The number of NoData pixel in the output L2 is too high

[RubenValDin]

Hello Olivier,

I am running MAJA-3.3.0-TM for 6 scenes:

S2B_MSIL1C_20190713T173909_N0208_R098_T14UNU_20190713T211057 S2A_MSIL1C_20181207T173711_N0207_R055_T14UNU_20181207T192326 S2B_MSIL1C_20190223T174329_N0207_R098_T14UNU_20190223T211941 S2A_MSIL1C_20181117T173611_N0207_R055_T14UNU_20181117T210544 S2A_MSIL1C_20190509T173911_N0207_R098_T14UNU_20190509T224452 S2A_MSIL1C_20190320T174031_N0207_R098_T14UNU_20190320T230030

I got the following error: The number of NoData pixel in the output L2 composite product is too high.

I have checked the log file (attached. It was ran with --debug mode) and found that the L2outNoDataRate (pourcentage) for one of the scenes is higher than the MaxNoDataPercentage allow, so I guess that's why it failed. However, from the log file I cannot figure out which one is the image.

Also, I am wondering why it fails here instead of creating a folder such as for the cloud cover when it doesn't fit the requirements and if there is a way of avoiding it.

Thank you very much

Ruben

S2A_MSIL1C_20181117T173611_N0207_R055_T14UNU_20181117T210544.SAFE.log

[olivierhagolle]

Hi Ruben, The list of products you are providing to MAJA is strange: 6 dates in 6 months. As MAJA uses multi-temporal information you have to provide all the dates in the time series, and not only the ones you selected.

What happens is that with what you provided, MAJA is not able to initialize the data set properly, and finishes the backward processing with no information on the surface reflectance to put in the composite image.

Could you please try downloading the whole data set with all the dates and start again the processing ? Olivier