The Canadian site is actually fascinating... Lots of stuff there about how
the magnetic field is NOT adequately represented by a dipole (What, no bar
magnet in the middle of the earth?)...
At 09:58 AM 2/11/2003 -0500, you wrote:
>Take a look at this interesting site. Knowing the location and
>characteristics of the Magnetic North Pole will help anyone remember how to
>correct a compass reading to True North.
>
><A
>HREF="http://www.geolab.nrcan.gc.ca/geomag/home_e.shtml";>http://www.geolab.nrcan.gc.ca/geomag/home_e.shtml</A>
>
>Another tip for those of us in the northern hemisphere. Find the time for
>your local sunrise and sunset. Divide that time period by two to determine
>your local noon. At that moment, the shadow of your tower points to true
>North.
This is only accurate to about a 1/2 degree (good enough for run of the
mill HF antenna pointing.. not particularly good enough with a 40 dBi dish).
1) The sun is 1/2 degree wide, so the shadow is somewhat indistinct. If
you're sighting with a sextant or transit, you can sight on the limb (edge)
of the solar image.
2) Local solar transit (noon) is exactly halfway between sunrise and sunset
times only when it happens to occur exactly at the solstice. Otherwise it's
off by a little bit. The saving grace is that you're essentially making a
linear approximation to a function that is sinusoidal underneath, and the
error over an interval of 1/(2*365), is going to be pretty small.
Example: Cleveland OH
March 21, 2003: rise, 629am, transit 1234pm, sunset 640pm - (29+40)/2
= 34.5
March 22, 2003: rise, 627am, transit 1234pm sunset 641pm - (27+41)/2
= 34
Thousand Oaks, CA
1 Jan 702am 1159am 457pm (62+57)/2 = 59.5
2 Jan 702am 1200n 458pm (62+58)/2 = 60 (00)
22 March, 557am 1203pm 609pm
21 March, 558am 1203pm 608pm
Splitting the rise/set times probably gives you local transit time to an
accuracy of 30 seconds, which corresponds to about 1/4 to 1/8 degree of
heading accuracy.
Given that the handy UNSO site actually gives transit time, you can use
that directly, but you're still limited by the optical accuracy problem
(1/2 degree wide source) and the fact that you need to know your lat/lon
accurately. A longitude error of 1/4 degree (around 10-15 miles)
corresponds to a 1 minute change in time.
UNSO Site for sunrise/set: http://aa.usno.navy.mil/data/docs/RS_OneDay.html
There are some other sites that will give you the azimuth of the sun, from
your location, at an arbitrary time, so you don't need to be constrained to
making your measurement at local noon. (if a cloud happens to come over at
just the wrong time!)
|