I am a bit new to GRASS, but I have managed to get a lot of good work done thus far. However, the following error is really throwing me off, because I can't find any help anywhere for it. I am using GRASS 7 that I downloaded today (9.23.13) on Windows 7.

Basically, I have a set of bands from a new Landsat 8 image acquired in May 2013. I want to convert these (GeoTiff format) to Top-of-Atmosphere using i.landsat.toar. According to what I've read, this should be the tool to use. I have the metadata file for the imagery and have made sure that i.landsat.toar is accessing it and that the computational region is set. However, I'm getting the following TWO errors.

Note this is the output from the tool's dialog box.

Error 1: The tool appears not to be picking up the Earth-sun distance or the Solar elevation angle from the metadata file

i.landsat.toar --overwrite --verbose input_prefix=LC80630462013143LGN01_B output_prefix=TOAR2_ metfile=C:\Users\Andrew\Desktop\All Projects\PTA\Mapping\Landsat\LC80630462013143LGN01_MTL.txt sensor=ot8

    Metada file is MTL file: new format
    RADIANCE & QUANTIZE from data of the metadata file
    ESUN evaluate from REFLECTANCE_ADDITIVE_FACTOR_BAND of the metadata file
     ACQUISITION DATE 2013-05-23 [production date 2013-05-28]
       Earth-sun distance    = 0.00000000
       Solar elevation angle = 0.00000000
       Atmospheric correction = UNCORRECTED

Error 2: Because of this (?) it thinks that BAND 1 (and all the rest of the bands) have calibrated DN's in the range 0 to 0.

     BAND 1  (code 1)
       calibrated digital number (DN): 0.0 to 0.0
       calibration constants (L): -0.000 to 0.000
       at-sensor radiance = -0.00000000 * DN + -0.000
       mean solar exoatmospheric irradiance (ESUN): -0.000
       at-sensor reflectance = radiance / -0.00000
     BAND 2  (code 2)
       calibrated digital number (DN): 0.0 to 0.0
       calibration constants (L): 330444430637036151551722492294
    49817417372212139642585088.000 to 0.000
       at-sensor radiance = -0.00000000 * DN + -0.000
       mean solar exoatmospheric irradiance (ESUN): 0.000
       at-sensor reflectance = radiance / -0.00000
  • Other people tell me ESUN is not being published yet for L8. gisagmaps.com/landsat-8-atco – Willy Sep 24 '13 at 2:06
  • No the ESUN is definitely in the metadata file, it just isn't being picked up by GRASS and i.landsat.toar – user22248 Sep 26 '13 at 17:42
  • The version included in GRASS 6.4.svn has been updated for the changed metadata format the other day. The update for the version in GRASS 7.svn is on the way. – markusN Oct 1 '13 at 19:54
  • Also the GRASS GIS 7 version should be fine now as Oct 15th 2013. – markusN Oct 17 '13 at 20:23
  • Any news on this? – BillyWill Apr 29 '14 at 5:41

Right now I have checked into SVN a set of updates for i.landsat.toar into SVN for Landsat8, provided by the author E. Jorge Tizado. They hopefully solve your problem (NASA changed the metadata format).

You can download the updated winGRASS binary by tomorrow morning either through the OSGeo4W installer ("advanced") or as standalone daily snapshot version from here.

| improve this answer | |
  • 1
    Also the GRASS GIS 7 version should be fine now as Oct 15th 2013. – markusN Oct 17 '13 at 20:21

As a suggestion, because the program is reporting that the file is a "new format", I would try comparing this metadata file with a LANDSAT 7 metadata file.

If you can identify differences, you may be able to tweak the LANDSAT 8 file into something the program can interpret.

| improve this answer | |
  • 1
    We have updated the GRASS module (hopefully) accordingly. – markusN Sep 28 '13 at 17:27

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.