[GRASS-dev] i.atcorr returns nan for Landsat 5 TM bands 1 and 2

[Alessandro Samuel Rosa]

Dear Grass Developers

I have been having a problem using module i.atcorr to process Landsat 5
TM bands 1 and 2. The resulting rasters are empty for these two bands.
For other bands the result seems to be correct. There are many posts
talking about this topic, but I did not find an answer. So I tried using
Grass 6.4 and Grass 7.0. Unfortunately the same error occurred. Next I
tried changing some parameters of the input file. What I found is that
changing the time of acquisition of the image the algorithm was able to
perform the calculations and return acceptable results. Bellow I show
the content of input file A, with which the module did not work, and
input file B, with which the module worked. Notice that the only
difference is an increase of one hour in the time of acquisition of the
image. By the way, it also worked if I decreased the acquisition time by
one hour. The same occurred to other images from other dates. From this
result I infer that the problem may rely on the calculations of sun
position.

I also tried contacting Mauro Antunes, the author of the original code
in Fortran. He was my teacher, so I have a close relation with him. I
asked him if he was interested in taking a look at the C++ code
implemented in GRASS, but he is not willing to do it. He has been
working on the Fortran code during the last years and provided me with a
copy. Using his Fortran code and the same input file that I used for
i.atcorr gave the expected results.

I hope this is of any help for fixing this old issue in the i.atcorr
module. Unfortunately I know neither C++ nor Fortran to help finding the
error.

The content of input file A is:

---
7                                                         # geometrical
conditions = Landsat 5 TM 
12 26 16.34 -54.125 -28.87    # month day hh.ddd long lat (hh.ddd is in
decimal hours GMT) 
1                                                         # atmospheric
mode = Tropical 
1                                                        # aerosols
model = Continental Model 
20                                                     # visibility [km]
(aerosol model concentration) 
-0.3                                                  # mean target
elevation above sea level [km] 
-1000                                              # sensor on board a
satellite 
25                                                     # band 1
(0.430-0.560)
---

The content of input file B is:

---
7                                                         # geometrical
conditions = Landsat 5 TM 
12 26 17.34 -54.125 -28.87    # month day hh.ddd long lat (hh.ddd is in
decimal hours GMT) 
1                                                         # atmospheric
mode = Tropical 
1                                                         # aerosols
model = Continental Model 
20                                                      # visibility
[km] (aerosol model concentration) 
-0.3                                                   # mean target
elevation above sea level [km] 
-1000                                               # sensor on board a
satellite 
25                                                     # band 1
(0.430-0.560)
---

Kind regards,
-- 

Alessandro Samuel-Rosa
---
PhD Candidate Graduate School in Agronomy - Soil Science
Federal Rural University of Rio de Janeiro
Seropédica, Rio de Janeiro, Brazil
---
Guest Researcher ISRIC - World Soil Information
Wageningen, the Netherlands alessandro.rosa at wur.nl | Phone 0031 06 4435
9563
---
Homepage: soil-scientist.net Skype: alessandrosamuel 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-dev/attachments/20140227/8f1fd341/attachment-0001.html>