How to process Landsat8 Collection2 L1 data with SeaDAS

[roxani2018]

Dear all,
Hi!
I encounter a problem when using l2gen to produce Rrs data from Landsat8 Collection2 L1 data. Follows are what it shows on my GUI interface:
============================================================================================
# PRIMARY INPUT OUTPUT FIELDS
ifile=/Users/mj/Desktop/LC08_L1TP_122048_20140422_20200911_02_T1/LC08_L1TP_122048_20140422_20200911_02_T1_MTL.txt
# # WARNING!!! file /Users/mj/Desktop/LC08_L1TP_122048_20140422_20200911_02_T1/LC08_L1TP_122048_20140422_20200911_02_T1_MTL.txt is not a valid input mission: Mission=null
============================================================================================
I have already downloaded OLI processor on SeaDAS 8.0.0, so what should I do next?
I really appreciate for your help.
Best whishes.

[OB SeaDAS - xuanyang02]

Please do SeaDAS-OCSSW -> SeaDAS/System Info and post the output here.

[roxani2018]

Hi,
Thanks for your reply. Follows is the output shown in System Info.
I am not sure what it stands for? Since it works well when processing Landsat8 Collection1 L1 data.
Hope to hear from you soon and thanks a lot.
Roxani

[OB.DAACx - SeanBailey]

SeaDAS does not yet support Landsat Collection 2 formatted data, neither for visualization nor for data processing. Visualization and data processing do work for Collection 1 formatted data. Unfortunately, we do not have an estimate on when support for Collection 2 will be added.

Regards,
Sean

[roxani2018]

Thanks Sean.
Now I successfully produced L2 data from Collection1 L1 data. However, I found there exists a lot of negative Rrs values, especially Rrs(655).
During AC process, variables are all set by default, and ancillary data has been caught first. Is it the wrong variable setting or something else?
Sorry to take up your time and waiting for reply.
Thanks,
Roxani

[OB.DAACx - SeanBailey]

Roxani,

This is a different question, so should probably be a new thread...but, since I'm here...

Landsat is, well, designed for land. The signal-to-noise is a wee bit below what is really required to do ocean color remote sensing - especially for the red bands (a bit better in the shorter wavelengths, though). Noisy data combined with low signals can easily result in negative Rrs retrievals.

For more information see:

Nima Pahlevan, Zhongping Lee, Jianwei Wei, Crystal B. Schaaf, John R. Schott, Alexander Berk,
On-orbit radiometric characterization of OLI (Landsat-8) for applications in aquatic remote sensing,
Remote Sensing of Environment, Volume 154, 2014 (https://doi.org/10.1016/j.rse.2014.08.001)

Nima Pahlevan, John R. Schott, Bryan A. Franz, Giuseppe Zibordi, Brian Markham, Sean Bailey, Crystal B. Schaaf, Michael Ondrusek, Steven Greb, Christopher M. Strait, Landsat 8 remote sensing reflectance (Rrs) products: Evaluations, intercomparisons, and enhancements, Remote Sensing of Environment, Volume 190, 2017 (https://doi.org/10.1016/j.rse.2016.12.030)

Sean

[roxani2018]

Hi Sean，
Sorry for the long delay of this reply. Thanks a lot for your detailed response, it really helps.
Best wishes,
Roxani