Hi Paul,

Thank you very much for the answer. But don't curse me yet :) I care, but I can probably work around it. I can modify my edge and add interior vertices that are on the wgs84 edge (tessellate) and thus reduce the error when projecting a point to the edge on the sphere.

--Hector

> Yes, you've got it right. The projected point is done in sphere world for

> speed, and then the point-to-point distance is done on spheroid. Most

> people don't care. Curse you for caring :)

> P

>> Hello,

>>

>> I am trying to verify the distance calculation (using geographies and

>> use_spheroid=true) between a line and a point.

>>

>> I have used GeographicLib to generate a geodesic (using the wgs84

>> ellipsoid) between points 0 0 and 10 50 (lng lat). I generate a point lies

>> on the edge (at the mid point) which is 3.<a href="tel:91451185%2025" value="+19145118525" style="font-size:12.8px">91451185 25.11977359. When I use

>> st_distance to compute the distance between the edge and the point I expect

>> the answer to be close to 0. But it is not.

>>

>> select st_distance(st_geogfromtext('SRID=4326;point( 3.91451185

>> 25.11977359 )'),st_geogfromtext('SRID=4326;linestring(0.00000000

>> 0.00000000, 10 50 )'), true);

>> produces: 274.54332271 meters. If I use use_spheroid=false the answer

>> is 274.10911308.

>>

>> Is this because the distance calculation between a point and a line is

>> done by projecting the point to the line assuming a perfect sphere? Or am I

>> doing something wrong?

>>

>> I am using Postgresql-9.5 and PostGIS-2.2 on Ubuntu 16.04.

_______________________________________________

postgis-users mailing list

[hidden email]
https://lists.osgeo.org/mailman/listinfo/postgis-users