Hello everyone, I would like to ask you some simple questions: (1) if I use high-quality DEM for Doppler terrain correction, does the decibeled sigma reduce the influence of surface roughness to the greatest extent, like this：
Two things are important here:
as stated in the SNAP histogram, these are just “rough statistics” (calculated from a subset of all pixels), so there is no guarantee that the given numbers are the true min/max values. If you want to go sure, you have to use the statistics tool which calculates statistics based on all pixels.
ArcMap also does not show you the entire range of values. It cuts the outliers so that your raster is presented in “nice” contrasts. Please have a look at this tutorial Export of products from SNAP, especially the chapter “Notes on contrast”, which explains this based on examples.
Thank you very much for answering my question and I will continue to study~
to answer your second question: Sigma0 corrects for the overall incidence angle so that you can compare images of different angles. This does not apply for the shape and orientation of topography (you need to apply Terrain Flattening) or objects/roughness at the surface. These will always be part of your backscatter signal (unless they are estimated and removed by polarimetric approaches, but this is not feasible with Sentinel-1).
First of all, thank you for your reply. Can I understand that the obtained sigma0 is the backscattering coefficient (without eliminating the influence of surface roughness)？ Finally, thank you again for your answer~
yes, this is correct. It is described at more detail here:
I’m very sorry that I’m busy with my teacher and elder martial brother recently. I didn’t log in in time. I want to say that I feel very honored to communicate with you in the forum.