ACOLITE forum

I'm trying to retrieve Chl-a concentration of a lake in Berlin (Muggelsee) in different dates with a "spectrum approach".
I have in situ data of Chl-a concentration but not of reflectance so I can't use a ELM o similar approach.
So I need to do some absolute atmospheric correction.

I'm using Sentinel 2A data only. Here is the RGB images of the days that I choose:

https://drive.google.com/drive/folders/ ... sp=sharing

I tried to apply ACOLITE to the granule containing the lake with standard parameters(SWIR, per pixel epsilon), with different raylight correction and also with NIR/SWIR here is what I got:



Some comments:

-I took the spectra at a single pixel (60m output), containing the measurement station:
(WGS 84 Dec. Deg.): 52.446141, 13.650048
and in 3 other locations in the lake.

-I thought to an "adjacency problem" although the station is very small.

-Some days have thin cloud so they are obviusly wrong (check RGB pictures)

-After ACOLITE correction I would have a peak at around 500 or 700 wavelength.

-For example 24/12/15 is one of the best days (clear, without cloud) but has an almost totally negative spectra

-For me there is not a sun glint problem.


Thank you very much for your time


Thank you very much for your time

I will have a look at your specific area and scenes! I've also noted that for some inland waters ACOLITE approach does give negatives, probably caused by adjacency effects in the SWIR bands. Perhaps using a fixed aerosol reflectance will solve some of these issues. (You could also use a lower percentile than the default 50th.)

Note that the end of December the sun will be very low, and there will be rather large errors (overestimation) on the Rayleigh correction (computed using plane parallel simplification). You are probably processing other scenes through the year though, where the sun will be high enough.

Unfortunately I cannot open your links with examples.

quinten wrote: ↑Wed Nov 01, 2017 11:17 am I will have a look at your specific area and scenes! I've also noted that for some inland waters ACOLITE approach does give negatives, probably caused by adjacency effects in the SWIR bands. Perhaps using a fixed aerosol reflectance will solve some of these issues. (You could also use a lower percentile than the default 50th.)

Thank you for your suggestion, but it's not working. And yes maybe this is caused by the adjacency effect but I get the same strong path for the center of the lake.

quinten wrote: ↑Wed Nov 01, 2017 11:17 am Note that the end of December the sun will be very low, and there will be rather large errors (overestimation) on the Rayleigh correction (computed using plane parallel simplification). You are probably processing other scenes through the year though, where the sun will be high enough.

Yes, I'm working also on some other scenes. Anyway, thank you for explanation

quinten wrote: ↑Wed Nov 01, 2017 11:17 am Unfortunately I cannot open your links with examples.

Why? Did you get some error?

Thank you
Michele