[GRASS-dev] [GRASS GIS] #3469: i.atcorr: Sentinel-2 support broken on some systems

[GRASS GIS]

#3469: i.atcorr: Sentinel-2 support broken on some systems
--------------------------+-------------------------
  Reporter:  sbl          |      Owner:  grass-dev@…
      Type:  defect       |     Status:  new
  Priority:  normal       |  Milestone:  7.4.0
 Component:  Packaging    |    Version:  unspecified
Resolution:               |   Keywords:  i.atcorr
       CPU:  Unspecified  |   Platform:  Unspecified
--------------------------+-------------------------

Comment (by mlennert):

 Replying to [comment:10 mmetz]:
 > Replying to [comment:9 sbl]:
 > > Replying to [comment:8 mmetz]:
 > > > The symptom can be cured by using double precision floating point
 variables throughout. Currently, i.atcorr uses single precision floating
 point variables. When using double precision floating point variables,
 Debian testing produces reasonable results. However, these results differ
 from Fedora 27, also using double precision floating point variables
 throughout. I would say that both results (Debian testing and Fedora 27)
 are slightly wrong because of the numerical instability of the 6s code in
 combination with the sentinel-2 parameters.
 > >
 > > Would you be able to provide a patch with those changes? I would not
 mind slightly wrong results, just I am able to proceed. I can update the
 computation once this is properly fixed...
 >
 > Patch i_atcorr_double.patch attached. i.atcorr is identical in 7.4 and
 7.5, therefore the patch works with both versions.

 Thanks for investigating this !

 I now get what looks like more reasonable results:


 {{{
 r.info -r test_atcorr
 min=0.3878489
 max=8513.962
 }}}

 compared to the original before correction:


 {{{
 r.info -r B08
 min=83
 max=7110
 }}}

 Interestingly, i.atcorr also runs much faster with the patch.

--
Ticket URL: <https://trac.osgeo.org/grass/ticket/3469#comment:11>
GRASS GIS <https://grass.osgeo.org>