地形跟随坐标下的中尺度模式气压梯度力计算误差分析及其改进方案

GRAPES_Meso模式预报存在南风偏大、虚假降水偏多等问题，且在大地形下游地区异常明显。平缓-混合坐标可以有效减小气压梯度力计算误差以及平流输送误差，而这两种误差与风场和水汽场预报密切相关。基于GRAPES_Meso模式选择四种平缓-混合坐标对一次典型的高原东部准静止锋降水过程进行模拟分析。模拟结果表明，较弱的天气形势演变下，平缓-混合坐标的改进效果比较明显，可以有效缓解风场预报偏差、虚假降水、虚假天气系统等问题，个例模拟的结果与实况更接近。

On the use of hybrid coordinates in numerical weather prediction models

The GRAPES_Meso model has some forecast problems, such as larger south wind and more false precipitation in the downstream region of the Tibetan plateau. The proper smoothed-level terrain-following coordinate (SLEVE T-F coordinate for short) could efficiently reduce the pressure gradient force error and the mass transport error that are closely related to the wind and water vapor forecasting, therefore, potentially increase prediction performance. Four kinds of T-F coordinates are selected here to simulate a typical quasi-stationary front process. The results of the simulation indicate that under the weak evolution of synoptic background, the improvement of the SLEVE T-F coordinate would be more significant. The SLEVE T-F coordinate can reduce the bias of the wind, false precipitation, false synoptic systems and so on, which results in small calculating error and the better performance of the simulation.