Earthdata Forum

Hello,

we downloaded and installed successfully SeaDAS 7.4, but running it on a VIIRS L1A file leads to the following error message :

VcstCalNetcdfInputItem::readNetcdfDynamic() Failure opening LUT file: /home/cats_qo/app/tools/seadas/current/ocssw/run/var/viirsn/cal/OPER/VIIRS_NPP_CAL_RSB_DYNAMIC_LUT_vxx.xxx.xx.xx_obpg.nc
VcstCalNetcdfInputItem::getData(iet) Failure reading netCDF dynamic file for LUT F_PREDICTED_LUT
ViirsGranule::read_lut() Error reading LUT data for group name F_PREDICTED_LUT
ViirsGranule::initialize_LUT_data() Failure reading LUTs.
ViirsGranule::LUT initialization failure
VcstViirsCal::Granule initialization failure

Error in initialization.

Hereafter is the content of the "ocssw/run/var/viirsn/cal/OPER" directory :
VIIRS_NPP_CAL_DNB_DYNAMIC_LUT_vxx.xxx.xx.xx_obpg.nc
VIIRS_NPP_CAL_RSB_DYNAMIC_LUT_vxx.xxx.xx.xxc_obpg.nc
VIIRS_NPP_CAL_STATIC_LUT_vxx.xxx.xx.xx_obpg.nc
VIIRS_NPP_CAL_STRAYLIGHT_DYNAMIC_LUT_vxx.xxx.xx.xx_obpg.nc
VIIRS_NPP_CMN_LUT_vxx.xxx.xx.xx_obpg.nc
VIIRS_NPP_GEO_LUT_vxx.xxx.xx.xx_obpg.nc

Apparently, SeaDAS 7.4 expects VIIRS_NPP_CAL_RSB_DYNAMIC_LUT_vxx.xxx.xx.xx_obpg.nc but this file is not present (replaced by "vxx.xxx.xx.xxc")

Is it possible to copy the required file from the 7.3.2 SeaDAS version and use it with SeaDAS 7.4 (or rename the "vxx.xxx.xx.xxc" in "vxx.xxx.xx.xx") ?

A second problem is with the installation of aqua data, that failed :

./install_ocssw.py --install-dir=/home/cats_qo/app/tools/seadas/7.4/ocssw   --git-branch=v7.4  --aqua --terra --oli --viirsn  --olci
...
...

Installing ocrvc (5 of 20)
Installing modis (6 of 20)
Installing modisa (7 of 20)
Installing hmodisa (8 of 20)
Installing modist (9 of 20)
Installing hmodist (10 of 20)
Installing oli (11 of 20)
Installing olci (12 of 20)
Installing viirsn (13 of 20)
Installing bin (14 of 20)
Installing bin3 (15 of 20)
Installing scripts (16 of 20)
Python version acceptable
Installing aqua-luts (17 of 20)
Error downloading MYD02_Emissive_LUTs.V6.1.35.30_OC2.hdf
Could not communicate with the server.
Please check your network connections and try again later.
Error - Could not install luts for aqua

Thank you for your help,

Jacques

I'm sure someone at NASA can explain the difference between the vxx.xxx.xx.xx and vxx.xxx.xx.xxc files, but since mismatches like the one you have are a problem at my site, I'll pass on what I have learned.

The NetCDF metadata for V2017124173600.GEO-M_SNPP.nc has
:source = "V2017124173600.L1A_SNPP.nc,
VIIRS_NPP_CAL_STATIC_LUT_vxx.xxx.xx.xx_obpg.nc,
VIIRS_NPP_CAL_RSB_DYNAMIC_LUT_vxx.xxx.xx.xxc_obpg.nc,
VIIRS_NPP_CAL_DNB_DYNAMIC_LUT_vxx.xxx.xx.xx_obpg.nc,
VIIRS_NPP_CAL_STRAYLIGHT_DYNAMIC_LUT_vxx.xxx.xx.xx_obpg.nc,
VIIRS_NPP_CMN_LUT_vxx.xxx.xx.xx_obpg.nc,
VIIRS_NPP_GEO_LUT_vxx.xxx.xx.xx_obpg.nc,
polar_wander.ascii" ; so VIIRS_NPP_CAL_RSB_DYNAMIC_LUT_vxx.xxx.xx.xxc_obpg.nc could be the one we should use, at least for recent files. On this system, however, /run/data/viirsn/instrument_defaults.par has the (old?) VIIRS_NPP_CAL_RSB_DYNAMIC_LUT_vxx.xxx.xx.xx_obpg.nc version, which is
also present on my system.   You could edit $OCSSWROOT/run/data/viirsn/instrument_defaults.par to refer to the vxx.xxx.xx.xxc version.  It could be
informative to compare results for the two versions.

As for the MODIS Aqua data -- more often than not install_ocssw.py fails at Installing aqua-luts.  I just run update_luts.py -v aqua until it succeeds, and make note of the files it downloaded.  Unfortunately, update_luts.py needs to edit $OCSSWROOT/run/data/modisa/msl12_defaults.par.   This means git status should show the file as "modified":
$ cd $OCSSWROOT/run/data/modisa/
$ git status
On branch v7.4
Your branch is up-to-date with 'origin/v7.4'.
Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git checkout -- <file>..." to discard changes in working directory)

  modified:   msl12_defaults.par

no changes added to commit (use "git add" and/or "git commit -a").  To verify that the changes match the files shown in the putput from update_luts.py -v, use $ git diff msl12_defaults.par
diff --git a/msl12_defaults.par b/msl12_defaults.par
index 69fecce..b55e593 100644
--- a/msl12_defaults.par
+++ b/msl12_defaults.par
@@ -17,7 +17,7 @@ gain=[0.9731,0.9910,0.9935,1.0002,0.9994,0.9996,0.9998,0.9989,1.0]
offset=[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0]
xcal_opt=[1,1,1,1,1,1,1]
xcal_wave=[412,443,488,531,547,667,678]
-xcalfile=$OCVARROOT/modisa/xcal/OPER/xcal_modisa_axc26f
+xcalfile=$OCVARROOT/modisa/xcal/OPER/xcal_modisa_axc39d
pol_opt=3
polfile=$OCDATAROOT/modisa/cal/polcor_modisa_2010b
vcnnfile=$OCDATAROOT/modisa/ocr_vc/modisa_ocr_vc_nn
I check that $OCVARROOT/modisa/xcal/OPER/xcal_modisa_axc39d* matches the file sizes and list of wavelengths of the previous set, in my case, $OCVARROOT/modisa/xcal/OPER/xcal_modisa_axc38d*.

The correct fix for VIIRS L1A is to modify  $OCSSWROOT/run/data/viirsn/instrument_defaults.par and add the "c" like George suggested.   The new line will be:

rsb_dynamic_lut_file=$OCVARROOT/viirsn/cal/OPER/VIIRS_NPP_CAL_RSB_DYNAMIC_LUT_v2.0.0.2c_obpg.nc

The new file has slightly different Time_Dependent_RSB_F_LUTs which extend into the future and produce better matchups.
We are pushing this change to the git repository so future installs and updates will get the proper LUT.
Thanks for pointing this one out.

don