ECMWF细网格10 m风预报在象山港海域的评估与订正-海洋预报

ECMWF细网格10 m风预报在象山港海域的评估与订正

单位：1. 武警海警学院航海系, 浙江宁波 315801;
2. 北京师范大学地理科学学部, 北京 100875;
3. 宁波市奉化区气象局, 浙江宁波 315500

分类号：P732.1

出版年·卷·期（页码）：2025·42·第三期（45-55）

摘要：

系统评估了欧洲中期天气预报中心EC细网格10 m风24 h预报产品在象山港海域的准确性，选取2016—2020年1月、4月、7月、10月的EC预报资料，与象山港不同区域3个代表站点的实测风速和风向资料进行对比分析，探讨概率密度匹配法及长短期记忆神经网络模型对预报的订正效能。结果表明：相比于平均风速，极大风速的EC预报结果与象山港海域的实测结果更接近。具体而言，EC对内港6级风的预报效果最好，外港预报效果的季节差异明显，春夏季对4级风的预报效果较好，秋冬季对7级风的预报效果较好。EC在外港极大风速预报上的优势明显，各站点风速预报偏差中位数（0.15～2.50 m/s）维持在较低水平，且春夏季偏差离散性小，秋冬季增大；相反，在内港区域的秋冬季表现出较低的偏差离散性。在风向预报上，EC对偏北、东南及西南风向的预报误差偏小，但对偏南风向的预报能力较差。特别地，对外港区域的风向预报整体表现更为稳定和准确。与概率密度匹配法相比，长短期记忆神经网络模型在风速预报订正上展现出全面优势，尤其在大风速等级下的订正效果明显更优，更有利于提升沿海及近海区域大风速预报的准确性。

This paper systematically evaluates the accuracy of the European Centre for Medium-Range Weather Forecasts（ECMWF） fine-grid 10 m wind forecasts at lead time of 24 hours in the Xiangshan Port area. The 2016-2020 forecasts of January, April, July, and October are validated against the observed wind speed and direction from three representative stations in the Xiangshan Port area. The paper also explores correction effort introduced by probability density matching method and Long Short-Term Memory（LSTM） neural network model. The results show that the peak wind speed forecasts are more accurate than the average wind speed forecasts. The ECMWF forecasts have the best quality for Beaufort scale 6 winds in the inner port area, and exhibit significant seasonal dependence for the outer port area, which shows a better quality for Beaufort scale 4 winds during spring-summer and Beaufort scale 7 winds during autumn-winter. The ECMWF forecasts perform well for maximum wind speeds in the outer port area, with median errors at all sites remaining at a low level（0.15～2.50 m/s）, and the error discreteness is small during spring-summer. The error discreteness is small during autumn-winter in the inner port area. The ECMWF forecasts perform well for northerly, southeasterly, and southwesterly winds, while bed for southerly. The wind direction forecasts are generally accurate in the outer port area. Compared to the probability density matching method, the LSTM algorithm demonstrates a comprehensive advantage in correcting wind speed forecasts, particularly for higher wind speeds, which suggests that the LSTM algorithm is suitable for correcting high wind speed forecasts in coastal and nearshore areas.

