turbid and coccolithophore flags in Level 2 and Rrs Level 3

[avalente]

Hello,

I read in forum that the turbid flag is a threshold at the red band and that the coccolithophore flag is related to the algorithm discussed by Moore et al. (2012); doi:10.1016/j.rse.2011.10.001. Could you please confirm and, if possible, direct me to a documentation on this?

Also, in relation to the use of coccolithophore flag by default in Level 3 data.
Does this happens also for Rrs and PIC Level 3 images?
As I understand, the Level 2 Rrs data would be masked with the coccolithophore flag and then binned for Level 3, so the answer would be yes for both (Rrs and PIC)?
But if this occurs for PIC, then the obvious large coccolithophore blooms are not present in PIC Level 3, correct?
 
Hope you can help. Thank you very much.

Best regards,
Andre

[OB.DAAC - SeanBailey]

Andre,
The coccolithophore flag currently in use is not related to the Moore (2012) paper (although probably should be...). It's a much older - Brown and Yoder (1994, doi:10.1029/93JC02156)
The flag is masked for Rrs at L3, but NOT for PIC (as you point out, that would mask the most interesting bits for that product :wink:)

Sean

[avalente]

Dear Sean,

Thank you. Could you please confirm that the cocco flag is on for Level 3 CHL?
Also, this brings me other questions.
If the flag is not on for PIC, then pixels with large blooms could also have some errors (dark-pixel failure?, guess that is a purpose for using that flag)?
Also, is there any way to have access to Level 3 Rrs without the cocco flag?

Thanks again! Hope you can help.

Best regards,
Andre

[OB.DAAC - SeanBailey]

Yes, it is on for the L3 CHL products.  The PIC algorithm is a blend of two algorithms.  One of those does not employ the GW94 atmospheric correction approach and is immune to the problems of non-negligible NIR reflectance that would certainly be present with high concentrations for PIC.   Since this is not true for the Rrs and it's derived products, we mask pixels flagged as containing coccolithophores for those products.  To get L3 Rrs that doesn't flag these pixels you would need to bin the input L2 files excluding that flag in the flaguse parameter...however doing so might void your warranty :razz:...as, well, there's a reason we flag it for Rrs and it's derivatives...

Sean

[avalente]

I see.. Ok thanks a lot for the information!