Different S1 processing result using GPT and Snappy


I am using a processing graph for S1 GRD data that works well using both GUI and SNAP Command-Line (Snap Version 6.0.4). I use two images as inputs because I apply Slice-Assembly tool. Since I need to process many pair of images, I would like to automatize the process with Snappy.

I have tested the xml graph in Python with a pair of images and it works, but the result is slightly different from SNAP Command-Line result: the pixels do not have the same value and there is border noise.

These are the results: SNAP Command-line (left) - no border noise / Snappy (right) - border noise.

I have also tested the graph using SNAP Command-Line in other computer (Snap Version 6.0.0). In this case, the result is the same that using Snappy: border noise. I think these differences are related to Snap version. Is it possible to update Snappy library so that I can obtain the same results as in Snap Version 6.0.4?


S1_processing.py (904 Bytes)
S1_processing_gpt.xml (10.0 KB)

The difference in the shore probably points to something being different with the DEM. Try running it again with the parameter to output the dem in the terrain correction set to true. Probably something isn’t loaded that should be. I’ll try debugging your script to see what could be missing.

Thank you so much. I have changed the output DEM parameter and it also appears the border noise when I run the graph using SNAP Command-Line, although the values of each pixel are still slightly different to Python’s result.