On the Accuracy of Acoustic Doppler Current Profilers for In-situ Measurements. A Proposed Approach and Estimations for Measurements in Tidal Channels

S. Jiménez-González¹, R. Mayerle1 and J. J. Egozcue²

¹Coastal Research Laboratory and Research and Technology Centre
Westcoast, University of Kiel.
Otto–Hahn–Platz 3, 24118 Kiel, Germany.
Tel. 49–431–8803641,
Fax. 49–431–8807303,
E-mail: rmayerle@corelab.uni-kiel.de

²Departament de Matemàtica Aplicada III,
Universitat Politècnica de Catalunya. Mod. C2,
Campus Nord. Jordi Girona Salgado 1-3,
08034 Barcelona, Spain.
E-mail: juan.jose.egozcue@upc.es

Although information on the accuracy of Acoustic Doppler Current Profilers (ADCPs) is available for laboratory conditions, very little is known about their performance in the field. This paper presents an approach for estimating the accuracy of these devices for field conditions. Measurements carried out in a tidal channel (mean tidal range of 3.5m) on the German North Sea coast were used. The measurements were gathered simultaneously from two vessels equipped with similar devices travelling side-by-side. 21 transects giving a total of 686 simultaneous profiles were considered. Only velocity profiles with the maximum point velocity magnitude values over 0.3 m/s covering ranges of depth-integrated velocities from 0.28m/s to 1.06m/s were considered in the analysis.

The estimation of the variability of point measurements, required in the computation of the probability intervals, was obtained on the basis of the simultaneous measurements. Logarithmic velocity distributions were fitted through the measured data by using weighted least squares regression techniques allowing the formulation of a point estimator. The accuracy in the resulting depth-integrated velocity values was estimated by computing several probability intervals on the basis of re-sampling techniques. The reduction in accuracy of the devices closer to the sea bottom was also accounted for.

The standard deviation for ADCP point measurements resulted about 0.06m/s and 0.14m/s for distances above and below 1m from the sea bottom respectively. The accuracy of the devices for the depth-integrated velocity values resulted approximately constant and equal to ±0.015m/s. No dependency between the length of the confidence intervals and the water depths or the depth-integrated velocities was identified