L1B problems

[cortivo]

I'm trying to process some L1 data but a have some problems in the L1B step...
I'm starting from L1 and the first step is to create a GEO file and at the end of the process seadas tell me that is ok!

However when I try to do the L1B process seadas tell me that the GEO file does not exists(but the GEO files is in the same folder where is te L1A file). It happens with all users with a dot in the user name...

Seadas searches for /home/pedro.****.GEO.L1A_LAC instead of for /home/pedro.paulo/****.GEO

What I'm doing wrong?

See the attached!

[gnwiii]

You aren't doing anything WRONG, just pushing the limits of historical conventions (which didn't allow "." in a username due to legacy usage: "chown user.group ..."; current usage has changed to "chown user:group ..."). Using an ocssw.test user account on macOS El Capitan can reproduce your problem.  There is no problem with command-line processing, so the broken test appears to be in the SeaDAS GUI.   When you push limits you have to accept some risk that you could encounter such problems and have to work around them.  If the bash shell isn't a good friend already there are excellent references and tutorials available in translations to many languages.

[OB SeaDAS - knowles]

I can confirm and replicate this.  We will look into it.  It is mistakenly interpreting what follows "." as the file extension and lopping it off.

For now, your choices are either:
   run from the GUI using the multilevel processor
   change "." to "_" in the directory name and run modis_L1B from the GUI
   run modis_L1B.py from the command line
  
Note: the extractor GUI will also fail with this naming convention.

I tested this case for the l2gen GUI and there is no problem there.

thanks
Danny

[OB SeaDAS - aynur]

Hi all,

Certain parts of the SeaDAS implementation made the assumption that there is no "." in a directory name. We will fix this in the next release, SeaDAS 7.5.

Thanks,

SeaDAS team

[byan0622]

Whether Seadas 7.4 can process viirs 1b data from MODAPS web site? e.g., NPP_VDNE_L1
VIIRS/NPP Day/Night Band 5-Min L1 Swath SDR 750m
NPP_VIAE_L1
VIIRS/NPP Imagery Resolution 5-Min L1 Swath SDR 375m
NPP_VMAE_L1
VIIRS/NPP Moderate Resolution 5-Min L1 Swath SDR and GEO 750m.

In fact, I opened one file in the above format using SeaDas7.4. I can see imagery but geolocation information is either wrong or missed. Would you please guide how I can use SeaDas7.4 to process viirs 1b data? Thanks.

[OB.DAAC - SeanBailey]

Use the NASA VIIRS L1A product distributed but the OB.DAAC.  The SeaDAS processing code includes support for this version of VIIRS data, but not the SDR data distributed by MODAPS.

Sean

[byan0622]

I visited the OB.DAAC for viirs sdr list. Is it VIIRS/NPP Imagery Resolution 6-Min L1B Swath SDR 375m? Does it contain all M abd I band of radiance or reflectance data? Thanks.

[OB.DAAC - SeanBailey]

The VIIRS L1A product we distribute contains the full sensor suite of bands.  SeaDAS is distributed with code to process the L1A to L1B (calibrate_viirs) and can output the IMG(375m), MOD(750m) and DNB (day/night panchromatic band) data, as well as the geolocation file (geolocate_viirs).  The L1B data are reflectance stored as scaled integers,  but the variable metadata also includes scale factors to convert to radiance if desired.

Sean

[byan0622]

We agreed with you. However, it seems that SeaDas tool is unable to process viirs l1b data to generate spectra and display lat./long of the pixels as it does for MODIS/Aqua or Terra. Is this true? Thanks.

[byan0622]

I meant that we have looked at the output data of viirs 1b from SeaDAS. The radiance output are fine, but there is no latitude and longitude data. It sounds like the VIIRS L1B data does not contain geometry data, SeaDAS only output the pixel numbers (line and column). Is this correct? My question is how we can generate spectra of selected pixels (pins) and how we can output reflectance with geolocation information as we can do for Aqua and Terra? Thanks.