How can I process the atmospheric correction in sentinel-2 images using SNAP for under water vegetation (marine and coastal ecosystem)? (from TOA to the water colomne).
It is a study of geographical distribution of seagrass in my region (SW of Madagascar). As it is about underwater vegetation, before classifying the image S2, I have to make (i) an atmospheric correction but apart from the atmosphere, before reaching the meadows (seagrass), the light must (ii) cross the sea-water and cause a loss of information. What I need is a is a protocol or guideline for S2 image pre-processing for this kind of study. Note that I am a beginner in the domain of remote sensin.
You should use an atmospheric correction developed for ocean colour applications. Using SNAP, you have two options:
1. The built-in C2RCC plugin (SNAP/Optical/Temathic water processing/C2RCC processors/*).
2. iCOR. Needs to be downloaded a part from SNAP and then installed as external plugin. (https://blog.vito.be/remotesensing/icor_available)
If you want to work in coastal shallow areas keep in mind the effects of bottom reflectance, turbidity or depth among others. Atmospheric correction processors may not be accounting for them.