| 摘要: |
| 将基于最优插值(OI)的同化并行模块植入第三代海浪模式WAVEWATCH III version3.14, 建立数据同化的台风海浪模式预报系统。该系统的强迫风场采用模型台风风场与台风来前海区背景风场混成的风场。以模式后报2010 年7 月严重影响南海北部的“康森”和“灿都”台风引起的海浪场为例, 首先对所构造的混合风场的台风海面风场结构进行定性检验, 并用高度计沿轨风速对混合风场精度进行定量验证。在此基础上, 海浪模式在混合风场强迫下边积分边同化。同化数据采用上述台风过境南海期间Jason-2 卫星高度计沿轨有效波高 (SWH)。值得指出的是, 同化时只取SWH 沿轨数据的一部分用于同化计算, 而另一部分沿轨数据则用于对同化分析结果进行检验。先后同化了4 条轨道上的SWH数据。将SWH 的同化分析与无同化的对照组结果分别与高度计测量SWH 比较, 发现同化较无同化可使均方根误差获得50%以上的明显改进。以同化分析场作为初始场, 同化影响预报(这里是后报)的时效性约在48 h 以内。本研究目的是通过同化高度计SWH 数据进一步提升台风海浪模式预报的准确度。 |
| 关键词: 台风浪 WAVEWATCH III 数据同化 最优插值(OI) 高度计有效波高(SWH) |
| DOI： |
| 分类号: |
| 基金项目:海洋公益性行业科研专项经费项目(201105002, 201005033);国家高技术研究发展计划(863 计划)课题(2012AA091801); NSFC-广东联合基金重点项目(U1133001) |
|
| A data assimilative model hindcast of typhoon waves in the South China Sea |
|
|
| Abstract: |
| The aim of this paper is to improve the accuracy of hindcasting typhoon waves by assimilating altimeter significant wave height (SWH) data into a spectral wave model. A parallel module of data assimilation based on optimal interpolation (OI) was developed and implanted into the full-spectral third-generation wind-wave model WAVEWATCH III version3.14 to set up a data assimilative wave model. The environmental wind field was blended into the tropical storm wind field and employed as surface forcing in the wave model. The hindcasting waves induced by Typhoon Conson (2010) and Chanthu (2010) during their passage over the South China Sea (SCS) were taken as examples. Besides qualitative analysis of structures of the blended wind fields, quantitative evaluations were also conducted using altimeter along-track winds. During the wave model integration, the along-track SWH data observed by Jason-2 satellite altimeter during Typhoon’s passage over SCS were assimilated consecutively. It is worth noting that only one part of the along-track SWH data was combined into the model during the assimilation run, the other part was used for validation of the analysis results. The SWH from the assimilation run and the control run were compared with the Jason-2 along-track SWH data. It was found that the consecutive assimilation of the observed SWH data reduced the root-mean-square error by approximately 50% in analyzed SWH. Furthermore, hindcasts using the analyzed SWH as initial fields indicated that the effect of assimilation on hindcasting may be retained in 48 hours. |
| Key words: typhoon waves WAVEWATCH III data assimilation optimal interpolation (OI) significant wave height (SWH) |
