[time-nuts] BYMLBI (Back Yard Medium LBI)

[Chris Albertson]

On Fri, Mar 18, 2011 at 8:36 AM, Brooke Clarke <brooke at pacific.net> wrote:
> Hi:
>
> For a number of decades I've tried to figure out a way to measure the period
> of the Earth's rotation accurate to at least 1 ms.


The US Navel observatory at Flagstaff was an 8 inch transit telescope.
 You can Google "FASTT" and read more.

You can measure to 1ms optically actually you can do 1000 times
better.  The trick is to look at many, not just one star and to use a
very accurate star catalog.   You don't really measure the position of
one star, you measure the position of a patch of sky that contains
many stars who's positions are known to great accuracy and then you
solve to find a best solution for the entire frame.    They measure
thousands of stars every  night and have been doing so for 30 years.

The result of this is that they know the local optical reference
frames is not stationary withing the frame defined by distant quasars.

One of the things to know is that an optical telescope can measure
position of a star to maybe 10x better than it's resolving powerYes
the stars look like little blurs but you can use some analytic methods
to find the pear of the blur.  The simplest method is just to take the
centroid.   You avoid issues with resolving close stars by ignoring
them and using only the "best" isolated stars.

There is MUCH detail to worry about if you want to match FASST's
results but person working at home can still do better then 1ms if you
set u a good program with automated data reduction and run it long
enough.   Almost all of the software you need to free although you'd
need to do considerable integration.   And that there is the learning
curve.
-- 
=====
Chris Albertson
Redondo Beach, California