参考文献：

[1] 李敏,王辉,金啟华.中国近海海面风场预报方法综述[J].海洋预报, 2009, 26(3):114-120.LI M, WANG H, JIN Q H. A review on the forecast method of China offshore wind[J]. Marine Forecasts, 2009, 26(3):114-120.
[2] 毛绍荣,林镇国,梁健,等.广东沿海强东北季风的概率预报方法[J].热带气象学报, 2003, 19(1):94-100.MAO S R, LIN Z G, LIANG J, et al. The probability forecast technique on strong northeast monsoon along the coast of Guangdong[J]. Journal of Tropical Meteorology, 2003, 19(1):94-100.
[3] 杨景泰,隋玉秀,肖剑,等.大连地区ECMWF细网格10 m风场与观测实况对比与检验[J].气象与环境学报, 2021, 37(1):73-81.YANG J T, SUI Y X, XIAO J, et al. Comparison and verification of 10 m wind field based on ECMWF fine grid and observations in Dalian[J]. Journal of Meteorology and Environment, 2021, 37(1):73-81.
[4] 连喜虎,刘桂艳,高松,等. EC细网格10 m风场产品在渤、黄海海域的预报能力检验与评估[J].海洋预报, 2020, 37(6):74-82.LIAN X H, LIU G Y, GAO S, et al. Evaluation of the ECMWF forecast skills for the 10 m wind field in the Bohai Sea and the Yellow Sea[J]. Marine Forecasts, 2020, 37(6):74-82.
[5] 薄文波,高山红,王永明.基于WRF模式的渤海海面风的预报与订正[J].海洋湖沼通报, 2013(3):37-44.BO W B, GAO S H, WANG Y M. Forecast and correction of the Bohai sea surface wind based on WRF model[J]. Transactions of Oceanology and Limnology, 2013(3):37-44.
[6] ISAKSEN L, JANSSEN P A E M. Impact of ERS scatterometer winds in ECMWF's assimilation system[J]. Quarterly Journal of the Royal Meteorological Society, 2004, 130(600):1793-1814.
[7] LEUTBECHER M, LOCK S J, OLLINAHO P, et al. Stochastic representations of model uncertainties at ECMWF:state of the art and future vision[J]. Quarterly Journal of the Royal Meteorological Society, 2017, 143(707):2315-2339.
[8] 曾瑾瑜,韩美,吴幸毓,等. WRF、EC和T639模式在福建沿海冬半年大风预报中的检验与应用[J].海洋科学, 2015, 39(7):75-85.ZENG J Y, HAN M, WU X Y, et al. Test and application of WRF,EC and T639 Models in strong wind forecasting in Fujian coast in winter half year[J]. Marine Sciences, 2015, 39(7):75-85.
[9] 方艳莹,申华羽,涂小萍,等. ECMWF细网格对浙江沿海10 m风预报性能评估[J].中国农学通报, 2019, 35(13):119-125.FANG Y Y, SHEN H Y, TU X P, et al. ECMWF fine mesh:forecast performance evaluation of 10 m wind in Zhejiang coastal area[J].Chinese Agricultural Science Bulletin, 2019, 35(13):119-125.
[10] 万瑜,曹兴,窦新英,等. ECMWF细网格数值预报产品在乌鲁木齐东南大风预报中的释用[J].沙漠与绿洲气象, 2014, 8(1):32-38.WAN Y, CAO X, DOU X Y, et al. The application of ECMWF refined net numerical forecast data in the southeast gale in Urumqi[J]. Desert and Oasis Meteorology, 2014, 8(1):32-38.
[11] 杨亦萍,史珩瑜,刘力源,等.冬春季动量下传引发浙江沿海灾害大风过程分析[J].海洋预报, 2022, 39(3):66-74.YANG Y P, SHI H Y, LIU L Y, et al. Analysis of the disastrous gale caused by downward transportation of momentum in the coastal areas of Zhejiang province in winter and spring[J]. Marine Forecasts, 2022, 39(3):66-74.
[12] 刘彬贤,陈宏,左涛,等.基于EC集合预报的渤海灾害性大风风速预报方法研究[J].海洋预报, 2018, 35(2):19-26.LIU B X, CHEN H, ZUO T, et al. EC ensemble forecast on severe gale speed in the Bohai Sea[J]. Marine Forecasts, 2018, 35(2):19-26.
