l2gen for MSIS2 with NIR/SWIR method

Please enter here to ask a question about any NASA Science related topics!
Post Reply
Posts: 7
Joined: Tue May 28, 2019 10:31 am America/New_York

l2gen for MSIS2 with NIR/SWIR method

by lzoffoli » Tue Jul 16, 2019 4:11 am America/New_York

I'm running l2gen in Seadas 7.5.3 to process Sentinel-2 images. I want to use the NIR/SWIR method as I'm working in shallow waters and for the default AC option I retrieved weird Rrs values (either negatives or higher than 1) even for the band at 665nm.

This is how my command line looks like:

l2gen ifile=/home/laura/Desktop/Glenan/Glenan2019/S2B_MSIL1C_20190709T112119_N0208_R037_T30TVT_20190709T131959.SAFE/manifest.safe ofile=/home/laura/Desktop/Glenan/Glenan2019/Seadas/20190709bis.nc l2prod='Rrs rhos' brdf_opt=0 aer_opt=-9  aer_swir_short=1613 aer_swir_long=2200  aer_wave_short=740 aer_wave_long=865 cirrus_opt=false cloud_thresh=0.018 cloud_wave=2130.0 maskhilt=0

I'm using center wavelengths detailed in Pahlevan et al 2017 (Sentinel-2 MultiSpectral Instrument (MSI) data processing for aquatic science applications: Demonstrations and validations). However, I get this error:

Processing scan #     0 (1 of 5490) after     32 seconds
Aerosol selection bands 739 and 864
Aerosol selection bands 1613 and 2200 not available for this sensor

There is a problem in my command line? or NIR/SWIR correction is not available for MSIS2?
Also, in an old post from last year it says that it's not possible to obtain 10m resolution pixel from seadas (https://oceancolor.gsfc.nasa.gov/forum/oceancolor/topic_show.pl?pid=40047;hl=sentinel%2010m) but all the bands are resampled to 20m. I wonder if something changed in new releases of Seadas.

Many thanks!


OB.DAAC - SeanBailey
User Services
User Services
Posts: 1171
Joined: Wed Sep 18, 2019 6:15 pm America/New_York

l2gen for MSIS2 with NIR/SWIR method

by OB.DAAC - SeanBailey » Tue Jul 16, 2019 8:09 am America/New_York

The MSI instruments on S2A and S2B have *slightly* different nominal band centers (see below).  If processing S2B, you'll need to use 1611 and 2184 for the swir channels.
Also keep in mind that Nima provided vicarious gain coefficients for S2A,  but for S2B we're still using gains of unity (so effectively no vicarious calibration). 
This may also be impacting your retrievals.

            S2A    S2B
Lambda(1) = 443    442
Lambda(2) = 492    492
Lambda(3) = 560    559
Lambda(4) = 665    665
Lambda(5) = 704    704
Lambda(6) = 740    739
Lambda(7) = 783    780
Lambda(8) = 835    835
Lambda(9) = 865    864
Lambda(10) = 945   943
Lambda(11) = 1613  1611
Lambda(12) = 2200  2184

The l2gen program still only outputs 20m resolution data from MSI.  This will not change.


Posts: 7
Joined: Tue May 28, 2019 10:31 am America/New_York

l2gen for MSIS2 with NIR/SWIR method

by lzoffoli » Tue Jul 16, 2019 9:07 am America/New_York

Hi Sean,
thanks a lot... I'll try with these wavelengths

about gains, I found some values for both Sentinel-2 satellite in the table 1 of this new Nima's paper:

Pahlevan, N., Chittimalli, S.K., Balasubramanian S.V., Vellucci, V. (2019) Sentinel-2/Landsat-8 product consistency and implications for monitoring aquatic systems. Remote Sensing of Environment 220, 19–29. https://doi.org/10.1016/j.rse.2018.10.027


Posts: 1
Joined: Tue Dec 24, 2019 11:56 am America/New_York

l2gen for MSIS2 with NIR/SWIR method

by lanzhiqishi1990 » Wed Dec 25, 2019 11:04 am America/New_York

Hi, friends,

are you change the band and are they work for aer_opt=-9? my question is after l2gen, i could not find the Rrs at any band during  aer_opt=-1? are you successful on aer_opt=-9?



Post Reply