Wind Speed and Direction Estimation from Wave Spectra using Deep Learning

Jiang, Haoyu

doi:10.5194/amt-2021-279

Cited by 2 publications

(2 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Through iterative training, MLP can automatically learn and extract features from input data, enabling classification and prediction of complex problems. It possesses strong nonlinear modeling capabilities and finds applications in areas such as image recognition, data analysis, and prediction [17][18][19] . To construct an MLP prediction model, part of the wind speed data measured by the faulty lidar A is used as input samples, while the wind speed data measured by the normally functioning lidar B at the same time is used as the expected output.…”

Section: Methodsmentioning

confidence: 99%

Interpolation of wind lidar data supported by MLP

Jiang,

Liu,

Wang

et al. 2024

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

To address the issue of invalid values in the measured wind speed data of the wind lidar, a data interpolation experiment was conducted by using the wind speed data outputted by the wind lidar as the base data source. Based on data preprocessing, the interpolation model was constructed by integrating MLP and the mean wind speed shear index. The accuracy of the interpolated wind speed data was evaluated. The experimental results show that the proposed interpolation method for wind lidar data achieved high accuracy, with a root mean square error of less than 0.75 m/s and an average relative error of less than 3.5%. This approach enhances the reliability and integrity of wind lidar data, providing a reference for the use of wind lidar data in wind resource assessment.

show abstract

Section: Methodsmentioning

confidence: 99%

Interpolation of wind lidar data supported by MLP

Jiang,

Liu,

Wang

et al. 2024

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…(2020) demonstrated the validity of ocean surface wind estimation based on the relationship with ocean wave spectra in the high‐frequency range measured by ocean wave buoys. Jiang (2022) established a deep neural network for the estimation of sea surface wind from wave spectra, although the relationship between high‐frequency ocean wave spectra and wind was not explicitly used. A small GPS wave buoy, Spotter (Houghton et al., 2021), incorporates a wind estimation function based on the methods of Thomson et al.…”

Section: Introductionmentioning

confidence: 99%

Ocean Surface Wind Estimation From Waves Based on Small GPS Buoy Observations in a Bay and the Open Ocean

et al. 2022

View full text Add to dashboard Cite

Ocean surface wind and wave information is important in a wide variety of areas, such as coastal disaster reduction, offshore structure design, and atmosphere‐ocean flux estimation. This study proposed a new method for ocean surface wind estimation from surface wave spectrum information measured by small global positioning system buoys. The concept of this method relies on the assumption that the high‐frequency part of the ocean‐wave spectrum is proportional to u∗f−4 ${u}_{\ast }{f}^{-4}$ where u∗ ${u}_{\ast }$ is the friction velocity and f $f$ is the frequency. The determination algorithm for the coefficient of f−4 ${f}^{-4}$ was optimized in this study. The wind direction was determined by the wave cross‐spectrum, assuming that the wind direction aligns the propagation direction of the high‐frequency part of the wave. The proposed wind estimation method was applied to bay and open ocean observations, and the performance of the proposed wind estimation method was similar between the bay and the open ocean. The proposed method improves the wind estimation especially in coastal areas and at high wind speeds in the open ocean compared with the previous method. The performance of the method of the previous study differs between the bay and open ocean due to their spectral shape differences. High‐quality wind and wave information can be obtained using the proposed method. If the mass deployment of small drifting buoys covered the global ocean, the information based on the proposed method could be considerably powerful, and could compensate for the weakness of satellite‐based wind and wave estimations.

show abstract

Wind Speed and Direction Estimation from Wave Spectra using Deep Learning

Cited by 2 publications

References 11 publications

Interpolation of wind lidar data supported by MLP

Interpolation of wind lidar data supported by MLP

Ocean Surface Wind Estimation From Waves Based on Small GPS Buoy Observations in a Bay and the Open Ocean

Contact Info

Product

Resources

About