Measuring shallow water flow velocity with virtual boundary condition signal in the electrolyte tracer method

• Published in: Journal of hydrology (Amsterdam) 2012-07, Vol.452-453, p.172-179
• Main Authors: Shi, Xiaonan, Zhang, Fan, Lei, Tingwu, Chuo, Ruiyuan, Zhou, Shumei, Yan, Yan
• Format: Article
• Language: English
• Subjects: Boundary conditions; Electrolyte tracer; Electrolytes; Flow rate; Flow velocity; Mathematical models; Sensitivity analysis; Sensors; Shallow water flow; Slope gradients; Velocity measurement; Virtual boundary condition
• ISSN: 0022-1694; 1879-2707
• DOI: 10.1016/j.jhydrol.2012.05.046

► An improved method with higher velocity measurement accurate was proposed. ► The method was tested under various flow rates, slope gradients, and positions. ► Sensitivity analysis on measured velocity was used for methods evaluation. ► The method shows stability and improvement for short distance velocity measurement. Accurate measurement of hill-slope shallow water flow velocity is an important issue in soil erosion prediction. The pulse model of electrolyte tracer advanced by Lei et al. (2005) indicated that the boundary condition is critical for accurate velocity measurement. To improve the measurement accuracy, Lei et al. (2010) improved the method by employing practically measured boundary condition instead of the pulse assumption, by an additional sensor close to the electrolyte injection location. The measurement precision was improved significantly as compared with that by the pulse model. However, the measured boundary condition is not the actual boundary condition, which may still cause errors in the measured velocities. The sensor for boundary measurement can otherwise be used to measure flow velocity if it can be rationally determined without measurement. In this study, modifications are made to the previous methods. The boundary condition is not measured by an additional sensor but is determined by fitting the model to the experimental data so as to determine virtual/real boundary condition (Virtual B.C.) at its actual position and the flow velocity simultaneously. The measured velocities and the estimated boundary condition are verified using the experimentally measured data involved three flow rates (12, 24 and 48Lmin−1), three slope gradients (4°, 8° and 12°) and six measurement positions (0.05, 0.3, 0.6, 0.9, 1.2, and 1.5m). The newly-proposed Virtual B.C. method improved the measurement precision of velocities as compared with those by the previous methods. An additional experimental data set is used for sensitivity analysis. This experiment involves three flow rates (12, 24 and 48Lmin−1), three slope gradients (4°, 8° and 12°), four measurement positions (0.3, 0.6, 0.9, and 1.2m), and three electrolyte injection durations (0.2, 0.3, and 0.4s). The sensitivity analysis on measured velocity indicates that the Virtual B.C. method shows significant advantage of stability, regardless of electrolyte injection duration. The improvement is relatively more significant for shorter distance measurement and shorter electrolyte injection duration.