I digitized contour lines from topographic maps and created a DEM. With the DEM, I estimated the boundary of a catchment area. Estimated positional accuracy of topographic maps is RMSE = 65 m, with a systematic bias towards East. Using an affine transformation (R-package vec2dtransf), I shifted the data towards West. The RMSE of the affine transformation is 30 m.

Visual check of the boundary using satellite images suggests that its actual position is more accurate than the former. However, I understand that there still is an uncertainty about the correct position and shape of the boundary.

I will use the estimated boundary to limit the area were to make predictions of soil properties. But I want to be more confident that predictions cover the real area of the catchment. Thus, I am planning to increase the area adding a buffer equal to the RMSE of the affine transformation (30 m).

I like to know if this approach is correct and if I am missing something.

  • See also grasswiki.osgeo.org/wiki/Contour_lines_to_DEM – markusN Mar 15 '14 at 22:53
  • It's hard to understand what problem you want to solve. I suggest you add screenshots that allow us to figure out the problem. By the way, the RMSE you are getting can be reduced if you want. Just try to adjust your control points and calculate once more the parameters of the transformation. It's a trial-and-error process. – Germán Carrillo Dec 29 '14 at 2:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.