5

In QGIS i created a grid (via "create grid"), which i later split into layers of horizontal and vertical lines. Afterwards i calculated the respective cumulative line length for polygons using "Sume line length". The results i achieved differ especially for horizontal lines quite strongly (~ 25% less) from expected values (in the case where i could compare, based on another publication (values for China)).

One suggestion was, that the difference may be due to measuring planar vs. geodetic distance. Does anyone know how QGIS measures distance with this tool? May the result be subject to the projection I use?

5
  • 1
    I think when using any kind of distance/areal calculations, the CRS plays a very significant role. My guess is that the differing results you obtained could be due to whichever CRS you used.
    – Joseph
    Commented Aug 1, 2016 at 14:52
  • what CRS are you using on the grid layer, and where are you getting the figures which you're comparing it with?
    – Steven Kay
    Commented Aug 1, 2016 at 19:40
  • Mollweide and World Mercator, projected as otherwise sum line length wouldn´t work. From this paper: pnas.org/content/109/26/10263.full The way i made the grid varies, but this should actually not explain the quite high differences. I just made a 25 Km grid. While vertical is only slightly higher ~10%, horizontal is ~25% lower, hence the ratio almost one, so I assume theirs a hidden problem which i cant find. What is further interesting choosing Mollweide or Mercator in my approach had only a very marginal effect. Commented Aug 2, 2016 at 8:31
  • 1
    @PhilippDahlem, see this answer by @Underdark. Sum line length uses layer CRS.
    – dof1985
    Commented Aug 3, 2016 at 9:12
  • check this comment. gis.stackexchange.com/questions/123036/…
    – Fran Raga
    Commented Aug 8, 2016 at 11:32

0