Subset as input for SEN2COR


I am running SNAP 5 on UBUNTU 16.04 and since SNAP cannot run SEN2COR, I am running SEN2COR stand alone in command line. However standalone and command line SEN2COR asks for *.SAFE as input. The area I am working is small so I can subset Sentinel 2 image within SNAP but it saves as BEAM-DIMAP format only. How can SEN2COR get BEAM-DIMAP format as input? My image is too big for processing 10 m resolution with stand alone/command line SEN2COR yielding MemoryError…

Any ideas I can retrieve 10 m resolution BOA reflectances?

Many thanks.


it is not possible to make a subset before running sen2cor because as you said the .SAFE format is needed. You should apply the atmospheric correction and then subset your image.
In fact, sen2cor looks for some “special pixels” with an specific spectrum during its processing to compute some correction parameters so if it would be possible to subset the image before, the quality of the correction would be deteriorated.


Thank you.
As far as I could understand atmospheric correction must be applied for 60 m bands, then to 20 m bands and at last for 10 m bands.

I could retrieve 20 m bands for atmopheric correction but when I try to apply SEN2COR for 10 m, it presents MemoryError and no bands with 10 m resolution are retrieved.

What can I do to retrieve 10 m resolution bands?

I do appreciate any insight…
Thank you.


you can run directly Sen2Cor with a resolution of 10 m, but there is a minimum memory needed in your system, please see:

Depending on the product you want to analyze, perhaps you could find it already processed on Copernicus Open Hub:

1 Like

“you can run directly Sen2Cor with a resolution of 10 m” , i applied this to my image but when i try to save it in tiff format, i find wrong as the file is multisize so i make resample to 10 m again.
What do you think?