Back to Search
Start Over
Data-driven modeling of Bay-Ocean wave spectra at bridge-tunnel crossing of Chesapeake Bay, USA.
- Source :
-
Applied Ocean Research . Jun2023, Vol. 135, pN.PAG-N.PAG. 1p. - Publication Year :
- 2023
-
Abstract
- • Developed data-driven models to estimate wave parameters and spectra near the CBBT. • Provided a framework to predict wave spectra in the frequency domain near bay entrances. • Used the developed models to hindcast wave spectra during the Halloween storm. The Chesapeake Bay Bridge-Tunnel (CBBT) was designed in the early 1960s and first opened on April 15, 1964. It is a 28.3-km bridge-tunnel system that crosses the mouth of Chesapeake Bay. Because there is a lack of reliable long-term observations of surface waves near the Chesapeake Bay entrance, accurate forecasts and hindcasts of wave conditions are essential for maintaining and expanding the bridge-tunnel infrastructure. To estimate wave parameters and energy spectra near the CBBT, novel composite data-driven models were developed using the wind, water level, and offshore wave data as input. The developed models provide satisfactory predictions of both integral wave parameters and energy density spectra of sea and swell waves at the Chesapeake Bay entrance. The developed models can rapidly hindcast the wave characteristics and spectra during an extreme event (i.e., the Halloween storm in 1991). This paper provides a novel framework for developing surrogate models to predict wave spectra in the frequency domain and hindcast historical wave climate, which can be applied to other sea-crossing bridges and/or tunnel sites near bay entrances. The data-driven models, based on deep neural networks, allow for estimating waves without a high demand for computational resources, and thus serve as a useful tool for the characterization and simulation of the complex wave environment at the interface of estuary and ocean. [ABSTRACT FROM AUTHOR]
Details
- Language :
- English
- ISSN :
- 01411187
- Volume :
- 135
- Database :
- Academic Search Index
- Journal :
- Applied Ocean Research
- Publication Type :
- Academic Journal
- Accession number :
- 163339591
- Full Text :
- https://doi.org/10.1016/j.apor.2023.103537