159 Estimation of Wave Height from Standard Deviation of Water Level Measured by a Low-Cost Water Level Sensor

Monday, 29 January 2024
Hall E (The Baltimore Convention Center)
Cliff Ehrke, San José State University, San José, CA; and P. E. Tissot, M. Vicens-Miquel, B. Estrada Jr., K. Mukai, and B. Glazer

As water levels continue to consistently rise, the frequency of coastal flooding in the United States is increasing rapidly. A vital measurement for coastal flood modeling on the beaches of large water bodies is wave run-up. However, despite national efforts to increase the availability and reliability of coastal monitoring stations, the Texas coastline still has no real time nearshore wave measurements. Past studies have shown that the standard deviation of measured water level time series can be correlated with the significant wave height (Kirk et al., 2022) for open coasts and freshwater NWLON stations. The purpose of this project is to test if the standard deviation of measured water level from a relatively new active microwave sensor made by Hohonu, Inc. can be a good indication of significant wave height. This private company develops accessible water level sensors and is helping to monitor water levels on broader geographical scale where the water level information is needed for decision making, but the National Ocean Service standards for installing and maintaining a water level station are not needed. The sensor was deployed on Horace Caldwell Pier in Port Aransas, TX and consistent measurements began on May 1st, 2023. To build a relationship between water level standard deviation and significant wave height the latter was estimated by using the Nearshore Wave Prediction System (NWPS) shortest lead time predictions at coordinates closest to the sensor. The nearshore spatial resolution of nwps predictions is as small as 500m. Upon installation, water level measurements were verified with a nearby NOAA tide gauge. The station is protected by jetties but water levels are equivalent. , A simple 1-2-1 filter was first applied to the Hohonu standard deviation time series to reduce noise. Once data had been collected, a nearest neighbor interpolation was used to time-sync measured data with the nwps model predictions, and outliers were removed from the model data. While building the model between standard deviation and significant wave height, various cutoff heights were tested to account for the lack of data for measured high wave events and improve the model correlation. Results and model options comparisons will be presented with the data found to be highly correlated and to be best represented by a quadratic polynomial at all cutoff heights.to potentially estimate wave height in a specific location along the Texas coastline.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner