Uncertainty analysis for non-intrusive measurement of river discharge using image velocimetry

To overcome operational difficulties of conventional instruments for measuring river discharge during high flows, a Mobile, Large-Scale Particle Image Velocimetry (MLSPIV) system has been designed and constructed. MLSPIV is composed of a LSPIV image processing program, an imaging device (digital camera or video camcorder), a camera positioning system, a controlling program for the camera and positioning system, and a truck for deployment to field locations. Non-intrusive measurement of river discharge using MLSPIV requires special consideration of several technical aspects: image transformation; use of favorable seeding surrogates and illumination conditions; identification of MLSPIV's error sources; and documentation of the total uncertainty of the discharge measurement. The purpose of this dissertation is to review these technical aspects, set up the MLSPIV system, estimate the uncertainty of velocity and discharge measured by MLSPIV in field conditions, verify its measured results with alternative technology, and demonstrate the potential of the newly developed discharge estimation method.;The framework for uncertainty analysis reveals that MLSPIV measurements are affected by twenty-seven elemental error sources. Most of these error sources were assessed through laboratory and field experiments and numerical simulations in this research. All the error sources are propagated through the data reduction equation for estimation of the accuracy of velocity and discharge measurements. A sensitivity analysis was conducted to rank of the error sources based on their contribution to the velocity measurement error.;Validation of MLSPIV was evaluated through field discharge measurements at USGS gauging stations. The existing stage-discharge relationships at the sites were used as the primary validation reference. Alternative technology such as a StreamPro ADCP and a real time LSPIV (RTLSPIV) were used as the secondary reference for MLSPIV discharge measurement results.;Lastly, the performance of MLSPIV was evaluated through an extensive series of measurements conducted during a rainfall event. The series of 43 MLSPIV measurements were aimed to capture the dynamics of the stream flow event during the rising and falling stages of a natural hydrograph and assess the validity of the current rating curve-based discharge estimation for monitoring stream discharges.