I'm processing OLCI/Sentinel-3A with SeaDAS 7.5 , and I don't understand why the surface reflectance (rhos) at 709 nm is always depleted as compared to the two nearby wavelengths (681 and 754). I speak about ocean spectrum, far from sunglint or clouds, with no vicarious gain applied. See enclosed png file. I expected a 709 nm rhos value approximately given by a linear interpolation between 681 and 754 rhos values. Is there some correction missing in Seadas for this particular wavelength ?
Keep in mind these are not likely to be the set of gains we apply when we finally get around to distributing L2 data...but they're probably close.