S3A mosaic problem

Dear All,

I tried to create a mosaic over Sentinel 3A products (OGVI).

  • 1th image (on the left) represents the original file with original projection;
  • 2nd image represents the OGVI output obtained from SNAP using SUBSET (over OGVI) and then MOSAIC with specific boundaries to speed up the process;
  • 3th image represents a REPROJECTION using original file with preserve resolution checked from original S3A file;
  • last image shows the correct output reprojection obtained using MOSAIC over the original S3A file;

It seems there is a problem because there are some missing information over France.
Any idea about the correct metodology to do a mosaic?

Thanks in advance for your collaboration,
Monica

Hi Monica,

In the mosaic function you have a label with Variables & Conditions, there you have to select with variables you want to have as bands in your output mosaic and the combination (AND/OR). Have you checked that you are using the right logic?

The only check done was on valid values of OGVI (!=NaN).
Here another example:


where on reprojected image there are a missing area.

Is there a problem on mosaic procedure?

I have similar problem to mosaic 2 scenes of S3 OLCI. I get the mosaic but i can not see adg443 band mask. I have the raster but only if use adg443 color classification show me same value for all pixels. :S

Can you post some image which show the problem?
Also your configuration of the mosaic might be of interest.

Thanks, attached the picture. I use snap gui interface but i would like build gpt code to next steps.

The image looks really good and I don’t see any issue.

For starting with gpt you can have a look at this guide:
https://senbox.atlassian.net/wiki/spaces/SNAP/pages/70503590/Creating+a+GPF+Graph

Type for example

gpt mosaic -h

For getting help on the mosaic operator.
However, the mosaic is a bit complex to configure.
You can save the configuration from the GUI of the Mosaic (File/Save Parameters…) or
you use the GraphBuilder, save the graph and andapt later the xml file.

Yes, is really nice the mosaic, but the flags (clouds, land) show the major index (red) of chla.

Thanks for help. I will post the result soon.