Spectral signature of Vegetation is weird

Hello experts!!!

I have got a weird kind of product after using SNAP. First I used Sen2Cor for correcting the image and then the corrected images were then fed into a graph which produces 20 spectral vegetation indices plus the nine bands. Later when extracting the values from these images the reflectance in green is always lower than in blue and red which is wrong. I have attached my graph and the final signature for vegetation I got. Does anyone have an idea of

why this might have happened?

Thank you.

I see no problem here. The reflectance of red and blue can be higher than green ad still the vegetation is perceived green by the human eye.

A test on different types of vegetation:

How is this related to the vegetation indices?