Slava,

The reason for the apparent discrepancy between SeaWiFS and MODIS bandpass weighted absorption in the blue portion of the spectrum has to do with the spectral response function for MODIS (both Aqua and Terra).  Below is a plot comparing the MODIS Aqua (blue line) and SeaWiFS (black line) RSRs for 412nm, with the water absorption spectra (red line).  Relative to SeaWiFS, MODIS has a larger contribution from the red/nir, resulting in a significantly higher aw value.

Regards,
Sean

Sean,

thank you for your replay.

Next question arises.
Atmospheric correction procedure removes contribution of rayleigh and aerosol components from total satellite signal, Lt, is spectral response function used in this procedure?

Thank you,
slava

Slava,

The short answer is 'Yes', all processing through to water-leaving radiance (Lw, not nLw) uses the total spectral response, or 'fullband' data. 
All necessary quantities (e.g. Rayleigh optical thickness) are computed for the total band pass.  For a more detailed answer, please see the Gordon and Wang 1994 paper:

"Retrieval of water-leaving radiance and aerosol optical-thickness over the oceans with SeaWiFS - a preliminary algorithm"

APPLIED OPTICS 33: (3) 443-452 JAN 20 1994

Regards,
Sean

I'd suggest:

Gordon HR
    Remote sensing of ocean color: A methodology for dealing with broad spectral bands and significant out-of-band response.
    APPL OPTICS 34: (36) 8363-8374 1995

http://oceancolor.gsfc.nasa.gov/cgi/publications.cgi?a=84

Sean,

thank you for your reply.
I have APPLIED OPTICS 33: (3) 443-452 JAN 20 1994
but have no APPL OPTICS 34: (36) 8363-8374 1995
I'll seek it.

Where can I find spectral response functions for SeaWiFS (1-8 bands)
and MODIS-Aqua (oceancolor bands and used for atmospheric correction bands)?
Better in table format, Maybe in TM?

Take my congratulation with your country holiday.
Thank you,
slava

Slava,

The spectral response functions for all sensors that we support are available in tabluar format via:

http://oceancolor.gsfc.nasa.gov/DOCS/RSR_tables.html

Regards,
Sean

Sean,

Sorry. My previously question is no correct.
I want to get SeaWiFS and MODIS-Aqua bandpass weighted functions the same as in Figure in your first answer to me (Date 2005-06-30 12:29).

Thank you,
slava

Slava,

All the information necessary to regenerate the plot I created, as well as the band-pass averaged quantities in the tables on the RSR page are linked from the RSR page:

The various spectra (aw, bb, F0, etc.) are linked at the top of the page, the spectral response functions are linked at the top of each sensors respective band-pass averaged quantities table.  For example, under the heading 'SeaWiFS' is a link 'Spectral Response Functions' which points to the file 'http://oceancolor.gsfc.nasa.gov/DOCS/RSR/SeaWiFS_RSRs.txt'.

Regards,
Sean

Sean,

thank you for your reply.

I want to back to my first message.
Physical sense of aw' values in http://oceancolor.gsfc.nasa.gov/DOCS/RSR_tables.html is unclear.
Say, I want to restore nLw (Rrs) spectrum by optical model.
What value of aw can I use?
So, the result depends on range of integration.

Thank you,
Slava

Slava,

The aw measurements are in units of inverse meter (1/m).  The values in the tables at  http://oceancolor.gsfc.nasa.gov/DOCS/RSR_tables.html were derived by convolving the data found at http://oceancolor.gsfc.nasa.gov/DOCS/RSR/water_coef.dat with the respective sensor response function.

Sean

hi Sean,

hope u don mind me barging in. i also have some questions regarding spectral response functions (SRFs).

what are the units for MODIS's SRFs?

i thought the SRFs are typically normalised to 1 (by dividing throughout by the largest response value), but the values shown in the tables at http://oceancolor.gsfc.nasa.gov/DOCS/RSR/spectral_response_comp.html does not seem to agree. is it because these have not been normalised, or some other reason?

also, what are the differences/similarities between the SRFs and the relative spectral response, as found at ftp://ftp.mcst.ssai.biz/pub/permanent/MCST/FM1_RSR_LUT_07-10-01/ ?

thx in advance.
bored

Boredin,

The values in the tables under http://oceancolor.gsfc.nasa.gov/DOCS/RSR/spectral_response_comp.html
are radiances (mW/m2/um/sr, which IS listed in the tables).  That web page describes a simple analysis
of the issues related to the use of the spectral response functions for the MODIS instruments.  The
spectral response functions were  convolved with a clear water radiance model - which has radiance units.

The RSRs (or SRFs if you prefer) for MODIS are indeed normalized to the maximum measured value.
See http://oceancolor.gsfc.nasa.gov/DOCS/RSR/Aqua_RSRs.txt and http://oceancolor.gsfc.nasa.gov/DOCS/RSR/Terra_RSRs.txt

These were derived from the values provided by MCST.

(note, the SeaWiFS RSRs are not normalized, but the normalization isn't necessary, since the normalization will occur
with the application of the RSRs )

Regards,
Sean