零侧边界通量方案在CMA-MESO模式中的应用

Application of Zero Lateral Flux Scheme in the CMA-MESO model

关键词:  零侧边界通量  有限区域守恒半拉格朗日方案  水物质守恒  极端暴雨  中尺度模式

Key words:  Zero Lateral Flux  conserving semi-Lagrangian-based limited-area models  conservation of water mass  extreme rainstorms  mesoscale model

摘要:

在全球或区域数值模式中实现水汽等水物质的守恒或收支平衡计算十分重要，缺乏质量守恒可能会导致虚假的水汽运动和过量的局部降水。目前我国的 CMA-MESO 模式中使用的平流方案是 PRM（Piece-wise Rational Method）方案，尽管该方案有较高的精度和正定保形性以及能够在全球模式中做到守恒，但是在有限区域模式中由于侧边界的处理难以做到在有限区域模式中的守恒或收支平衡。为了解决模式平流方案在有限区域模式中的守恒问题，研究了一种新的简单且可忽略计算成本的有限区域模式半拉格朗日方案质量守恒的零侧边界通量方案（Zero Lateral Flux, ZLF）将其应用在 CMA-MESO模式中。研究先通过理想试验结果表明 ZLF 方案具有良好的守恒性和保形性，能够更好保持物理量场的分布和强间断物理量场的守恒。然后将该方案加入 CMA-MESO 模式中，通过实际个例预报试验和连续预报试验结果表明 ZLF 方案能够抑制高估的降水，减少虚假降水预报，显著改善降水落区预报。对于极端暴雨而言，ZLF 方案对于降水量级和降水落区预报改善效果都非常显著。ZLF方案有效改进了 CMA-MESO模式的水物质不守恒问题，提高了模式的降水预报效果。

Abstract:

It is important to achieve conservation or break-even calculations of water substances such as water vapour in global or regional numerical models where the lack of mass conservation may lead to spurious water vapour movements and excessive local precipitation. At present, the advection scheme used in CMA-MESO model is the PRM (Piece-wise Rational Method) scheme. Although this scheme has high accuracy and positive, definite shape retention and is able to achieve conservation in the global model, it is difficult to achieve conservation or break-even in the limited-area model due to the treatment of lateral boundaries in the limited-area model. In order to solve the conservation problem of the model advection scheme in a limited-area model, a new and simple limited-area model semi-Lagrangian scheme mass conservation method (Zero Lateral Flux, ZLF) with negligible computational cost has been developed and applied in the CMA-MESO model. In this study, the ideal experimental results show that the ZLF scheme has good conservation and conformality, and can better maintain the distribution of the physical quantity field and the conservation of strong discontinuous physical quantity fields. Then the scheme is added to the CMA-MESO model, and the results of actual case forecast experiments and continuous forecast experiments show that the ZLF scheme can suppress overestimated precipitation, reduce false precipitation forecasts, and significantly improve precipitation-falling area forecasts. For extreme rainstorms, the ZLF scheme has a very significant improvement effect on the forecast of precipitation magnitude and precipitation area. The ZLF scheme improves the water non-conservation problem of the CMA-MESO model (effectively) as well as the precipitation forecasting effect of the model.