I would like to use Sentinel-2 Level 1C images to produce EVI maps using the following equation :

EVI=G×(ρnir-ρred)/(ρnir+(C1×ρred-C2×ρblue )+L)

Can I use the Level 1C image as it is, or should I perform atmospheric correction first?

Because EVI is supposed to be an atmospheric resistant index using the coefficent C1 and C2, I'm not too sure if atmospheric correction is necessary or not. My images, from a semi-arid area, are mostly cloud free.

1 Answer 1


You should perform atmospheric correction first, because rhos in NIR/red/blue in EVI are atmospherically-corrected or partially atmosphere corrected (Rayleigh and ozone absorption) surface reflectances. Check wiki.

Besides you can use Level 2A data, which are surface reflectance data.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.