Questions about QAAV6 source code

[a.topo_s]

Why was the ‘check aph443 range’ code (lines 412-440) added in the source code of the QAAv6 algorithm in l2gen (https://oceancolor.gsfc.nasa.gov/docs/ocssw/qaa_8c_source.html)?

[a.topo_s]

Hello everyone,

I noticed that in the pml_iop_calculate.c file （https://oceancolor.gsfc.nasa.gov/docs/ocssw/pml__iop__calculate_8c.html）, there is also a "check aph443 range" code segment （line 87-105）, and the comments mention that "this sanity check is an empirical fix from the Lee model." I have searched through the algorithm's technical documentation and published literature but haven't found any explanation or description of this empirical fix.

Could someone please point me to where I can find the basis or a reference for its use?

Thank you very much!

[OB.DAACx - SeanBailey]

This one required a bit of sleuthing...and contacting the developer who originally provided the code used in l2gen. It has been determine that this change went into the code base circa 2006, when updates were made going from version 2 to version 3 of QAA. The current code is version 6. As to exactly why the code was added, we can only surmise that it was a check to ensure the inversion was stable. I know this may be an unsatisfactorily incomplete answer, but short of calling up Zhongping Lee (who might also not recall, as it has been almost 2 decades), this is the best we can provide.

Regards,
Sean

[a.topo_s]

Hi Sean,

Thank you very much for your reply! I emailed Professor Zhongping Lee, and he mentioned that the "check 443 range" procedure is entirely based on their data and experience. It's intended to ensure that the aph440 product remains within a reasonable range even when there are issues with the quality of the Rrs data.

Thanks again for your support!

[noramila]

Glad Professor Lee confirmed that the "check aph443 range" stabilizes QAAv6 results, especially when Rrs data quality is uncertain.