雷达径向风观测在华北区域数值预报系统中的实时三维变分同化应用试验

以实现业务应用为目标开展了区域多部雷达径向风观测资料的三维变分直接同化应用研究。重点对背景场误差协方差的方差和尺度因子进行调整,形成能够与其他常规观测资料协同同化的雷达径向风同化方案,并建立了京津冀6部多普勒雷达观测资料的实时预处理系统。基于上述工作开展2011年汛期京津冀多普勒雷达径向风观测资料在华北区域快速更新循环同化和预报系统中的实时同化和对比试验,并对应用效果进行了初步评估。实时同化试验期间京津冀地区6部雷达经过质量控制后的径向风数据质量和同化情况的分析结果表明,同化系统有效地吸收了雷达径向风的观测信息并形成合理的分析增量,其中,S波段雷达观测的径向风数据数量、质量和稳定度均明显优于C波段雷达;整体来看,雷达径向风同化对地面和高空要素预报性能的影响基本为中性,且主要影响时段集中在最初的6 h。但降水预报评分结果表明,雷达径向风同化从降水强度、落区和范围等方面均明显提升了系统对对流尺度降水的短时预报性能。同时也应该看到,受制于目前3 h一次的同化更新频率,雷达资料同化的效果往往到对流临近时次才能体现。

The real-time radar radial velocity 3DVar assimilation experiments for application to an operational forecast model in North China

In recent years, research efforts have been carried out to develop the radar data pre-processing system and assimilating strategy for the purpose of future operational implementation for BJ-RUCv2.0. There are 4 S-band and 2 C-band Doppler radars involved in this study, comprising the operation radar network in North China. A radar data preprocessing system was built on the base of the counterpart of VDRAS with the functions to perform quality control, mapping and error statistics. With techniques of tuning the background error covariance (BE) and length scales developed, a two-step scheme was designed to assimilate radial velocity along with other conventional observations. The pre-operational experiments of BJ-RUCv2.0 with assimilation of radial velocity were performed during the flooding season from 1 June to 31 August 2011. The results were evaluated and compared along with another identical experiment running in parallel but without assimilation of radial velocity. Monitoring on quality of the radial velocity data entering the data assimilation (DA) system revealed that the DA system effectively absorbed the information brought by the radial velocity and generated reasonable analysis increments. But obviously, the valid amount, data quality and reliability of radial velocity observations from S-band radars were much better than C-band radars. The evaluations for cycling forecasts for both experiments were performed against conventional surface and radiosonde observations. In general, from the 3-month results, no significant improvements were indeed found against conventional observations because of the localized impact of radial velocities in space and time. But more positive signals were shown by precipitation forecast skills, especially for short-term forecasts and the precipitations with larger intensities.The cases of a squall line and a local-forced convection were selected for more detailed discussion respectively. It was found that the more the assimilation time of radar radial velocity approached the occurrence time of convection, the more significantly the precipitation forecast skill was improved. For the well-organized convection such as squall line, the modification of low-level dynamical conditions brought by the assimilation of radial velocity was the key factor of the improvement of forecast. For the local-forced convections, the more favorable thermal circumstances were also built along with the modification of dynamical conditions. But it is still needed to be noted that, it is usually at the approaching time of convection that the effect of radial velocity assimilation gets to be more prominent.