基于漂浮式测风激光雷达观测数据的海上预报风速订正OA北大核心CSTPCD

This study focused on correcting short-term wind speed forecasts extracted from GFS global forecast data sets at corresponding locations and heights.Data measurement mode and calibration method of the LiDAR data were discussed,and wind speed at three different heights(50m,80m,and 100m)from floating LiDARs were employed.The results indicated that directly using the Kalman filter method with sliding adaptive weights to generate the average hysteresis deviation to correct the wind speed did not yield satisfactory results.Statistical analysis suggested that better standard deviation results can be obtained by lowering the ratio of the average hysteresis deviation in the forecast correction formula.When the average hysteresis deviation ratio was reduced to 0.1-0.5,the wind speed standard deviation after correction by the Kalman filtering method was generally lower than the wind speed standard deviation before correction,and the best ratio was 0.3.Using this hysteresis deviation ratio for wind speed correction at different forecast times and periods for these three LiDAR devices revealed that,except for individual data sets,more than 90%of the wind speed data sets exhibited reduced standard deviation after correction compared to the uncorrected data sets.After the introduction of this improved algorithm,the wind speed forecast improvement rate at each height ranged from 5%to 20%.