Luke,

As usual, and as Roger well knows, it is difficult to sneak one by you!

Yes a simple value for the radius of Earth is 20/pi * 10^6 meters

If you use the approximation to Pi recommended by Zu Chongzhi
(http://aleph0.clarku.edu/~djoyce/mathhist/china.html)
of 355/113 to compute the radius of Earth you get:
(20*113/355)*10^6 m or
6.366197 * 10^6 m, precise to the nearest meter.

If you approximate Pi by 3.1415926 you get a radius of
6.366198 * 10^6 m, precise to the nearest meter.

I tried to verify your statement to the effect that: "We now know however that 
the
length of the quadrant is really 0.023% larger than originally surveyed, ..."  
and
got hopelessly lost in Chapter 4 of the Explanatory Supplement (© 1992) and in
Appendix X of Bowditch, 1977 Ed.   I have a renewed respect for geodesists and
those folk who delve into arcane matters and come up with elegantly sensible 
ways
of locating things.

I do recall from my earlier thinking on the matter that the 20/pi radius falls
between the polar radius and the equatorial radius and is close to the radius 
of a
sphere that has the same surface area as one of the ellipsoids used to 
approximate
the geoid.

In any event I made those things on Polaris Boulevard
(http://sciencenorth.on.ca/AboutSN/polaris/index.html )
to have diameters of 12.732 m or 12732 mm on the CAD drawings.

The steel was cut with a plasma torch all under numerical control.  When we
checked alignments and directions while installing, Celeste was within a mm in 
all
dimensions except when the wind got blowing.  No doubt strut vortices were shed.
Our latitude is so close to 45 degrees that she resonated quite well.  Not as 
bad
as the Tacoma Narrows Suspension Bridge, but she danced several cm.  So you'll
notice a little stiffening here and there to break the resonances.  She is very
well behaved now.  Almost boring!

The spherical terrazzo on Terra was laid in by an elderly Italian gentleman 
whose
family had been practicing the art for generations.  (And it is still standing 
up
well in spite of harsh environmental treatment, including freeze/thaw, some 
salt,
skate boards and the dancing of little shoes as kids celebrate being at Sudbury,
on the "top of the world"!)

I figured that an approximation between polar diameter and equatorial diameter 
was
good enough for government work.

Besides it was fun to talk about the origins of the meter!

Now if someone would just explain all of this ellipsoid of revolution, geoidal
separations and equipotential surface stuff, we might have a chance to 
understand
where the 0.023% "error" is!  *SIGH*  (In any event I figure that thermal
expansion/contraction episodes will get the thing passing through the correct
value from time to time.)

Good stuff, Luke. Thanks.

Tom Semadeni



> Tom,
>
>         Perhaps one of the lists geodesists could give a more complete answer
> but I believe the original French definition of the meter was the
> subtended arc-length (along a great circle) of one ten-millionth of a
> guadrant (a guadrant = 90 deg. or pi/2 radians). This survey, by the way
> took seven years to complete! Nothing like a government contract...
>
>         Therefore, since arc-length = radius * theta (in radians) we have the
> following expression:
>
> 10^7m / (pi/2) = r = 2*10^7m / pi
>
>         With a reduction in scale of one million (divide above by 10^6) we
> have:
>
> 20/pi m
>
>         We now know however that the length of the quadrant is really 0.023%
> larger than originally surveyed, I hope you have made allowances for the
> corresponding error. Just kidding.
>
> Best,
>
> Luke
>

Reply via email to