2

I have done several algorithms for topographic correction in R including simple cosine, Minnaert, SCS. The input data aspect and slope for the hillshade raster have correctly been done in radians. The hillshade has some spots where the cos (i) is slightly below zero and these spots appear black in the corrected Landsat image. The white surroundings of the black spots are zero or slightly above. That means that cos(i) is around 90° or above in these areas. Such areas receive no direct radiation from the sun but diffuse light.

Can somebody help me with this - I mean do you know an algorithm to solve this problem? Is this phenomenon called overcorrection?

Thanks a lot S.

3

I solved this problem via the C-Correction (see literature). The C is calculated with the help of a linear regression including the raster data of the Landsat image to correct and the raster data of the hillshade for the scene (C=b/m, b=Interzept, m=gradient). This algorithm is suitable to avoid the black spots (after correction) in the areas illuminated only with diffuse light and no direct sun light.

This is my second question I answer myself. I cannot believe that there is no other person who is able to answer this question in this forum...? It seems to me that in this forum are only people who ask questions and none who answer ;-)

  • Because there was no answer I had to find the solution by myself. – Stacky Jul 8 '13 at 13:15
  • 1
    Also a link to the C-correction would be helpful for future reference. – Kirk Kuykendall Jul 8 '13 at 14:15
  • Easy to find with any search engine...there are several papers to find. – Stacky Jul 12 '13 at 10:44
  • Thats strange, I have had a lot more artifacts with C-correction – mace Mar 9 '16 at 12:43

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.