I’m working with SentinelHub WMS to get Sent. 2 NDVI, and I compare it with previous indices I got with bands from AWS, pre processing with Semi Autom. Plugin, and is not the same value for a given pixel, making DOS1 atmosferic correction for both indeces.
Does anyone know why is this? Should I work always with same “resource” of information to get comparable indeces?
I am not familiar with Semi. Autom. Plugin so difficult to comment on this. Sentinel Hub WMS is as well based on AWS data so there should not be a difference.
Is it because of the bands used (B8A versus B8)?
NDVI = (NIR - Red) / (NIR + Red)
Sentinel 2 - Band 8a (NIR, 20m), Band 8 (NIR, 10m), Band 4 (R, 10m)
Thanks! I’m creating NDVI both from Band 8 and 4 so it should not be that
the problem. Regards
The link says its better to use B8A for NIR…
NDVI = (B8A - B04) / (B8A + B04) vs NDVI = (B8 - B04) / (B8 + B04)