In the beginning of 2021 we decided to share with you an analysis-ready dataset of 2019 Sentinel-2 L2A data on the global scale (here is the original article). This means that this data was:
- cloud-free (accuracy according to s2cloudless cloud detector)
- gap-filled (using temporal interpolation between valid observations)
- available in equidistant temporal intervals (to comply with our OCD tendencies)
This post serves as an announcement that we have further improved our global mosaic algorithm and prepared an update to our analysis-ready dataset for the year 2020. Having such a dataset at the grasp of you hands means that you can simply use it in order to make your big EO ML application come to life, without the need to first process petabytes of raw images just to get the input data.
You can find all the necessary information about the updated collection in the link below, but just to give a quick description, here is a list of changes:
- we added all available bands for L2A
- we implemented a
QM) pixel-level data quality band
- we ran the process over a more detailed global geometry
- How to Make the Perfect Time-Lapse of the Earth by Lubej M.
- Tools & Applications