Negative values of water-leaving reflectances with Polymer v4.8

Post Reply
MohMograne
Posts: 2
Joined: Fri Dec 14, 2018 11:49 am
company / institution: LOG
Location: Wimereux

Negative values of water-leaving reflectances with Polymer v4.8

Post by MohMograne »

Hello,

I used Polymer v4.8 to perform the atmospheric correction to get the water-leaving reflectance for OLCI-A sensor. In the resulting spectra I found some negative values of water-leaving reflectance at 709, 885 and 1020 nm. Why such negative values ?

How Polymer accounts for the fluorescence correction ?

Thanks.
Mohamed
User avatar
fsteinmetz
Site Admin
Posts: 306
Joined: Fri Sep 07, 2018 1:34 pm
company / institution: Hygeos
Location: Lille, France
Contact:

Re: Negative values of water-leaving reflectances with Polymer v4.8

Post by fsteinmetz »

Hi Mohamed,

Indeed it is possible that some reflectances are negative.
The output of Polymer is not directly the output of the water reflectance model. The final water reflectance is recalculated from the observed spectrum by subtracting the atmospheric path reflectance (eq 10 of Steinmetz et al 2011). This subtraction may lead to slightly negative values, especially in the red/NIR where Rw is close to zero.

Concerning the fluorescence : the water reflectance model does not account for fluoresence. But this does not matter, because the fluorescence bands are not used in the whole atmospheric correction. Thus, the atmospheric path reflectance is calculated at the bands 681 and 709, without being affected by the potential fluorescence signal at these bands. Since we subtract the atmospheric path reflectance from the observation, the potential fluorescence signal in the final water reflectance should be preserved.

However, water vapour correction is currently (as of v4.9) not applied to these bands, so this also may affect the retrieval of the fluorescence signal.
Post Reply