How to calculate large distance taking the Earth curvature into consideration

[BennettWilliams]

I want to calculate distance of two points that is far away and at different place on Earth, for example flight-distance between major cities around the world.  It will be desirable to take the Earth curvature into the calculation.  Dose the "Distance or Near" function in ArcGIS do this?  I don't believe so.  How to do this and is there a script exist to do this?

Thanks in advance,

[AnthonyGiles]

Ming,

In the Nautical Solution Extension you can use the Production Feature Builder to create great circle distances:

http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//0106000001rr000000

Link to the extension site, not sure how much it is but you can download a 60 day trial:

http://www.esri.com/software/arcgis/extensions/nautical-solution/what-you-get

Regards

Anthony

[DaleHoneycutt · Accepted solution]

See also this blog: Calculating Geodesic Distance Between Points

[BennettWilliams]

Hi Anthony & dmhoneycutt,

Both of your reply are very helpful!

From Anthony's replay I learned that the problem is referring as "great circle distances".  I did further Google search and found the math formula for solve it in this site: http://en.wikipedia.org/wiki/Great-circle_distance.

Because I plan to do the calculation with c# code anyway (for special selection and post processing), I can use the formula to calculate the great-circle distance.

dmhoneycutt's replay prompt me to do a search in the ESRI SDK help, I found there is some interface to get the great-circle distance of two points.  So I can skip the formula and direct use the interface.

Thanks both of you for opening the door for me!

Ming

[AnthonyGiles]

Ming,

You can use the Haversine formula to calculate the distance, here is a site that has the code in various languages:

http://www.codecodex.com/wiki/Calculate_Distance_Between_Two_Points_on_a_Globe

You will find that the haversine formula can have an error of nearly 3%.

A better formula to use is the Vincenty formula which is more accurate, there are loads of examples out there:

http://www.gavaghan.org/blog/2007/08/06/c-gps-receivers-and-geocaching-vincentys-formula/

regards

Anthony