Investigating the Discrepancies between Acoustic and Radar Measurements of Water Level Using Data Analysis and Computational Modeling

Observations indicate that there is a discrepancy between water level measurements from NOAA’s acoustic gauges and radar gauges. NOAA’s acoustic gauges use a downward-looking acoustic sensor that measures the instantaneous distance from the sensor to the water level. To create a calm water surface for measurements, the sensor is placed in a PVC tube that penetrates the water surface to a certain depth. This tube is called the stilling well. In this thesis, discrepancies in water level measurements from acoustic and radar sensors were explored through data analysis and computational modeling. Specifically, the correlations between the error, which is defined as the discrepancy between the water level measured by the acoustic sensor (inside the stilling well) and the true water level measured by the radar sensor (outside the well), and important environmental factors are examined. Furthermore, a computational model was developed to simulate various scenarios of interactions between the flow and the stilling well. The model was validated with the experimental study of Shih et al. (1981) and applied to various scenarios to examine the possible sources of errors. Gaussian humps and pits were added at the bottom of the domain to study the effect of the scour pit and pile up of soil at the bottom of the stilling well on the error. Dependencies of error on the distance between the well’s bottom and the sea bottom is also analyzed. Furthermore, the effect of supporting piles on the errors in water level measurements was explored for different current velocities, and cases with various well-to-pile distances were simulated.