using sen2cor to convert TOA product into BOA product in sentinel 2A image from 2016 February

I have an image of Lahore city from sentinel 2A which dates to 2016 February 17. I used sen2cor algorithm in SNAP to convert the level 1C to level 2A product. But I kept receiving the empty folders. The processed image file appeared inside the (unzipped) original folder with 2A instead of 1C. Everything appeared normal except when I tried to open the processed image in SNAP, there was nothing inside the R10m, R20m and R60m folders inside IMG_DATA. There was an exception inside R20m folder though which had a H5 file format named “.database.h5”. Guessing that I might have done some mistake installing the sen2cor plugin, I copied the image in my friend’s laptop to try the process in her SNAP. (she had used sen2cor with sentinel 1C images in the past and had no issue). But I had no luck, I still got the empty folders. Please provide me some guidelines.

have you tried to run sen2cor from the command line? It’s easy to apply and less prone to configuration errors.
Please see this comment here: Sen2Cor-02.05.05-win64 - AttributeError: 'L2A_Tables' Object has no attribute '_L2A_Tile_PVI_File'

Hi ABraun, thank you for reaching out. I tried to run sen2cor from command line. It went smoothly at first then after few minutes it gave me this message: AttributeError: ‘L2A_Tables’ object has no attribute ‘_L2A_Tile_PVI_File’. Please provide me further guidelines.

Hi again, I recently saw your past comments on other’s post. I will start with 60 m then with 20 m and finally with 10 m. I will try this first.
Thanks again!

actually, to get 10 m resolution, no flag has to be set, just
L2A_process S2_FOLDER.SAFE

Hi ABraun

I tried everything accordingly and it worked great. It takes lot less time than in SNAP as well. However, I ran into something. While the command line worked perfectly yesterday, I started to get no results at all today and learned that somehow I have overwritten the L2A_Process.bat file. Because when I went inside the folder it says “0 KB” beside the batch file. Is there a way to download it separately or should I reinstall sen2cor? Please let me know.

That is strange, I cannot imagine how this happened, because it is only executed, not actively written.
But you can download the bundle as a zip file from the sen2cor homepage and only use the bat file inside to replace the empty one.

Hi ABraun
I will try that and let you know. Thank you very much once again.

Hi @binita and @ABraun

i am converting 2016 and 2017 Sentinel2 L1C data, but it seems sen2cor 2.8 ( stand alone ) is not supported, is there any other way to convert the L1C to L2A. using 2.8. ( can I convert older sentinel 2L1c data to new format)

I am able to use sen2cor 2.5 stand alone version, but i am not that keen to use 2.8 for 2018 to 2020 and 2.5 for older version.

any insight will be really helpful

What makes you think that?

Have you seen these instructions? Sen2cor (detailled instructions)

Hi @ABraun

thanks for the response, sen2cor 2.8 release note states that sentinel2 L1C processed using PSD < 14.2 are not supported with sen2cor 2.8. The Sentinel2 L1C data from 2016 and 2017 has PSD version of 12.

I am using standalone version of sen2Cor 2.8 ( Linux version)

oh I see.
Have you checked if L2A products already exist for these images?

Otherwise, you can always run sen2cor from the command line very easily.

There are no L2A products. ( checked in couple of places )
I am running se2cor 2.8 in command line and it fails.

is there a way to use sen2cor 2.8 ( command line ) to process L1C data with PSD version < 14.2

just to be clear - it fails because of the PSD version or do you encounter a different error message?

yes it fails due t o PSD version

so would it be an option to call L2A_process.bat from the 2.55 version for these products from the command line by entering the entire path to the script?

I am able to run using sen2cor 2.5 version, my concern is that i have now that data for 2018, 2019 and 2020 is processes with 2.8 where as older data is processed using 2.5 version.

i am not clear is it ok to have data with processed with different version of sen2cor, as the correction applied in 2.8 is better than 2.5

we have already processed a large amount of data for 2018, 2019 and 2020, 2021, we are going back to 2017, 2016, and 2015 ( where we hit this issue )

now I understand, maybe this can be answered by @Jan

1 Like

Hello @abhiataero

You can use both 2.5 and 2.8 with confidence. The improvements in Sen2Cor 2.8 are mainly:

  • Improvement of the spatial homogeneity of surface reflectance between adjacent tiles.
  • Improvement of the terrain correction algorithm in mountainous areas, in order to avoid artefacts for pixels under semi-transparent clouds.
  • Improvement of the terrain correction algorithm, in order to avoid adjacent tile discontinuities in terrain correction for seemingly flat areas with gentle slopes.
  • Improvement of the Scene Classification of snow/ice, enabling the detection of snow above frozen lakes and icebergs in Arctic and Antarctic Sea.

With regard to processing with Sen2Cor 2,8, when I process an L1C 02.04 baseline product via the command line using L2A_Process.bat (just default, no switches) it tells me that:
“Old product version 14.2 detected - will be updated to 14.5”

It’s runnning as we speak, with no issues so far.

Cheers

Jan

S2 MPC Operations Manager

2 Likes

Hi @Jan

Thanks for the reply, it solved the issue ( did some test on 8 to 10 samples it worked fine)

I had to make two changes add AUX_DATA and QI_DATA empty folder, which I was missing

Thanks @ABraun for all your help :slight_smile:

Regards

Abhijit

1 Like