Both edges of a profile are contaminated by geometry, not by bad electronics. Near the surface the beams intersect the surface before the transducers are deep enough, and near the bed the outer beams touch the bottom before the centre of the profile does.
Three mechanisms are usually at work:
- Side-lobe contamination. A small fraction of transmit energy leaks outside the main beam and reflects from the surface or bed early, mixing into bins that should be clean water.
- Blanking distance. The first bins after transmit are unusable because the receiver is still ringing; nothing can be measured there.
- Beam spreading. As range grows, the sample volume grows and horizontal averaging blurs small-scale structure.
Practical mitigations: keep the transducer well below the surface (typically 0.5 to 1 m of draft, more in rough water), set an honest blanking distance, discard the bins flagged as contaminated by the processing software, and estimate discharge at the edges by extrapolation from valid bins rather than trusting the last few cells. If the site is shallow, a phased-array instrument with a smaller footprint reduces the contaminated fraction.
See the ADCP family overview and the deployment notes on river and estuary work.