针对WRF模式中行星边界层参数化过程倾向项的扰动方法  

Stochastic perturbation on planetary boundary layer parameterization tendencies in the WRF model

作　　者：武天杰[1] 闵锦忠[1] WU Tianjie;MIN Jinzhong(Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters,Nanjing University of Information Science&Technology,Nanjing 210044,China)

机构地区：[1]南京信息工程大学气象灾害教育部重点实验室/气象灾害预报预警与评估协同创新中心,江苏南京210044

出　　处：《大气科学学报》2021年第2期240-251,共12页Transactions of Atmospheric Sciences

基　　金：国家重点研发计划资助项目(2017YFC1502100)。

摘　　要：针对WRF模式中行星边界层参数化过程中的不确定性,发展了一种针对行星边界层参数化过程的随机物理扰动方案(SPPBLPT),该方案针对行星边界层计算的温度、风场、水汽倾向项进行扰动。使用该方案、多行星边界层参数化方案、多参数扰动方案及针对WRF模式总倾向的随机物理过程扰动(SPPT)方案对2014年7月进行对比实验,发现使用较大格点方差的SPPBLPT方案能有效降低地面温度与风场的误差,也能降低降水的预报误差,而其他方案对预报改善不明显。针对地面温度和风场的BS评分显示,SPPBLPT方案通过降低可靠性评分(提高可靠性),显著改善了集合预报对温度与风场的可能性预报,同时该方案能显著提高降水的GSS评分,在所有实验中,较大格点方差的SPPBLPT方案表现最好。针对行星边界层参数化过程的随机物理扰动方案(SPPBLPT)能显著提升集合预报系统性能,但是该方案的扰动参数的设置还需要进一步研究。In this study,targeted at the uncertainty in the parameterization process of the planetary boundary layer in the WRF model,a Stochastic Perturbation on Planetary Boundary Layer Parameterization Tendency scheme(SPPBLPT)was developed.This scheme perturbs the temperature,wind and water vapor tendencies.A comparative experiment was conducted in July 2014 using the SPPBLPT scheme,multi-planet boundary layer parameterization scheme,multi-parameter perturbation scheme,and Stochastic Physical Process Perturbation(SPPT)scheme.It was observed that the SPPBLPT scheme with a large grid-point variance can effectively reduce the error of 2 m temperature and wind field,and it can also reduce precipitation error,while other schemes do not improve the forecast significantly.The BS scores for the 2 m temperature and wind field show that the SPPBLPT scheme significantly improves the possibility of ensemble forecasting for temperature and wind field,by reducing the reliability score,thereby increasing performance.At the same time,the scheme can significantly improve the GSS score of the precipitation.In all experiments,the SPPBLPT scheme with the larger grid-point variance had a superior performance.Although the SPPBLPT scheme reduced the FSS score,the reduction was not significant.It is thus shown that the SPPBLPT scheme can significantly improve the performance of ensemble forecasting systems,yet the setting of perturbation parameters for this scheme requires further investigation.

关 键 词：行星边界层 随机物理过程扰动 集合预报 WRF模式 

分 类 号：P456.7[天文地球—大气科学及气象学]