'Resampling' or 'Super-resolution' for calculating Normalized Difference Moisture (Water) Index?

Optical Toolbox
#1

Dear friends,
I am trying to calculate NDMI from Sentinel-2 images. What method would be preferable among ‘resample’ and ‘super-resolution’ as SWIR band has 20m resolution whereas NIR band is in 10m resolution?
My main observations from using both the methods are,

  • A pixel which has high NDMI values (calculated after super-resolving) shows even higher values when calculated after Resampling

  • Similarly, a pixel which has lower NDMI values (calculated after super-resolving) shows even lower values when calculated after Resampling

I performed resampling in all 3 methods available in SNAP and all of them gave similar observations as mentioned above.
Does it mean that performing ‘resampling’ might lead to overestimation?

Thank you.

#2

(B8 - B11)/(B8 + B11) which would be at best at 20m resolution
according to

#3

@gbrelstaff Thank you very much for your response, I will go through it.