Study On Multiple Time Scales Data Analysis And Assimilation Of Ocean Currents Based On HF Radar

Ocean current is one of the most basic elements in oceanographic research,which has an important effect on physical,chemical and biological processes.As a new observational method,HF radar has advantages compared to traditional observations.This method is featured with large observational domain and high temporal and spatial resolutions,which has great values in the study of dynamic characteristics of offshore ocean currents and numerical assimilation of models.HF radar has been improved continuously since its development,and its accuracy has been able to fulfill the actual research needs,but there are still some problems in its applications.For example,when studying the dynamic characteristics of offshore ocean currents,previous studies lacked the spatial and temporal integrity of the region.Meanwhile,there are few studies on extreme events.The data assimilation of HF radar is mostly based on structured mesh,but there are limited studies on unstructured mesh.If the scheme of structured mesh is directly applied to unstructured mesh model,the errors caused by frequent interpolation between structured and unstructured meshes will significantly impact the model accuracy.In this research,the HF radar data of Zhoushan and Jiangsu sea areas were obtained and the abnormal data were eliminated by quality control.Based on years of HF radar data in Zhoushan sea area,the dynamic process of surface ocean current at multiple time scales is interpreted.Based on the FVCOM model,a ThreeDimensional Variational Assimilation(3D-VAR)module for unstructured meshes was developed,and a universal radar assimilation scheme was established.The scheme was applied to Zhoushan and Jiangsu models,the assimilation test of HF radar in two sea areas was carried out,and the assimilation effect was analyzed.The results are summarized as follows:(1)Zhoushan sea area has regular semi-diurnal tides,and the current movement is mainly dominated by clockwise rotating current.Interannual variation of the residual current in the Zhoushan sea area is not significant,however,obvious intermonth variation of the residual current was detected.Strong southward coastal residual current in winter and strong northward residual current in summer.The correlation between wind and residual current was analyzed.During periods of high wind speed,especially during extreme events,The correlation coefficient and complex correlation coefficient of wind and residual velocity are above 0.9.In conclusion,the analysis of HF radar data can well reflect the temporal and spatial characteristics of ocean currents in Zhoushan sea area.(2)A 3D-VAR assimilation module for unstructured mesh was developed based on FVCOM,the 3D-VAR scheme was used to carry out surface current assimilation in space and the Nudging scheme was used to extend the assimilation effect of 3D-VAR to the time dimension.A universal radar assimilation scheme was constructed based on the established assimilation module,which can automatically download the radar data from the server and complete the processing and quality control of the radar data,and then carry out surface current field assimilation.In addition,for the purpose of operationalization,the scheme can set different assimilation parameters according to different types of radars,which is extensible.(3)The radar assimilation scheme was applied to the Zhoushan and Jiangsu sea models,and the influence of the assimilation on the model results was analyzed.The 3D-VAR assimilation can effectively improve the correlation between model and radar data in space,and effectively reduce the error between them;the Nudging assimilation avoids the velocity shear caused by the flow field assimilation in space.In addition,the vertical comparison with the observation data of buoys shows that the assimilation also influences the flow field below the sea surface,which gradually decreases with the increase of water depth.Although the model only considers surface current assimilation,it still has a great influence on temperature and salinity.With the continuous assimilation,this effect will gradually expand,and significantly affect the spatial structure of the model's temperature and salinity field,indicating that ocean current data assimilation can affect the simulation of material transport model.