Wrong SeaWiFS PAR on MLAC swath borders

[dem1]

Hi,

We found that the default SeaDAS l2bin flags filtering for SeaWiFS GAC&MLAC PAR:
flaguse=LAND,NAVFAIL,FILTER,HIGLINT (in the file l2bin_defaults_PAR.par)
seems not enough for MLAC products because there is a vertical band with very low PAR values on MLAC swath border.
It seems the HISATZEN can remove these wrong PAR pixels.
GAC are not impacted maybe because the swath is thinner. I don't know if other sensors have a similar PAR issue.

Example: SEASTAR_SEAWIFS_MLAC.19990620T160022.L2.OC.nc, see MLAC PAR vs GAC on attached screenshot, with the suspect vertical band in purple.

SEAWIFS_PAR_MLAC_VS_GAC.png (305.39 KiB) Not viewed yet

Have a nice day,
Julien

[OB.DAACx - SeanBailey]

Julian,

Thanks for alerting us to the issue. Adding HISATZEN is a good stop-gap measure for the MLAC data, however it will exclude more than is necessary. Once we have identified the cause for the erroneous data in the SeaWiFS MLAC PAR output (it appears to only affect SeaWiFS MLAC), we determine how best to resolve it. However, we will likely wait for a future mission reprocessing to implement any change to the product.

Regards,
Sean

[dem1]

Thanks Sean, ok so for the moment we will remove HISATZEN SeaWiFS MLAC PAR pixels. Note that the CLDICE flag is also wrong on the vertical band (we use the CLDICE flag to construct a cloud fraction product and we can see the issue on it), I don't know if other L2 flag are also impacted.

Best regards,
Julien