Measuring Waves in Difficult Places: New Approaches to Observing Waves in Hurricanes and Sea Ice
Guardado en:
| Publicado en: | ProQuest Dissertations and Theses (2025) |
|---|---|
| Autor principal: | |
| Publicado: |
ProQuest Dissertations & Theses
|
| Materias: | |
| Acceso en línea: | Citation/Abstract Full Text - PDF |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Resumen: | New approaches are applied to study ocean surface wave dynamics in hurricanes and the coastal Arctic. In hurricanes, arrays of drifting buoys and airborne radar are used to characterize the wind speed dependence and spatial distribution of ocean surface roughness caused by waves. Hurricane-generated waves, and the drag imparted by their roughness, contribute to coastal flooding, cause infrastructure damage, and modify exchanges of momentum and heat between the atmosphere and ocean—important controls on storm intensification offshore. In the coastal Arctic, a novel method called Distributed Acoustic Sensing is combined with machine learning to measure waves from a submarine fiber-optic cable offshore of Oliktok Point, Alaska. This system can be used for the subsequent study of wave-ice interactions and to monitor wave action in regions susceptible to coastal change. Both approaches leverage innovative, low-cost sensing modalities to measure waves in hard-to-reach environments with exceptional spatial resolution. These measurements enhance our ability to understand, monitor, and predict the impacts of the ocean on coastlines.Drifting buoy observations in hurricanes Ian (2022) and Fiona (2022) are merged with modeled surface wind speeds to determine the evolution of wave slope at high wind speeds. Wave slope is quantified using the mean square slope, which is commonly used as proxy for ocean surface roughness. At low-to-moderate wind speeds (≤ 15 m s-1), slopes increase linearly with wind speed. At higher winds (> 15 m s-1), slopes continue to increase, but at a reduced rate. At extreme winds (> 30 m s-1), slopes asymptote. The mean square slopes are directly related to the wave spectral shapes, which over the resolved frequency range (0.03 to 0.5 Hz) are characterized by an equilibrium tail ( −4) at moderate winds and a saturation tail ( −5) at higher winds. The asymptotic behavior of wave slope as a function of wind speed could contribute to the reduction of surface drag at high wind speeds.An airborne radar is then combined with the drifting wave buoys to provide a multiscale view of hurricane-generated waves. Wave slopes measured by the radar, which include waves 0.2 m and longer, saturate in a similar manner to the buoy-measured slopes. A method to infer the shape of the spectral tail from 0.5 Hz to 3 Hz using collocated mean square slope observations from each instrument is introduced. The method is able to recover the frequency -5 tail characteristic of the saturation range expected at these frequencies based on theory.Next, a dense array of buoy observations in Hurricane Idalia (2023) is used to investigate the spatial distribution and dependence of mean square slope on wind, wave, and storm characteristics. Inside Hurricane Idalia, buoy-measured mean square slopes have a secondary dependence on wind-wave alignment: at a given wind speed, slopes are higher where wind and waves are aligned compared to where wind and waves are crossing. At moderate wind speeds, differences in mean square slope between aligned and crossing conditions can vary 15% to 20% relative to their mean. These changes in wave slope may be related to the reported dependence of air-sea drag coefficient on wind-wave alignment.Lastly, in the coastal Arctic, two new data-driven models for estimating ocean surface waves from distributed acoustic sensing (DAS) submarine cable strain rate are developed using supervised machine learning. The new models are trained on target data from pressure moorings at three sites along 27.1 km of cable and are benchmarked against an empirical transfer function method previously used to estimate waves from DAS. A model which uses convolutional neural networks to transform frequency-wavenumber spectra to pressure spectra outperforms the benchmark in wave height and period prediction when evaluated on the cable at Oliktok Point. Regression-based machine learning is useful for estimating waves from DAS data when the pressure-strain relationship varies temporally and spatially across different wave conditions. |
|---|---|
| ISBN: | 9798293848775 |
| Fuente: | ProQuest Dissertations & Theses Global |