NCEP/NOAA全球预报产品在江苏近岸及长江口外的精度评估

Accuracy assessment of the NCEP/NOAA global forecasting product around Jiangsu coastal region and Yangtze River estuary

利用2021年1—7月江苏近岸和长江口外13个测风站、3个波浪浮标的观测数据,与美国国家环境预报中心/美国国家海洋和大气管理局(NCEP/NOAA)全球预报系统的风场、波浪预报数据进行对比。结果表明:与上一代预报产品相比,目前全球业务预报产品在风场预报方面的准确度明显提高,江苏近岸地区的风速预报误差明显小于长江口外地区,24 h风速和风向预报的均方根误差分别为2 m/s和45°。对大风天、大浪天环境下全球预报产品精度的评估表明:大风天风速预报精度在长江口外降低,在江苏近岸无明显变化,风向预报精度在长江口外无明显变化,在江苏近岸提高;大浪天波高、浪向的预报精度分别降低、提高。在台风极端天气条件下,全球预报产品对长江口的风场仍有一定的预报能力,但预报的最大风速的出现时间滞后6~9 h,且无法预报出风速剧烈变化的情况。尝试基于长短期记忆网络方法利用NCEP/NOAA预报及观测改进业务预报,可以改善短期单站预报水平。

The observation data of 13 wind stations and 3 wave buoys around Jiangsu coastal region and Yangtze River estuary from January to July 2021 are used to compare with the wind and wave forecasts from the National Centers for Environmental Prediction/National Oceanic and Atmospheric Administration(NCEP/NOAA)global forecasting system.The results show that compared with the previous generation version of the NCEP/NOAA global forecasting system,the current version improves significantly in wind forecasts.The wind speed forecast error outside the Yangtze River estuary is greater than that in the Jiangsu coastal region,where the root mean squared errors of 24 h wind speed and direction forecasts are 2 m/s and 45 degrees,respectively.Wind speed forecasting accuracy on windy days decreases outside the Yangtze River estuary,but does not change obviously in the Jiangsu coastal region.Wind direction forecasting accuracy does not change obviously outside the Yangtze River estuary,but increases in the Jiangsu coastal region.The accuracy of wave height and direction forecasts on high wave days decreases and increases,respectively.Under extreme weather conditions,the global forecasting system also demonstrates capability to forecast wind around the Yangtze River estuary,but the timing of maximum wind speed is delayed by 6~9 h,and the drastic wind speed change cannot be predicted.Combing the Long Short-Term Memory method and the NCEP/NOAA data can improve the short-term single-station forecast.