Ocean Wave Optimization using In-situ Buoy Measurements
Presentation Type
Presentation
Full Name of Faculty Mentor
Erin Hackett, Marine Science; Douglas Pastore (Ph.D. Student)
Major
Marine Science
Presentation Abstract
Ocean-atmosphere interactions are highly dynamic and are largely related to prevailing wind and wave conditions. Accurate modeling of waves in various types of physical models affected by the near-surface region is paramount – such as in numerical weather prediction models, electromagnetic wave (EM) propagation simulations, and climatological models. For example, EM propagation is greatly influenced by forward scattering from the sea surface, thus high-fidelity wave models are commonly used to represent the sea surface. Because measured wave fields can be more complex than their model representation, and high-fidelity simulations often require more information (higher resolution) than buoy measurements can provide, it is not straightforward to use these wave models to replicate wave fields measured by wave buoys. In this study, in-situ buoy data are optimized to a wave model that includes both the wind and capillary wave portion of the wave spectrum, and swell are modeled using a narrow band swell spectral model. Optimization of the in-situ data to the model is performed using particle swarm optimization (PSO), a machine learning technique. The buoy data were collected during the Coupled Air-Sea Processes and Electromagnetic Ducting Research East field campaign in Duck, NC. PSO iteratively estimates parameters of the wave model until the model optimally matches the in-situ (buoy) data. Optimized models are visually compared to the corresponding buoy data as well as quantitatively compared. The results demonstrate how buoy data can be used to optimize a wave model to improve simulations that include a sea surface.
Start Date
13-4-2023 3:00 PM
End Date
13-4-2023 3:20 PM
Disciplines
Oceanography
Recommended Citation
Bruno, Madison, "Ocean Wave Optimization using In-situ Buoy Measurements" (2023). Undergraduate Research Competition. 111.
https://digitalcommons.coastal.edu/ugrc/2023/fullconference/111
Ocean Wave Optimization using In-situ Buoy Measurements
Ocean-atmosphere interactions are highly dynamic and are largely related to prevailing wind and wave conditions. Accurate modeling of waves in various types of physical models affected by the near-surface region is paramount – such as in numerical weather prediction models, electromagnetic wave (EM) propagation simulations, and climatological models. For example, EM propagation is greatly influenced by forward scattering from the sea surface, thus high-fidelity wave models are commonly used to represent the sea surface. Because measured wave fields can be more complex than their model representation, and high-fidelity simulations often require more information (higher resolution) than buoy measurements can provide, it is not straightforward to use these wave models to replicate wave fields measured by wave buoys. In this study, in-situ buoy data are optimized to a wave model that includes both the wind and capillary wave portion of the wave spectrum, and swell are modeled using a narrow band swell spectral model. Optimization of the in-situ data to the model is performed using particle swarm optimization (PSO), a machine learning technique. The buoy data were collected during the Coupled Air-Sea Processes and Electromagnetic Ducting Research East field campaign in Duck, NC. PSO iteratively estimates parameters of the wave model until the model optimally matches the in-situ (buoy) data. Optimized models are visually compared to the corresponding buoy data as well as quantitatively compared. The results demonstrate how buoy data can be used to optimize a wave model to improve simulations that include a sea surface.