Research On Realization Of Four-Dimensional Assimilation Based On Set Optimal Interpolation

Sea surface temperature,as an important physical quantity in ocean data,is a key factor affecting the earth’s air-sea exchange.Its changes will directly affect climate change,the formation of typhoons and changes in underwater temperature.Therefore,the accurate prediction of sea surface temperature has certain scientific significance.However,both ocean numerical models and observation instruments have unavoidable random errors.Therefore,in order to obtain high-precision data products,data assimilation methods are needed for analysis.The law of time evolution has important research significance in four-dimensional assimilation.However,in the current four-dimensional assimilation methods,due to the large amount of computation and difficulty to achieve,in order to reduce the difficulty of the experiment,people often ignore the role of temporal correlation in data assimilation.Therefore,this paper conducts corresponding analysis and research on this problem,discusses the feasibility of adding time correlation to a simple data assimilation method,and designs and conducts a series of preliminary experiments to verify the effectiveness of the method.It has been proved that this method can not only shorten the experimental calculation time,but also fully reflect the time correlation between variables,so as to achieve the purpose of four-dimensional assimilation.On the basis of Bayesian theory,this paper firstly proves the rationality of adding temporal correlation to ensemble optimal interpolation from the aspect of principle.The error analysis of the sea surface temperature data of the POM offshore area ocean model is carried out,and the ensemble optimal interpolation method is used to carry out a control experiment.Then,based on the four-dimensional assimilation method proposed in this paper,the four-dimensional data assimilation experiments of single point and area are carried out to study the applicability of the method.In the ideal experiments based on the algorithm design of ensemble-optimal interpolation,experiments using assimilation have significantly improved the effect of the background field.In the single-point experiment based on the new ideas put forward in this paper,the time correlation is added to the assimilation process.The experiment can control the time evolution trend to improve the assimilation quality and make the deviation between the analysis data and the real value smaller.Different assimilation points and the choice of time frequency will also affect the final result.In the final regional assimilation experiment,using the four-dimensional assimilation method to assimilate in different sea areas will produce different performance effects.The assimilation effect will be better when the time interval value is consistent with the fluctuation frequency of the spectrum.If it is inconsistent,it will bring negative effects to the experiment,the specific reason is that in the frequency domain of the mode,the energy of the time signal is more divergent.