MODIS Level-1 processing problems

Post Reply
gumley
Posts: 37
Joined: Thu Feb 18, 2010 2:36 pm

MODIS Level-1 processing problems

Post by gumley »

Dear IMAPP Colleagues,

I recently received a report from a Terra direct broadcast user about a problem with processing of Terra MODIS data. He reported that Terra MODIS Level 1 processing was failing in the geolocation stage. In the event that other users are experiencing the same problem, here is a synopsis what happened.

1. The problem occurred with Terra processing only (Aqua works fine).

2. The geolocation software version is IMAPP MODIS Level 1, version 1.5 from 2003 (more on this in a moment).

3. The files named "leapsec.dat" and "utcpole.dat" were both updated recently (these files can cause problems if they are not updated regularly).

4. The latitude and longitude values computed for Terra in the geolocation processing were all set to missing values (e.g., -999).

After I provided some guidance, the user checked some of the parameter limits set for Terra in the geolocation processing. The conclusion was that the lower limit for Terra's angular momentum Z component needs to be reduced in the file named

geolocation/static/Terra/GEO_parameters.dat

In particular, this line

ang_mom_z_limit = -7.60E9 -7.50E9

should be changed to

ang_mom_z_limit = -7.94E9 -7.50E9

After the user made this change to GEO_parameters.dat, he was able to process all recent Terra MODIS data with no problems.

However, this is a short term fix at best, because the IMAPP team here at SSEC no longer supports the IMAPP MODIS Level 1 geolocation or calibration software (which was last updated in 2004). MODIS Level 1 processing software for Terra and Aqua direct broadcast applications has been provided since 2005 by the MODISL1DB package developed and maintained by the NASA Ocean Color Biology Processing Group:

http://seadas.gsfc.nasa.gov/modisl1db/

MODISL1DB contains the operational version of the MODIS Level 1 geolocation and calibration algorithms (i.e., the same version used at NASA GSFC to process global MODIS data). It is also a subset of SeaDAS, so if you already use SeaDAS in a DB environment, then you have all the functionality of MODISL1DB.

That said, I understand that there are DB users out there who may have limited experience with Linux systems (MODISL1DB only runs on Linux), or who may be running IMAPP processing on a Windows system. If you are in this situation, I recommend that you investigate the IMAPP Virtual Appliance, available at

http://cimss.ssec.wisc.edu/imapp/imapp_va_v1.0.shtml

The IMAPP Virtual Appliance provides a complete processing solution for Terra and Aqua MODIS, AIRS, AMSU, and AMSR-E, and it runs on Microsoft Windows, Linux, or Mac platforms. It can run on a very modest Windows PC, with just 1 CPU and 512 MB of RAM, and it can be fully automated to process data in real-time. It also provides built-in automatic updates for items such as leapsec.dat/utcpole.dat files, calibration lookup tables, and even new versions of processing software (e.g., MODISL1DB). The IMAPP Virtual Appliance *is* a supported project here at SSEC, and we can provide help and advice on how to integrate it into an existing processing system.

Regards,
Liam.
egarbi
Posts: 6
Joined: Tue Sep 28, 2010 9:45 am

Re: MODIS Level-1 processing problems

Post by egarbi »

We have here a server running old version of imapp and we was not able to process terra data since Dec 14, now problem has been solved.
Thanks a lot
Post Reply