A Global Coastal Sea Surface Wind Field Correction Method for HY-2C Scatterometer
Lanjie Zhang,Xuchen Dai,Yanfang LV,Biao Zhang,Haoming Song,Yibo Zhang
DOI: https://doi.org/10.1109/jstars.2024.3465607
IF: 4.715
2024-10-16
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:The coastal sea surface wind (SSW) field can be influenced by factors like land pollution and land terrain characteristics, resulting in lower accuracy than the open ocean. This article proposes an SSW field correction method that combines the light gradient boosting machine model with the cascade forest framework (CFLGBM) using Haiyang-2C (HY-2C) scatterometer to improve the quality of SSW field retrieval in global coastal regions. This method considers precipitation, sea surface temperature, and ocean currents, which influence the SSW, achieving significant correction performance in global coastal regions. Relative to the ERA5 wind field, the corrected SSW speed (SSWS) and SSW direction (SSWD) show root mean square errors (RMSE) of 0.89 m/s and 19.28°, with improvements of 40.67% and 36.93% over the HY-2C Level 2B (L2B) product. In comparison with the National Data Buoy Center buoy, the corrected SSWS and SSWD reduced the RMSE of the HY-2C L2B product by 40.69% and 29.77%, respectively. Moreover, the variation in RMSE of the corrected SSW field with distance from the coastline was analyzed. The corrected SSW field shows the most notable improvement within 0–50 km of the coastline, with up to 0.97 m/s reduction for SSWS and 15.69° reduction for SSWD. The CFLGBM SSW field correction method provides more accurate coastal wind field data for HY-2C and can be further extended to other scatterometers for coastal SSW field correction.
imaging science & photographic technology,remote sensing,engineering, electrical & electronic,geography, physical