Rainbows

Post Reply
arb
Posts: 14
Joined: Wed May 29, 2013 12:41 pm

Rainbows

Post by arb »

Hi, I'm seeing some odd artefacts whenever I run polar2grid.
It looks as though areas which are black or white are being turned into rainbows, see attachment.
Does anyone recognise this problem?
Attachments
npp_viirs_true_color_crefl_None_20131127_124307_tco_lcc.a.gif
npp_viirs_true_color_crefl_None_20131127_124307_tco_lcc.a.gif (153.09 KiB) Viewed 19969 times
davidh
Posts: 116
Joined: Tue Jun 04, 2013 11:19 am

Re: Rainbows

Post by davidh »

We've noticed this too. As far as I can tell it has to do with the corrected reflectance algorithm (which I don't really have control over). In past cases I've looked at the bands used by CREFL and didn't see anything out of the ordinary so I have always blamed CREFL. Let me know if you notice anything that might give us an idea what's causing this. In the mean time I'll look at a case that Kathy gave me today and see if I find any answers.
kathys
Posts: 487
Joined: Tue Jun 22, 2010 4:51 pm

Re: Rainbows

Post by kathys »

We think we have found the problem you are seeing and the solution. The problem occurs when we are trying to combine bands to create the RGB file that have different regions missing data. If you try to create an RGB using CREFL input files where a FOV has 2 or 1 band missing, you will get these types of colors. The solution is to mask FOVs where 1 or more of the bands you are using to create the RGB are missing. We are working on a patch to the polar2grid software that will fix this. Sorry for any problems this might have caused.

Kathy Strabala
jenniffer07
Posts: 1
Joined: Mon Jan 27, 2014 3:56 am

Re: Rainbows

Post by jenniffer07 »

i think its not an odd artefacts but it may happen in some circumstances.. :)
kathys
Posts: 487
Joined: Tue Jun 22, 2010 4:51 pm

Re: Rainbows

Post by kathys »

This issue has been fixed in the polar2grid version 1.2 software release. The problem was that incorrect saturation and fill values were being stored in the crefl files. The bug was corrected, and this updated version of the crefl software is now executed as part of the crefl2gtiff.sh processing script. We have also notified the crefl software developer about this problem.

Kathy
Louise
Posts: 1
Joined: Thu Oct 08, 2015 3:57 am

Re: Rainbows

Post by Louise »

really different one!!!!!
arb
Posts: 14
Joined: Wed May 29, 2013 12:41 pm

Re: Rainbows

Post by arb »

I'm still seeing this type of artefact in polar2grid v2.0, see attachment.
Not in bright areas like previously but in dark sea between bright clouds, so it looks like a polar atoll!
Attachments
viirs_201511131226_rgb_crop1.jpg
viirs_201511131226_rgb_crop1.jpg (19.06 KiB) Viewed 19581 times
davidh
Posts: 116
Joined: Tue Jun 04, 2013 11:19 am

Re: Rainbows

Post by davidh »

Hi Andrew,

Would it be possible to get access to the data for this case too, like the data drop out problem you saw? My guess is something with remapping and combining the various resolutions that are overlapping oddly. One thing you could try until I have a chance to look at the data is running the P2G commands but don't provide the I-band data. This will result in an unsharpened image. If this output image looks good then I'll have to see why the I and M band data differs so much.

Thanks,
Dave
Post Reply