SNAP vs Python different raster statistics

I am trying to calculate some basic statistics (min, max, median, percentile) from the S1 rasters that I am working on. The calculated statistics from the same raster using python numpy is somewhat different from the exact calculation of raster statistics in the Color Manipulation window of SNAP. Which one should be taken as the accurate one ? What is the reason of this difference ?

Could you please give examples how you derived both statistics?
With screenshots at best.

I am sorry!
I made a mistake by confusing two different images with two different polarizations.
The statistics are calculated the same.

That’s good to hear. Would have been strange.

One possible explanation could have been a valid pixel expression applied go the data in the band properties in SNAP, while python looks at the raw raster values only.

1 Like