In this study, we assess the validity of the Weather Research and Forecasting (WRF) numerical model for estimating precipitable water vapor (PWV) at the Ali observatory, located in the southwestern part of the Tibetan Plateau. We have run WRF in three nested domains with different horizontal resolutions, centered at the Ali observatory, and the ability of the WRF model on estimating PWV conditions at an astronomical observatory on the Tibetan Plateau with complex terrain has been discussed. In the validation, we make use of a verification sample of 1 yr of radiosonde data set obtained directly at the Ali radiosonde station in 2016, which is about 25 km north of the observatory. The results from model with the highest horizontal resolution of 1 km and the temporal resolution of 0.5 hr in domain 03 are presented in this article. On the Tibetan Plateau, it should be noted that the WRF model slightly overestimates the PWV, while there are high correlation coefficients between the PWV from model and that from observation. In summary, the WRF model shows an excellent performance in estimating PWV at the observatory, and a good agreement between model simulation and radiosonde observation has been found, including for the extremely low value of the PWV (≤1 mm), thus allowing it to complement the PWV estimation above the Tibetan Plateau.

在这项研究中，我们评估了位于青藏高原西南部阿里天文台的天气研究和预报（WRF）数值模型用于估算可降水量水汽（PWV）的有效性。我们已经在以阿里观测站为中心的三个水平分辨率不同的嵌套域中运行了WRF，并讨论了WRF模型在具有复杂地形的青藏高原天文观测站估算PWV条件的能力。在验证中，我们使用了2016年直接在阿里无线电探空仪站获得的1年无线电探空仪数据集的验证样本，该站位于天文台以北约25公里处。本文介绍了在域03中具有1 km的最高水平分辨率和0.5 hr的时间分辨率的模型的结果。在青藏高原上，应该注意的是，WRF模型略微高估了PWV，而来自模型的PWV与来自观测的PWV之间具有较高的相关系数。总而言之，WRF模型在估算天文台的PWV方面表现出出色的性能，并且在模型仿真和探空仪观测之间已经发现了很好的一致性，包括PWV的极低值（≤1mm），从而使其能够补充青藏高原上方的PWV估算。