I created a classified image of my study area using random tree classifier method as shown:- enter image description here Fig 1: Classified Image of Landsat-8

enter image description here Fig: Classified Image of Sentinel-2

I now want to classify image based ON NDVI value based on these images as reference. I classify Landsat image setting NDVI value as shown. enter image description here Fig: NDVI map of Landsat-8

I also classified Sentinel-2 image using NDVI value using above 2 reference images as shown. enter image description here Fig: NDVI map of Sentinel-2

But, on classifying these images land cover classes the NDVI values for landcover classes for these figures (3) & (4) are different for Landsat-8 (fig 3) & Sentinel- 2 (fig 4). Is this because they are different satellite images i.e. Landsat-8 & Sentinel-2 and it's like that??

1 Answer 1


Did you apply the gain and biases to your images ? Are they radiometrically coherent (TOA, BOA ?)

To be sure to have aligned S2 and Landsat-8 images, check the HLS (Harmonized Landsat-Sentinel) images created by NASA. The drawback of those is that S2 images are downsampled to 30m.

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.