0

I have to find a circle with a certain radius that touches two points. I have the two points in a lat/lon form (EPGS:4326), and for the calculus I'm converting into Geocentric coordinates (EPGS:3857). Once I found the center, I calculate the distance with pitagora, and I get the exact distance, but when I convert back to lat/lon, And I calculate the distance with haversine or using geodesic (GeoPy) I get a distance much different than the one got in the others. (One example is 1600meters vs 1122 meters)

Where can be the problem? It's caused by the accuracy of the trasformation or other things?

GforGIS
  • 3,126
  • 3
  • 19
  • 38
GokuGunZ
  • 1
  • 3
    The problem is that Web Mercator (3857) is useless for distance. The poles are infinitely far from the Equator, so all distances not measured along the equator range from wrong, to very wrong, to absurdly wrong. You've basically described the exact problem with the projection itself, as a definition. There has got to be a canonical Question of which this is the duplicate, but I don't have the time to search for it right now. – Vince Mar 28 '24 at 02:46
  • 2
    Possible duplicate - https://gis.stackexchange.com/q/396004/2856 – user2856 Mar 28 '24 at 02:59

0 Answers0