Estimating random uncertainty of depth-averaged velocities measured by moving-boat acoustic Doppler current profilers

Abstract

This paper presents a mathematical model for estimating the standard uncertainty of depth-averaged velocities measured by moving-boat acoustic Doppler current profilers. A general form of the presented uncertainty model was developed based on the law of propagation of variances and dimensional analysis. It was then calibrated using 30 datasets of stationary ADCP measurements, in which standard uncertainties were available from statistical analysis of the data. Because the model utilizes velocity data collected at a site, it accounts for all random error sources including ADCP system noise and ambient turbulence encountered at the site; it also accounts for the cross-correlation of ADCP depth cells in velocity measurements. The presented uncertainty model can be used in field surveys or data post-processing. It provides a useful tool for assessing the quality of ADCP measured depth-averaged velocities. A moving-boat ADCP measurement on the Mississippi River is presented as an application example. This paper also explores some insights on ADCP velocity profile and time series data.