地波雷达表层径向流资料的多重网格三维变分同化方法研究

The HF radar surface radial currents data assimilation is a important method to raise the accuracy of marine forcast significantly, and the studing of the assimilation method is a frontier topic. The traditonal assimilation methods always be single-scale and most of the methods only can work when there are two or more than two radars exist, furtermore, the adjustment of dynamic consistency is concerned rarely in most traditional methods. This study aims to improve the HF radar surface radial currents data assimilation method from three respects: firstly, the study will use the multigrid 3D-Var method to assimilate the HF radar surface radial currents directly, the method can extract the multi-scale observation information of longwaves and shortwaves in turn correctly and can take the correlation of HF radar currents temporal and spatial characters into accound, that can make up for the weakness of the traditional method only extracting the sigle-scale information; secondly, in order to get the more reasonable and sufficient information of the currents, the shallow water wave eqation will be imported into the cost function as a weak constraint condition to adjust the momentum field and mass field; thirdly, the study will assimilate the HF radar surface radial currents directly, which can make a complete utilization of the radar data. The work will provide a new assimilation method for HF radar and supply better analysis field for the marine environment forecast, that can raise the accuracy of marine forcast significantly. Therefore , the study have the great significance in the academic and good application prospect.

地波雷达表层径向流资料同化是有效提高近海海洋环境预报准确度的重要手段之一,其同化方法研究属国内外前沿课题。然而其传统同化方法均是单尺度的,大多只能同化多颗雷达的交叠区数据,且较少涉及动力一致性调整。本研究从三个方面改进地波雷达表层海流数据同化方法:一是采用多重网格三维变分法同化地波雷达表层径向流数据,利用其从长波到短波依次提取多尺度观测信息的优点,有效考虑观测数据的时空相关,弥补传统同化方法只能提取单尺度信息的不足;二是在同化中将浅水波方程作为弱约束条件引入到目标泛函中,对动量场和质量场同时进行调整,获得物理依据更为充分的海流信息;三是直接进行地波雷达表层径向流资料同化,充分利用非多颗雷达信号交叠区的数据资料。本研究工作将能够获得一种新的地波雷达径向流数据同化方法,为海洋数值预报提供更好的分析场,有效提高海洋环境预报准确性,因此具有重要的科学意义和良好的应用前景。

本项目基于传统三维变分数据同化思想，发展了地波雷达表层径向流多重网格三维变分数据同化方法，并将浅水波方程作为弱约束条件引入到目标泛函中，研究了一种包含动力约束的地波雷达表层径向流资料多重网格三维变分数据同化方法。该同化方法具有能够由长波到短波依次提取多尺度信息的优点，与传统三维变分同化方法相比能够有效提高表层海流同化的效率，降低误差，并且对中小尺度信息提取的效果有明显提升，此外，浅水波方程的引入，实现了动量场和质量场的同时调整，也实现了优化海底底摩擦系数的功能，通过该方法获得的海底底摩擦系数在POMgcs数值模式中的引入，能够有效改善海表面高度场的模拟效果，降低模拟误差。. 该项目研究成果能够为我国近海海洋环境预报提供更好的分析场，有效提高海洋环境预报精度，为沿海重点保障区域防灾减灾提供有效保障，具有良好的应用前景。