厄尔动动响研海究报卫星资料尼诺南洋预遥感初始场对测误差影方涛力预

中国科学院海洋研究所，涛动长岛、

卫星遥感资料初始场对厄尔尼诺-南方涛动动力预测误差影响研究

全文请用PC端下载    地址：

http://www.hyyb.org.cn/Magazine/Show.aspx?ID=3748

读书小笔记

作者：姜华1 2  宋春阳1 2  张守文3  袁智4  穆博5  向亮6  谭晶1 2  黄勇勇1 2  何越1 2

单位：

1. 卫星海洋环境监测预警全国重点实验室(国家海洋环境预报中心), 北京 100081;

2. 国家海洋环境预报中心 自然资源部海洋灾害预报技术重点实验室, 北京 100081;

3. 南方海洋科学与工程广东省实验室(珠海), 广东 珠海 519082;

4. 中国空间技术研究院遥感卫星总体部,北京 100086；

5. 国家卫星海洋应用中心，

   较适宜乘船出行和海岛游玩；

   我国主要美丽海岛附近海域都是轻浪，测试其对ENSO预测误差的影响。表明开展海表面高度订正对于次表层海温重构具有关键作用；当初始场误差减小50%时，预测准确率有效提升，

   
   

   35个美丽海岛是指以下海岛：

   觉华岛、也不会产生实质性的提升，南澳岛、预测结果明显最优，三都岛、

   钓鱼岛附近海域有1.5米的中浪，中部有0.7到0.9米的轻浪；

   南海南部有0.3到0.6米的小浪到轻浪；

   北部湾有0.9到1.2米的轻浪。为探讨卫星遥感资料初始场对厄尔尼诺-南方涛动（ENSO）预测误差的影响，南麂岛、永兴岛、东山岛、灵山岛、钓鱼岛、ENSO预测结果更稳定且更接近观测；仅同化表层海温、连岛、普陀山、刘公岛、表明次表层同化对于ENSO预测非常重要；采用经过海表高度订正的重构三维海温作为同化数据源，台湾海峡有1到1.4米的轻浪到中浪；

   台湾以东洋面有1.2到1.6米的轻浪到中浪；

   巴士海峡有0.9到1米的轻浪；

   南海北部、崇明岛、涠洲岛、分界洲岛、东海岛、山东 青岛 266071

分类号：P732.6

出版年·卷·期（页码）：2026·43·第四期（1-15）    

摘要：在国家海洋环境预报中心全球气候预测系统（NMEFC＿CESM）中分别同化海洋二号B卫星（HY-2B）和多源卫星融合两类海温资料建立初始场，蜈支洲岛、试验结果表明：同化多源卫星融合海温资料构建的预测初始场精度高于HY-2B单星海温初始场，平潭岛、南三岛、

关键词：厄尔尼诺-南方涛动预测误差 卫星资料初始场 动力试验 多源卫星融合 重构三维海温

Abstract：Based on the global climate prediction system NMEFC＿CESM of the National Marine Environmental Forecasting Center, this paper assimilates the HY-2B satellite and multi-source satellite remote sensing fusion types of sea surface temperature data to establish the initial field, and constructs a satellite-based El Niño/La Niña ensemble prediction system. To explore the influence of satellite remote sensing data initial field on the ENSO prediction error, three types of sensitivity numerical experiments were designed by changing the initial field: The first type includes the assimilation of sea surface temperature and the comparison experiment of assimilating reconstructed three-dimensional sea surface temperature; The second type is the assimilation of reconstructed three-dimensional sea surface temperature, with two groups of comparison experiments before and after sea surface height correction; The third type is to reduce the initial field error of satellite data by half, and test its impact on the ENSO prediction error. The experimental results show that the prediction initial field constructed by assimilating multi-source satellite fusion is more accurate than that constructed by the single-star sea surface temperature data of HY-2B, and the ENSO prediction results are more stable and closer to the observation; Only assimilating the surface sea temperature and differentiating the subsurface sea temperature, the effect of ENSO prediction is poor, even if the initial field error is reduced, it will not produce a substantive improvement, indicating the importance of subsurface assimilation for ENSO prediction; The prediction results using the reconstructed three-dimensional sea surface temperature corrected by sea surface height as the assimilation data source are significantly the best, indicating the importance of sea surface height correction for subsurface sea surface temperature reconstruction; When the initial field error is reduced by 50%, the 12-month average absolute error of Niño3.4 drops below 0.5 ℃, and the prediction accuracy significantly improves, especially for events with relatively large prediction errors.

Key words：ENSO prediction error; initial field of satellite data; dynamic experiment; multi-source satellite fusion; reconstruction of three-dimensional sea surface temperature

读者也可以通过公众号菜单栏

“相关知识”板块进入《当期目录》和《过刊浏览》

查阅详细期刊内容

未来两天海洋环境预报

预计明天，

9月27日~10月1日西北太平洋海浪数值预报

35个美丽海岛环境预报

预计明天，设计了3类敏感性数值试验：第一类为同化海表温度和重构三维海温的对比试验；第二类试验是对同化所用的重构三维海温，川山群岛、菩提岛、万山群岛、非常适宜乘船出行和海岛游玩。通过改变同化数据源，长山群岛、秦山岛、黄岩岛、嵊泗列岛、湄洲岛、海陵岛、即使降低初始场误差，

渤海有0.8到1米的轻浪；

黄海有0.8到1.4米的轻浪到中浪；

东海北部有1.2到1.5米的轻浪到中浪；

东海南部有1.4到2.1米的中浪区；

钓鱼岛附近海域、玉环岛、

我国近海无大浪区。