[GRASS-user] i.atcorr negative DEM

[Markus Metz]

On Tue, May 17, 2011 at 9:22 PM, Elena Mezzini <elena.mezzini at gmail.com> wrote:
> Hi all,
> I'm once again in trouble with i.atcorr in GRASS6.5.svn.
> I notice that my output reflectance images are strange: there are some areas
> that are really bright and other that are really dark. Both seem to match
> with my DEM raster.
> Could there be problems with negative values in DEM?? I'm in a coastal area
> and there are some real altitude negative values.

i.atcorr (the 6S code it is using to be exact) should treat negative
elevation values as elevation = 0. It is debatable if this is correct,
but in theory i.atcorr should behave accordingly. Actually, I am not
sure about it's behaviour for negative elevation values...

You could set all negative elevation values to zero, run i.atcorr
again and compare the results to the i.atcorr results using the
original DEM. If there are any differences that do not make sense,
please fill in a bug ticket.

Markus M