[13] 吴俞,冯文,李勋,等. ECMWF细网格10 m风场产品在南海海域的预报检验[J].西南师范大学学报(自然科学版), 2015, 40(9):204-212.WU Y, FENG W, LI X, et al. On forecast test of ECMWF 10 m wind field products in South China Sea[J]. Journal of Southwest China Normal University (Natural Science Edition), 2015, 40(9):204-212.
[14] 方楠,姜舒婕,闫晓敏,等.基于长短期记忆神经网络的风速超短期快速滚动预报技术[J].气象科技, 2022, 50(6):842-850.FANG N, JIANG S J, YAN X M, et al. Research on ultra shortterm fast rolling prediction technology of wind speed based on LSTM neural network[J]. Meteorological Science and Technology, 2022, 50(6):842-850.
[15] 吴迪,田宏强,刘辉,等.基于相似卡尔曼滤波的安徽省WRF模式风速预报订正[J].气象与环境学报, 2023, 39(4):31-37.WU D, TIAN H Q, LIU H, et al. Bias correction of wind speed forecasts for the WRF model in Anhui province based on the analog Kalman filter method[J]. Journal of Meteorology and Environment, 2023, 39(4):31-37.
[16] MONAHAN A H. The Gaussian statistical predictability of wind speeds[J]. Journal of Climate, 2013, 26(15):5563-5577.
[17] 王洁,郭鹏,何晓凤,等.基于测风塔观测资料的近地层风速平面订正[J].气象与环境学报, 2020, 36(6):115-121.WANG J, GUO P, HE X F, et al. Research on the correction method of gridded wind speed data based on wind tower observation[J]. Journal of Meteorology and Environment, 2020,36(6):115-121.
[18] 刘志杰,刘彬贤,王锐,等.基于传统和深度学习技术的黄渤海域大风预报方法研究[J].海洋预报, 2022, 39(6):34-43.LIU Z J, LIU B X, WANG R, et al. Research on the gale prediction methods in the Yellow Sea and Bohai Sea based on traditional and deep learning technologies[J]. Marine Forecasts,2022, 39(6):34-43.
[19] 钱磊,邱学兴,郑淋淋.基于概率密度匹配方法的WRF模式阵风风速误差订正[J].气象科技, 2019, 47(6):916-926.QIAN L, QIU X X, ZHENG L L. Error correction of WRF model gust speed based on probability density function matching method[J]. Meteorological Science and Technology, 2019, 47(6):916-926.
[20] 曹萍萍,肖递祥,徐栋夫,等.基于概率匹配的西南区域模式定量降水订正试验[J].气象科技, 2018, 46(1):102-111.CAO P P, XIAO D X, XU D F, et al. Calibration of SWCWARMS quantitative precipitation based on probability matching method[J]. Meteorological Science and Technology, 2018, 46(1):102-111.
[21] 方楠,谢国权,阮小建,等.长短期记忆神经网络(LSTM)模型在低能见度预报中的应用[J].气象与环境学报, 2022, 38(5):34-41.FANG N, XIE G Q, RUAN X J, et al. Application of long shortterm memory neural network (LSTM) model in low visibility forecast[J]. Journal of Meteorology and Environment, 2022, 38(5):34-41.
[22] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8):1735-1780.
[23] 张丽旭,赵敏,付旭强,等.近22年来象山港海域水环境变化趋势及R/S预测研究[J].海洋湖沼通报, 2010(1):115-120.ZHANG L X, ZHAO M, FU X Q, et al. The variation trend of seawater environment of Xiangshan harbor and study of their R/S forecasting in recent 22 years[J]. Transactions of Oceanology and Limnology, 2010(1):115-120.
[24] 李婷,黄秀清,黄晓琛.基于遥感技术的海岸带的生态综合评价--以象山港为例[J].海洋湖沼通报, 2023, 45(1):90-97.LI T, HUANG X Q, HUANG X C. Integrative assessment of ecosystem based on remote sensing--Xiangshan Bay as a case[J]. Transactions of Oceanology and Limnology, 2023, 45(1):90-97.
[25] 曹蕾,乔馨慧.两次强降雪过程EC预报能力检验[J].黑龙江气象, 2020, 37(2):16-18.CAO L, QIAO X H. Examination of EC forecasting ability during two heavy snowfalls[J]. Heilongjiang Meteorology, 2020, 37(2):16-18.

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】