Abstract

The study presents the new system for operational short-range numerical weather prediction with the grid spacing of 1 km for the Moscow region that considers the features of urbanized surface and is based on the COSMO-Ru1M model configuration. This system is implemented in the Hydrometcenter of Russia. The results of the trial testing of the optimum model configuration for the Moscow region using observations from the dense network of weather stations and MTP-5 temperature profilers are presented. High prediction capabilities of the new forecasting system are demonstrated. The approaches to the minimization of the time of calculations for the technology chain implemented at the Roshydromet supercomputer are described. A case study of modeling with the grid spacing of 500 m versus 1000 m for summer convective weather events in the Moscow region has been analyzed.

中文翻译：

---

开发用于莫斯科地区天气和恶劣天气事件数值预报的高分辨率业务系统

摘要

该研究提出了一种新的短程数值天气预报业务系统，该系统考虑了城市化表面的特征，并基于COSMO-Ru1M模型配置，对莫斯科地区的网格间距为1 km。该系统在俄罗斯的Hydrometcenter中实施。呈现了使用来自气象站和MTP-5温度剖面仪的密集网络的观测结果对莫斯科地区的最佳模型配置进行的试验测试结果。展示了新预测系统的高预测能力。描述了在Roshydromet超级计算机上实现的技术链计算时间最小化的方法。