| The analysis,assimilation and effective application of weather radar data and its derived products provide important data support for analyzing and monitoring high impact weather,early warning for severe weather events,and reducing the loss of economy,life and property.Due to many influencing factors and complex evolution process of meso-micro scale severe convective weather system,there are still many key problems in data variational assimilation and doppler velocity wind field retrieval,such as large assimilation error and inaccurate wind retrieval results,which affect the timeliness and accuracy of short-impending prediction,and need to be studied and analyzed by meteorological scientists and technicians more deeply.Therefore,centres in application research of variational problem in weather radar data processing,this paper includes three aspects:construction of computer-friendly forms of the Green's function in variational assimilation,DVAD method in wind field retrival and expression,and the three-dimensional wind field inversion of new high spatial-temporal resolution phased array weather radar.The main contents are as follows:1.Variational method is an important method for data assimilation.In order to solve this kind of inversion method derived from data assimilation,a computer-friendly form of the Green's function is needed to be constructed.The double series expression of Green's function of SKGE equation in rectangular domain is obtained by using the image method.By the eigenfunction expansion method,the Green's function of SKGE equation in rectangular domain is decomposed into logarithmic singular component and regular component in which the logarithmic singular component of Green's function can be clearly expressed as an elementary function by complete summation,and the regular component is expressed as the sum of two infinite series.Convergence analysis shows that the expressions are computer-friendly and uniformly convergent,and can be approximately calculated by direct truncation of series.The polar coordinate expression of Green's function in circular region is constructed,and the expressions of data assimilation and boundary condition retrieval are presented.The computer-friendly Green's function makes it possible to construct an objective selection algorithm of weight coefficients and complete data assimilation of radar and rain gauges.The observation of Changzhou radar and rain gauges in the observation area,are used to realize the optimal selection of weight coefficient and radar rain gauges data assimilation and the improvement effect of data assimilation is evaluated.2.By studying the linear and nonlinear wind field of more than ten kinds of parameter characteristics,the difference between wind field translation speed and VAD mean wind,the difference and linear transformation relationship of linear wind field center,wind vector center and DVAD detection center are analyzed and distinguished,and the feasibility of calculating ambient mean wind and vorticity in r Vd mode is analyzed.The obtained results are as follows:(1)In linear wind field,(_#,_#)obtained by VAD and VAD methods are the average wind vectors above the radar,rather than the actual ambient average wind.The calculated vertical wind profile products are the wind profiles in the calculation area of radar detection area with different heights,and their values change linearly with different observation time and distance between radar and the dynamic center of wind field.(3)In DVAD method,the major merit of representing radar velocity in the r Vd framework is its direct link to the well-known polynomial functions.r Vd field is more suitable for observing the movement of wind field than Vd field.However,the movement of the center of the r Vd quadratic curve is not equal to the actual movement of the linear wind field,but a linear transformation relationship.The transform coefficients depend on the magnitudes and signs of those first-order derivative terms of the wind field.(3)In terms of vorticity observation,by changing'-_),the r Vd difference of a linear wind field,?r Vd is proved to be a set of parallel lines with a slope of_+/_+.The theoretical research results are applied to the analysis of the wind field and precipitation field of the extreme wind and hail weather in Huaian,Jiangsu Province on April 30,2021.It is found that the extreme strong wind was caused by downburst generated by a supercell.There were divergent wind fields near the ground surface.Wind speed in the outflow area in the forward direction of the downburst increased remarkably.Cyclones and anticyclones appeared on two sides of downburst forward direction.All these prove the complexity of nonlinear wind field in severe weather events and the challenge of wind field retrieval.3.The newly developed phased array weather radar data is used for 3D wind field retrieval and analysis.Because of its distributed and highly cooperative scanning mode,high elevation detection range and high spatiotemporal resolution data quality,phased array weather radar has huge advantages in realizing short-range,meso-micro scale system detection,analyzing the fine structure of rapidly changing severe convective weather precipitation system,and 3D wind field retrieval.A strong wind and hail weather process on August 5,2021 detected by phased array radars located in Xiongan is used to study the advantages of phased array radar in detecting,tracking and analyzing meso-micro scale severe convective weather system.(1)Through the analysis of array phased array radar data and the retrieved 3D wind field,the spatio-temporal intensive observation of horizontal and vertical wind fields are obtained,which can advance the analysis of the fine structure and process of meso-micro scale severe convective cell.(2)Qualitative analysis shows that the retrieved wind fields from 3D variational method,two-way integral weight method and direct synthesis method have similar air flow structure characteristics,and are different in some details such as the retrieval data range and air flow scales.(3)Quantitative error analysis of the wind vectors of the three wind field retrieval algorithms is present.The error may come from the ill-posed problems in the mathematical solution of the wind field retrieval,inappropriate assumptions to the wind field,the uncertain Z-Vt relationship,the information distortion of the iterative solutions,etc.The results of quantitative error analysis show that there are still many problems in the wind field retrieval algorithms need to be solved.The research results provide a forefront study for the application of phased array weather radar data in severe weather event early warning.In a word,this paper presents some theoretical and application studies on several key questions in weather radar information processing.Some technical problems in theoretical analysis and practical application are solved,so the relevant methods can be widely used into future operational work. |
PDF Full Text Request