Hi, Alex.
I have not seen this in a discharge measurement, but I have seen a similar feature in a test with a RiverPro anchored in a lake. I really hope I would have noticed it in my own discharge measurements, and I have had a new look at some of them now without seeing the feature. The anchored riverpro shows a HIGHER velocity magnitude (and east and north velocity) in the first bin below the surface bins in WinRiver when I average a lot of ensembles. See attached screenshot. (The reason it was anchored is that we used the data to check if a still lake was actually still when testing fro flow disturbance). If not averaging, we only see low velocity magnitudes in the top bins and higher velocity magnitudes in the rest of the bins, since they are more noisy.