Wind Speed Extraction Based on High Frequency Radar Retrieved Wind-Driven Current
| Autor: | Yingwei Tian, Biyang Wen, Cui Wen |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Field (physics)
Mean squared error Correlation coefficient Astrophysics::High Energy Astrophysical Phenomena Geotechnical Engineering and Engineering Geology Wind speed law.invention law Physics::Space Physics Wave height Range (statistics) Astrophysics::Solar and Stellar Astrophysics Environmental science Electrical and Electronic Engineering Radar Image resolution Physics::Atmospheric and Oceanic Physics Remote sensing |
| Zdroj: | IEEE Geoscience and Remote Sensing Letters. 18:1555-1559 |
| ISSN: | 1558-0571 1545-598X |
| DOI: | 10.1109/lgrs.2020.3004402 |
| Popis: | High frequency (HF) radar often indirectly inverts wind field from wave height field, so the accuracy, range, and spatial resolution of the estimated wind field are usually limited by the wave field estimation performance. However, the sea surface current including the wind-driven current can often be accurately measured by radar. Thus, based on the wind-driven current estimated by HF radar, a new wind speed inversion algorithm is proposed in this letter. First, the relationship between wind speed and wind-driven current speed is established according to the ocean dynamics theory, but it contains several uncertain parameters. To avoid solving these parameters, an artificial neural network is used to train an accurate model between wind-driven current speed and wind speed. Final, the wind-driven current estimated by radar is substituted into the model to extract wind speed. A field experiment shows that the average correlation coefficient between the radar-estimated wind speed and the reference wind speed is 0.86, and the average root mean square error is 2.38 m/s. In addition, the proposed algorithm has larger measurement range and better spatial resolution than traditional algorithms. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|