Error in Geodetic Coordinates

That's "close enough" for a BALLPARK number.

It works "okay" for Latitude (anywhere), but you need to multiply it by the cosine of the Latitude, in order to get a BALLPARK number for your Longitude difference (at a given Latitude).

Loyal

no

Yes ballpark due to the radius changing depending on location.

In any small angular difference curvature has no effect so the right triangle approach work I believe.

However, why is there a difference between lat and lon?? Both are earth centric correct?

Just look at a globe...

The meridians (lines of constant Longitude) converge at the Poles.

0.01" of Longitude is much different @ 45° Latitude, than it is @ the Equator.

Loyal

ah yes

Plane trigonometry is not a good tool for these computations. I would approximate using the following methods listed in order of increasing accuracy and complexity.

As a rough approximation I would use s = R times theta where s is arc distance, R is radius of the sphere and theta is the angle ( IN RADIANS!)

A better approximation can be can be obtained using a great circle computation using the law of cosines: cos_c = cos_a * cos_b + sin_a * sin_b * cos_C where a, b are co latitudes of end points and C is the difference in longitudes. Answer to the equation is the cosine of side c. Convert to angle using arcosine then multiply by radius of sphere to get spherical distance.

See PDF here http://geodesyattamucc.pbworks.com/f/GreatCircleComputations.pdf for some graphics and more details including some Matlab code.

Working on an ellipse is more challenging. Numerous web resources exist for both the computation (INVERSE or IMVERS3D on NGS site) and the algorithms. Bowring's method is pretty straightforward.

Cheers,

DMM

I use 100' per second of lat. 70' per second of longitude where I am. Just for a rough idea of the shift between points. If you look at the circumference of the earth in feet it's close to the number of seconds in a circle times 100. So to get to .01' it's the 4th decimal place of a second of arc. More or less.

Remember... the Earth isn't a sphere.

Thanks all!!