Calibration of an acoustic pipe sensor through bedload traps in a glacierized basin

Quantifying sediment transport in small mountain basins is of great relevance to assess the morphological and ecological dynamics of the entire channel network and to predict flood hazards. In high-elevation, glacierized basins, seasonal variability in sediment transport is dramatic, but despite the relevance of such basins in many regions worldwide, very few investigations have tried to quantify it. Since direct methods to assess bedload transport are time consuming and practically challenging at high flows, indirect surrogate methods, allowing continuous measurements over time, are highly desirable. Yet, these methods require calibration to provide reliable estimations. The present research is focused on the calibration of an acoustic pipe sensor in the recently established (Spring 2011) monitoring station in the Saldur basin, a high-elevation glacierized watershed in the Eastern Italian Alps. The acoustic pipe signal (which is amplified through 6 channels having different gains) was calibrated against samples collected over 26 sampling periods using “Bunte” bedload traps along a cross-section 12m upstream of the pipe. Samples were collected from June to August 2011 during daily discharge fluctuations (ranging from 1.40 to 3.63m3s−1) due to snow- and glacier-melt, featuring very different bedload rates (up to 0.14kgs−1m−1). In order to calibrate the pipe sensor signal, the average number of impulses was plotted against the corresponding unit bedload rates for the associated bedload sampling periods. As expected, the signal from the two most sensitive channels of the acoustic sensor resulted dampened even at low discharges, and thus could not be used for calibration and bedload assessment. Instead, power laws (R2 from 0.76 to 0.92) relating the number of impulses per minute to unit bedload rate were obtained using channels having intermediate and low sensitivities, with higher correlations associated with the less sensitive channels.