# Calculate area of ellipsoidal polygon with QGIS

There is online service of GeographicLib, by Karney, where one can calculate the area of ellipsoidal polygon without need of reprojection.

Can one calculate the area of geometry in QGIS so as ellipsoidal polygons (in order to get the most accurate area), and not with `\$area` that gives planar area after a equal-area projection of a geometry?

Internally, QGIS uses GeographicLib to calculate Area for a polygyon on an ellipsoidal geodesic using the WSG84 Geodesic definitions.

You should be using an `EPSG:4326` projection with lat/lon coordinates, or some other ellipsoidal projection.

There are two ways to access the ellipsoidal geodesic area:

1. Using the `Identify Feature` tool, click on the polygon of interest and look under the "Derived" data. It should read "Area (Ellipsoidal - EPSG:7030)".
2. Use the "Field Calculator" and make use of the `\$area` variable. (Note the use of the dollar sign `\$` - this is required for getting the ellipsoidal area instead of the planimetric area)

Simple answer: no. QGis is a purely cartographic programme, all the computations it performs are carried on the Cartesian plane. Moreover, QGis erroneously assumes the counter-domain of any cartographic projection to be an infinite plane, which may invalidate its geometric computations in the vicinity of counter-domain edges (what are sometimes called "interruptions").

• QGIS uses an ellipsoid for area calculations. Doesn't that mean it assumes the earth's surface is not flat, but three-dimensional? Please provide some references to support your assertion that QGIS calculates area on a Cartesian plane.
– csk
Commented Dec 7, 2018 at 19:02
• @csk It is up to you to prove that QGis is a geodetic programme, rather than geographic. Open the manual and try to find a section on Geodesy. Commented Dec 9, 2018 at 14:04
• Doesn't this suggest, directly from Martin Dobias, that QGIS does calculate on an ellipsoid? Commented Feb 5, 2019 at 16:27
• @Gabriel C. Why don't you give it a try? Download the world administrative borders form Natural Earth. Open that dataset in QGis with Marinus of Tyre's projection and check for instance the area of Greenland. Then reproject the layer with an equal area projection, say the Sinusoidal. Open the reprojected layer in QGis and query again the Greenland feature. Commented Feb 6, 2019 at 8:48
• I just did and I'm still unconvinced: I calculated Greenland's area 3 times from that shapefile. Once while project CRS was set as equidistant cylindrical (ellipsoid automatically set as none/planimetric) which results in 8229000.9328 sq.km. I then set the project CRS to earth sinusoidal (ellipsoid set to Earth 2000) which results in 2171418.4078 sq.km, only 5000 km off from the official figure which could easily be explained by generalization of the coastline. Using same projection but setting ellipsoid to none/planimetric, the result was 8229000.9328 sq.km. Commented Feb 6, 2019 at 14:38