adjacent VIIRS granules cannot be merged

Errors and unexpected results
Post Reply
User avatar
hproe
Posts: 504
Joined: Sat Nov 27, 2010 3:46 pm

adjacent VIIRS granules cannot be merged

Post by hproe »

Hi -

Every now and then I encounter adjacent VIIRS granules (from CLASS) that cannot be merged. Just now (using build from 2014-05-27 15:26) I have stumbled again over such a case. Please tell me where to stage the two GMODO files.

HP
User avatar
jayh
Posts: 424
Joined: Thu Jan 15, 2009 10:34 pm

Re: adjacent VIIRS granules cannot be merged

Post by jayh »

Hi HP-

You can post them on our ftp site:
ftp ftp.ssec.wisc.edu
login is anonymous/your email
bin
cd pub/incoming
put the files here and let me know the filenames.

Thanks, Jay
User avatar
hproe
Posts: 504
Joined: Sat Nov 27, 2010 3:46 pm

Re: adjacent VIIRS granules cannot be merged

Post by hproe »

Hi Jay -

The filenames are uploading. Their names are:
GMODO-SVM14-SVM15-SVM16_npp_d20140521_t1303091_e1308495_b13282_c20140528080130405816_noaa_ops.h5
GMODO-SVM14-SVM15-SVM16_npp_d20140521_t1308507_e1314311_b13282_c20140528080243792348_noaa_ops.h5

thanks, HP
User avatar
jayh
Posts: 424
Joined: Thu Jan 15, 2009 10:34 pm

Re: adjacent VIIRS granules cannot be merged

Post by jayh »

Hi HP-

Those two granules are being displayed together fine for me. I've tried 1.4, the 5/20 nightly, and the 5/28 nightly on a Mac and Windows 7 machine. I just used the band 14 BrightnessTemperature for my display. I displayed with and without a rubber band box in the region selector.

Bob and I don't think there was an issue with the 5/27 nightly, so we aren't sure what could be occurring. Is there a more specific case where you are seeing your error?

Thanks, Jay
User avatar
hproe
Posts: 504
Joined: Sat Nov 27, 2010 3:46 pm

Re: adjacent VIIRS granules cannot be merged

Post by hproe »

Good people -

Sorry, I fear I was not clear enough- The problem is with the rubber band. I cannot extend it across the two granules. The box can only be opened in the earlier granule. It disapperas when I try to stretch it into the later granule. The problem sticks to the specific granule pair across sessions and earlier builds. In the same session I can work always with other granule pairs where the rubber band works as expected. I got problematic granule pairs irrespective of channel type (I, M or DNB), but have not been able to spot any systematic behaviour that would produce them. Just guessing, it might have to do with the selected geographical area, i.e. with granule straddling the 0deg longitude (like the staged example. I will be keeping an eye on future downlaod.

HP
User avatar
jayh
Posts: 424
Joined: Thu Jan 15, 2009 10:34 pm

Re: adjacent VIIRS granules cannot be merged

Post by jayh »

Hi HP-

Thank you for the clarification. However, Bob and I are not seeing the issues you are with these granules. Bob installed the 5/27 nightly and like me, tried a variety of displays and rubber band selections, and we were successful. I've attached pics at the bottom of the forum.

We were trying to think of what would cause this behavior to occur on your end and not ours. If you could try these files on a fresh session or with a fresh McIDAS-V folder?

We certainly have documented these issues in other inquiries with other NPP files, so we are aware that these type of problems are present:
http://mcidas.ssec.wisc.edu/inquiry-v/?inquiry=1704 (similar problem with rubber band selection when a scan line is missing)
http://mcidas.ssec.wisc.edu/inquiry-v/?inquiry=1333 (we do have examples of dateline issues from this inquiry)

Hopefully we can diagnose why we are getting different results.
Thanks, Jay
Attachments
display.png
region_selection.png
User avatar
hproe
Posts: 504
Joined: Sat Nov 27, 2010 3:46 pm

Re: adjacent VIIRS granules cannot be merged

Post by hproe »

Okay Jay, you have convinced me that my problem must have local roots. In fact, starting with a fresh McIDAS-V folder the rubber band behaves normal again! I will go through my McIDAS-V folder in order to find the culprit.

HP
User avatar
jayh
Posts: 424
Joined: Thu Jan 15, 2009 10:34 pm

Re: adjacent VIIRS granules cannot be merged

Post by jayh »

Hi HP-

Glad you were able to get things working. Please let us know what you find. I'll make the programmer aware and see if he has any thoughts as well.

Thanks, Jay
User avatar
hproe
Posts: 504
Joined: Sat Nov 27, 2010 3:46 pm

Re: adjacent VIIRS granules cannot be merged

Post by hproe »

Hi Jay -

After starting with a fresh McIDAS-V having found area selection okay as reproted before, I loaded the VIIRS and RGB plugins, the minimum of add-ons needed for the current work. And the problem re-appeared. I then deleted the McIDAS-V folder, re-installed the most recent build ov McIDAS-V and restarted the notebook. I then boldly re-installed my old McIDAS-V folder and after having done some 10 VIIRS RGBs area selection across adjacent granules still works fine. Obviously I am very pleased with the result, though I do not understand what was going on. Hopefully, the problem does not creep up later on.

HP
Post Reply