热带海洋学报 ›› 2010, Vol. 29 ›› Issue (5): 17-23.doi: 10.11978/j.issn.1009-5470.2010.05.017cstr: 32234.14.j.issn.1009-5470.2010.05.017

• 海洋水文学 • 上一篇    下一篇

Envisat ASAR海浪谱资料的最优插值同化试验

任启峰1,2, 张杰3,4, 尹训强3,4, 杨永增3,4   

  1. 1. 中国科学院海洋研究所, 山东 青岛 266071; 2. 中国科学院研究生院, 北京 100039; 3. 国家海洋局第一海洋研究所, 山东 青岛 266061; 4. 海洋环境科学与数值模拟国家海洋局重点实验室, 山东 青岛 266061
  • 收稿日期:2010-01-06 修回日期:2010-03-23 出版日期:2010-10-30 发布日期:2011-10-10
  • 作者简介:任启峰(1980—), 男, 山东省莱芜市人, 在读博士生, 主要从事海浪资料同化研究。E-mail: renqifeng@yeah.net
  • 基金资助:

    海洋公益性行业科研专项(200705027)

Optimal interpolation assimilation experiments based on Envisat ASAR ocean wave spectral data

1. Institute of Oceanology, CAS, Qingdao 266071, China; 2. Graduate University of CAS, Beijing 100039, China; 3. First Institute of Oceanography, SOA, Qingdao 266061, China; 4. Key Laboratory of Marine Science and Numerical Modeling (MASNUM), SOA, Qingdao 266061, China   

  1. 1. Institute of Oceanology, CAS, Qingdao 266071, China; 2. Graduate University of CAS, Beijing 100039, China; 3. First Institute of Oceanography, SOA, Qingdao 266061, China; 4. Key Laboratory of Marine Science and Numerical Modeling (MASNUM), SOA, Qingdao 266061, China
  • Received:2010-01-06 Revised:2010-03-23 Online:2010-10-30 Published:2011-10-10
  • About author:任启峰(1980—), 男, 山东省莱芜市人, 在读博士生, 主要从事海浪资料同化研究。E-mail: renqifeng@yeah.net
  • Supported by:

    海洋公益性行业科研专项(200705027)

摘要:

基于第三代海浪数值模式LAGFD-WAM, 分别利用4种不同形式的各向同性背景误差相关函数进行了Envisat 高级合成孔径雷达(ASAR)海浪谱资料的最优插值同化试验。与4个浮标实测数据的比较验证表明, ASAR海浪谱资料的最优插值能够有效地改进海浪模式有效波高的模拟。4种不同形式的各向同性背景误差相关函数的同化结果相差不大, 决定同化效果好坏的关键是对相关距离尺度的选取。针对自回归形式的背景误差相关函数的试验结果表明, 相关距离尺度的量级在400—500km时同化效果最好, 此时同化后的模式有效波高均方根误差比未同化时相对减小了26%。

关键词:  海浪谱, 最优插值, 背景误差相关函数, 相关距离尺度

Abstract:

With the third generation wave model named LAGFD-WAM, optimal interpolation assimilation experiments are performed based on Envisat Advanced Synthetic Aperture Radar (ASAR) ocean wave spectral data using four different iso-tropic background error correlation functions. The experiment results are compared with observation data from four different moored buoys. The results show that the optimal interpolation of ASAR wave spectral data can effectively improve the sig-nificant wave height (SWH) simulation of the wave model. The differences of assimilation effects among four experiments with different background error correlation functions are not obvious. The key to assimilation effects lies in the choice of cor-relation scale length. For the auto-regressive background error correlation function, the experiment results indicate that the assimilation effect is best when the correlation scale length is assumed to be from 400 to 500km, and that the root-mean-square error of model SWH data relatively decreases by 26% compared with that without assimilation.

Key words: ocean wave spectrum, optimal interpolation, background error correlation function, correlation scale length