Hi there – I’m using the s3tbx to successfully transform OLCI radiances to TOA reflectances. However I notices only that there are some evident stripes in the aforementioned bands, you can see the stripes in the following screenshot which shows the radiances:
I checked several images and all of them have the same issue. Am I missing something? is any additional correction for these absorption bands needed?
Unfortunately, these stripes can’t be removed at the moment. But a processor which can do it is planned for next year.
The vertical stripes are caused by the so-called “smile effect”.
You can read about it here and here.
The L2 files you can get from ESA and Eumetsat are already corrected for this effect.
Thanks for the quick reply.
I do need all OLCI TOA reflectances bands in order to apply some spectral transformations. I saw that in the documentation you pointed out (SENTINEL-3 OPTICAL PRODUCTS AND ALGORITHM DEFINITION ), in table 2 is mentioned that the “reflectance correction switch” is off in several bands.
- Does it mean that for all those bands the “smile effect” hasn’t been corrected in the current radiances to TOA reflectances processor?
- Or actually there is no “smile effect” correction at all applied in the radiances to TOA reflectances processor?
The documentation refers only to the L2 standard products you can download from ESA and Eumetsat.
it is also mentioned in the section above table 2
Correction is not made on all bands but restricted to a predefined subset for which those variations can induce significant distortions when used in algorithms based on a fixed wavelength scheme.
In SNAP this effect is not corrected at all. Not yet.
Thanks again. As I mentioned before, I do need TOA reflectances, “smile effect” corrected TOA reflectances.
Question is, when is the processor are you talking about going to be released? or could it be possible to get use the L2 reflectance correction that actually is applied to the 21 OLCI bands and write all the outputs?, not only the ones needed for the L2 processing.
Thanks a lot for your help.
The processor might come until summer. Probably too late for you.
All bands are written to the L2 products, but sometimes the correction is applied and sometimes not.
Have a look at the ESA land products or the Eumetsat water products. Depending on your application.
But here the reflectances are not TOA anymore.
If you want to have reflectances for water and land it is just not possible at the moment. Especially if they should be smile corrected.
Maybe you can merge the water and the land products. But probably also not a good solution.
If you accept the smile effect you can simply use the Radiance-to-Reflectance processor.
Definitely too late for me. I need a consistent time series of smile corrected TOA reflectances over land. Since this is needed for the Multiply project and we (Assimila) are part of the consortium (Brockmann is a partner too), perhaps we can take this offline and figure out how to tackle the issue.