible applications.
>
> Tangible for who ?
>
> For the average pedestrian there are no *current* tangible applications
> where cesium level time-keeping isn't plenty.
>
> However, the same would have been said about chronometers and quartz
> clocks at various times in the past.
>
> To answer your question we would need to look about 20-30 years
> into the future, which seems to be the median time for better
> timekeeping to break through to the wider public, even if they do
> not know it has happened, (ie: longitude navigation, digital telephone
> networks, GPS.
>
> Peeking 20-30 years into the future is an unsolved problem, so I
> would argue that your question is unanswerable at this time.
>
> --
> Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
> phk at FreeBSD.ORG         | TCP/IP since RFC 956
> FreeBSD committer       | BSD since 4.3-tahoe
> Never attribute to malice what can adequately be explained by incompetence.
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to
> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
>