Treatment of echo-sounder data for real-time bridge scour monitoring: an application on the Po river, Italy

The manuscript describes the use of an echo-sounder as a monitoring device for the bed elevation downstream of a bridge pier on the Po river in Italy. The echo-sounder furnishes every hour a sample of data measured with a frequency of 1 Hz. Data are processed in order to obtain hourly values; the latter may be obtained as a simple arithmetic mean when the data sample is relatively regular, but the post-processing requires instead some more care when the data samples are more puzzling (typically, when a strong bimodality of the data emerges, likely as a result of operating conditions). Based on an accurate analysis of situations chosen on purpose, an algorithm is setup to compute hourly values reducing the bias determined by the presence of the so-called ‘Unsure’ data and by bimodal data distributions. The application of the post-processing algorithm to long-term series provides a reliable temporal evolution of the bed elevation, that may be used to assess in real-time the bridge hazard, as well as to interpret the morphologic dynamics of the site.