ToDo.txt

James:
- New adapter for averaging pixels across multiple timesteps (SNAP operator)
- Set up new server
- Integrate MDN into SenCast (author may be about to release as package)

Jasmin:
- C2RCC auxdata doesn't seem to work for S2
- Review effects of missing auxdata on C2RCC

Dani:
CURRENT - Select bands for datalakes adapter & PP algorithm (for James until next week)
CURRENT - implement post processing filtering of non-physical/impossible values

Other:
- input_data folders in Creodias are called 'L2' rather than 'L1'
- Implement a logger that writes the log to a file (one file per product group)
- MPH not working for partial products which come from Idepix (date: 2020-04-20, wkt: sui)
- MPH immersed_cyanobacteria is transparent instead of dark blue at zero values (QL)
- Add Loisel IOP algorithm
- RGB output?


Newly discovered by Dani:
- if choosing only 'C2RCC' or 'L-FLUO' as processors, IDEPIX is no longer applied automatically

Completed:
- Distinguish between NRT and NTC data (OLCI)
- PP method integration (Soomets et al.) 
- Make the notification of failing ERA5 requests more obvious
- fix strange S2 polymer forms when mosaicked (check with a few Lake Constance images)
- Enable L1 data access via CreoDIAS API (older Sentinel-2 data)
- Merge output into single nc file
- Remove corrupted files if GPT fails or download is interrupted
- Prevent mosaicing of S3A and S3B files
- Implement option to use new paramaeter file even if one exists in output folder - Ensure all data reprocessed
- Flags for mosaiced products are missing -> add some logic to create them
- Masks for mosaiced products are missing -> add some logic to create them
- Wavelength metadata for polymer and mosaics
- Implement rerun logic for datalakes
- Moving average smoothing