对数值模式物理过程与动力框架耦合方案进行了分析, 在GRAPES模式中引进了ECMWF模式中的Wedi (1999) 耦合方案, 此方案考虑了两个时间层物理过程的影响, 即将上游点和到达点物理过程的影响同时反馈到动力框架中, 它是二阶时间精度的耦合方案, 比原一阶精度耦合方案更加合理和准确。利用GRAPES模式, 设计了Wedi耦合方案与原耦合方案的连续数值试验。试验结果表明, 在逐日降水预报中, Wedi耦合方案与原耦合方案降水预报的T评分 (Threat score) 互有高低, 但差别不大; 预报的降水量可能有较小的变化, 但不能使预报降水的等级出现本质的改变。但对其平均后的分析可知, 新引进的Wedi耦合方案对预报降水评分具有正的贡献。即新耦合方案完善和改进了GRAPES模式, 对提高模式的降水预报准确率具有正效果。
The paper investigates current coupling scheme of physics and dynamics of the numerical prediction model. The Wedi (1999) coupling scheme of physics and dynamics is introduced into the Global/Regional Assimilation and Prediction System (GRAPES). The impact of two-time-level physics is implemented in the Wedi scheme, i.e. the physics contribution at the departure point together with the arrival point give feedback to the dynamic core of the model. So the Wedi scheme is the second-order accurate coupling scheme and is more reasonable and precise than the old one-time-level scheme in the GRAPES in theory. The consecutive numerical experiments are performed by using of the Wedi and old coupling scheme in the GRAPES model. Results show that the Threat scores of precipitation forecasting with the Wedi and old coupling scheme is tinily different, the forecasted precipitation may have a little variation but it cannot change the magnitude. Based on the analysis of the averages, it is found that the Wedi scheme has a positive impact on precipitation forecasting skill score, that is to say, the Wedi coupling scheme improves the GRAPES model and has a positive effect on enhancing the precision of precipitation forecasting.