the SUM states regarding the data downloaded outside the growing season:
You might notice that the system starts downloading products three months before the defined season start date. This behaviour is intended and the extra data is used to initialize the L2A processor.
Is there any estimation on how important this data is, especially if there’s already a good amount of data within the growing season?
In this case, is it possible and would it make sense to either disable this download, or - if already downloaded - to at least temporarily exclude these files from L2 processing?
Three months were recommended by the developers of that processors. I don’t have any knowledge of how important this initialization period is, and how it will reflect in the output of the other processors. I suspect it’s mostly used for cloud detection, although I’m not sure about that either.
You can try to reduce its length by changing the downloader.start.offset configuration value.
As for disabling the processing of the downloaded products, you’d have to chage their status in the downloader_history table.
I assume setting it to status 4 would be appropriate.
As I described in this post, we still don’t process as fast as I would have hoped to. We still have more than 1200 producs outside the growing season which are not processed to L2, so I was hoping I could reduce the time on this side a bit. But if the output quality can’t be ensured, I’d rather not.