I have polygons of countries with its administrative divisions, I would calculate the areas of each division in a country accurately.

I got the bounding boxes for each country, then I divided its width (longitudely) by its height (latitudely).

If the

  • ratio = 1, the country extends equally South-North as West-East
  • ratio < 1, it extends West-East
  • ratio > 1, it extends South-North

But I made my own criteria:

  1. If the ratio < 0.7, I used an Albers equal-area projection +proj=aea with the 1st SP parallel is 1/6 degree below the northernmost latitude of the country, and the 2nd SP is 1/6 degree above the southernmost latitude.
  2. If the ratio is between 0.7 and 1.3, I used Lambert azimuthal equal area +proj=laea centered at the center of the bounding box.
  3. If ratio > 1.3, I used transverse cylindrical equal-area +pro=tcea with central meridian is at the center meridian of the country and k_0=0.9996

My question Would someone approve it to me, that it lets me get accurate areas relatively better than a projection for the whole continent or the whole world, or isa it boring time taking method?

  • 1
    1) You need to decide what precision you are interested in, to square km, square m, square cm, square mm :-), 2) it is not clear what the source projection you have identified the shape and size of the States, they are all different projections; 3) would you like to calculate the exact area of the States with regard to elevation +another accuracy in height; 4) Knowing the shape of States, I would choose one or two equal-area projection and bore their original dates, for example, the centroids of each state, and then calculated its area, in any case, continue with your experiments... – Cyril Mikhalchenko Dec 5 '18 at 11:11
  • My precision would be sufficient in sq.meter, I think, I would have large difference in area for large countries (e.g. Russian Federation) if I follow my criteria, bcz of its large administrative divisions (for example Sakha Republic is so large) but for european countries is better – user9322960 Dec 5 '18 at 11:25
  • In order to perform the calculation with such accuracy, you must have the original cartographic data with an accuracy of at least 1:10,000 scale on the whole Earth, a cheerful picture is obtained... – Cyril Mikhalchenko Dec 5 '18 at 11:34