Visualizations of Google Colab bands at L2A seem to be incorrect

[stienheremans]

Dear,

I have found an issue in the Google Colab visualizations of the time series. it seems to me that the band order from 'allbands[L2A]' is incorrect, as it includes 'id' and 'code_cultu' before the actual relfectance bands. But in the dataset (using Belle-ile) this seems not to be the case, it seems that there first you have the doa and then directly the reflectance values of the bands. As a consequence, the bands_idxs in the 'plot_sample' function for L2A are incorrect. It states that B2 is in position 3, but in the dataset position in 3 is actually B4. In the plots, I first noticed the strange behaviour for B11 and B12, which seem to be always (near) 0. I think this is because the values for 'CLD' and 'EDG' are plotted as B11 and B12, respectively. So, I think B2 in the plot is actually showing B4, B3 is actually showing B5, B4 is actually showing B6 and so on... Eventually B11 is actually showing CLD and B12 is actually showing EDG.
I don't know if it affects the modelling, I did not yet go through that part in detail yet.

Kind regards,
Stien Heremans