Calculating distance between geographic points

[Guest lmoelleb]

Hi,I am fooling around with a small utility where I need to calculate the distance between two geographic points giving the latitude and longitude. More precisely I am trying to calculate the point at a given distance from the reference point. I am using the following formula, but my test shows that it is something like 10% off the size:longitude.Radian += distance.X / ((40075000.0 / (2.0 * Math.PI)) * Math.Cos(Latitude.Radian));latitude.Radian += distance.Y / (40007000.0 / (2.0 * Math.PI));I am aware that you should normally try to use the latitude between the two points, but my base point is always the center of the object, so I should be fine - and nothing will be bigger than 1km anyway.Does anyone have the correct parameters to use with Flight Simulator? I got these of the SDK but probably misunderstood one or another thing. My current formula use meters, but if I get something else, I think it is within my limits to convert it to meters (after all, I do not work for NASA). :)/Lars

[Guest luissa]

Hi Lars,Please see:http://forums.avsim.com/dcboard.php?az=sho...pe=search#11001Regards, Luis

[Guest JohnC]

Lars, I can't remember the spherical geometry formula for the distance between two points, but I DO remember there were a number of terms which contained the sines of angles VERY close to 0 with numbers like 0.00000000000001265 and the cos of angles VERY close to 1 like 0.9999999999749. Very small angles are subtended for most practical purposes when measuring small distances on the surface.So unless you use (VBA?) code which accomodates a very large number of decimal points, the final values will be off by a significant percentage.I assume you need real precision here that works across the whole planet.......:-bluegrab

[rhumbaflappy 113]

Hi Lars.Here's a javascript website to calculate great circle distances for any 2 lat-long pairs:http://williams.best.vwh.net/gccalc.htmThe sourcecode can be viewed in Internet Explorer, and the author gives his e-mail address.You can save this webpage from IE, and you can then use it offline, if desired.Dick

[Guest JohnC]

This guy Williams must sleep with a globe beside his bed.........his paper was the one I used to transport the formula to Excel but found that small distances were not accurate...or at least they didn't agree with MicroSoft Streets and Trips.

[Guest lmoelleb]

Hmm, strange I managed to avoid that thread when searching, I must have typed wrong in the search :)As far as I can see, this formula is close to what I am having, except I tried to use the values of the SDK, so that might be the problem.I use the .Net framework so if needed I can use the decimal calss, but a double should do the trick. Notice that only one COS is used based on the latitude (not the distance between the points) so it should be fine.Thanks for the help, I will move on when time allows (long weekend coming up, so there is hope). Worst case I will just do some "reverse" calculation to find the required numbers, whatever they might be. :)/Lars