[GRASS-dev] i.topo.corr problems

Markus Neteler neteler at osgeo.org
Mon May 16 01:25:51 PDT 2016

On Mon, May 2, 2016 at 8:28 PM, Manamana <joanna.mardas at wp.pl> wrote:
> I have a problem with i.topo.corr
> Data:
> Landsat 7 ETM+ WGS 84/zone38N
> SRTM 3arc WGS84
> What I did: I did the atmospheric correction of landsat bands with dos1 I
> did the illumination model for SRTM, in a different location and mapset
> since it is in a different coordinate system and then I used r.proj to
> import the illumination model to the mapset with my landsat data. But! When
> I run i.topo.corr the results are not as they should be. The output map is
> in dark grey, almost black, just one colour. I thought that maybe it is
> because the resolution of SRTM is different than resolution of Landsat.

... this is possible.

My suggestion: Instead of processing SRTM in a different location,
import it right away into the UTM38N location of your Landsat data
through reprojecting on the fly with bilinear resampling.
Extra hint: Avoid SRTM 3arc but use the newest 1 arc-second for global
coverage (~30 meters).

You get it from EarthExplorer (http://earthexplorer.usgs.gov/) which
can be used to search, preview, and download Shuttle Radar Topography
Mission (SRTM) 1 Arc-Second Global data. The collections are located
under the Digital Elevation category. [1]

GRASS 7.1.svn (nc_spm_08_grass7):~ >

# set computational region to e.g. Landsat channel 3:
g.region raster=lsat7_2000_30 -p

# Import with reprojection on the fly. Recommended parameters:
# resample   Resampling method to use for reprojection -  bilinear
# extent   Output raster map extent
#    - region: extent of current region
# resolution   Resolution of output raster map
#    - region: current region resolution - limit to g.region setting from above
r.import input=n35_w079_1arc_v3.tif output=srtmv3_resamp resample=bilinear \
  extent=region resolution=region title="SRTM V3 resampled to Landsat7

# beautify colors:
r.colors srtmv3_resamp color=elevation

Keep in mind that SRTM is space borne.

I have now added a similar example to the r.import manual page (will
show up online in a few days) and also some Wiki pages.

Hope this helps,

[1] Pro tip:
Power users who know the tile name may grab it directly from this web
server (avoid to look at the entire directory in a Web browser, it may
overload it since the page is huge!

eg North Carolina tile (note that it is in HGT format rather that
GeoTIFF as obtained from EarthExplorer):

More information about the grass-dev mailing list