论文部分内容阅读
本文主要介绍3.5维雷达资料同化技术以及增加雷达垂直风场反演信息对该同化技术的影响,利用WRF模式建立了雷达变分同化系统,对同化效果进行检验。3.5维雷达资料同化技术是变分同化技术的扩展,使用多时次雷达观测资料,采用方程约束增加动力学、热力学信息,逐步反演得到风场、动力场和热力场。该技术使用3组连续多普勒雷达体扫资料,分三步获得分析增量以便在背景场基础上更新模式变量。首先利用连续两个时次的格点化三维径向风场来反演三维风场,然后利用连续两个时次的三维风场来依次反演出水平气压场与热力学变量。同时分析空间采用B-样条基底来降低空间的维数,并增加观测信息在空间上的扩散。但该技术也表现出垂直速度强度不够,浮力抬升作用不强的缺陷。针对本缺陷,本文讨论了通过物理初始化方法由雷达回波反演得到的垂直速度,并将该垂直速度作为观测信息同化进入3.5维变分同化中,考察原雷达资料同化方法以及改进方案对随后降水的改善。结果表明,对降水有明显的改善。
This paper mainly introduces the 3.5-dimensional radar data assimilation technology and the effect of increasing the vertical wind field inversion information on the assimilation technology. The radar variational assimilation system is established by WRF mode to test the assimilation effect. 3.5 Vlauda data assimilation technology is an extension of variational assimilation technology. Using time-lapse radar observations, equations and constraints are used to augment the dynamics and thermodynamics information to obtain the wind field, the dynamic field and the thermodynamic field step by step. The technique uses three sets of continuous Doppler radar body scan data to obtain analysis increments in three steps to update the pattern variables based on the background field. Firstly, the three-dimensional wind field is retrieved by using gridded three-dimensional radial wind fields in two consecutive times, and then the horizontal pressure field and thermodynamic variables are sequentially inverted by using two consecutive three-dimensional wind fields. At the same time, the analysis space uses B-spline base to reduce the dimension of space and increase the spatiotemporal diffusion of observation information. However, this technique also shows the drawback that the vertical velocity is not strong enough and the buoyancy lifting effect is not strong. In view of this defect, this paper discusses the vertical velocity obtained from radar echo inversion by physical initialization method and assimilates this vertical velocity into the 3.5-dimensional variational assimilation as observational information, investigates the original radar data assimilation method and the subsequent improvement to the subsequent Precipitation improved. The results show that there is a clear improvement in precipitation.