Scholarlink

Using hourly rainfall data from the meteorological big data cloud platform and based on the fuzzy verification neighborhood method, the forecast performance of the China Meteorological Administration's North China Numerical Forecast Model System (CMA-BJ), China Meteorological Administration's Mesoscale Weather Numerical Forecast System (CMA-MESO), and the Ruitu Northeast Numerical Forecast Model System (CMA-DB) during the rainfall process in the main flood season of Liaoning from 2022 to 2023 was evaluated. The results show that without considering the bias within a radius of 40 km, the average hit rate of the model forecast during the Northeast cold vortex rainfall process is below 0.8%, with a maximum of 3.94%. The success rate of the three models’ 12-h forecasts is higher than 24 h, with an average of 14.5%. During the subtropical high rainfall process, the hit rate and TS score of all times of the three models are below 9%, with a maximum of 8.2%. The forecast effect during the typhoon rainstorm process is significantly better than the Northeast cold vortex and subtropical high rainstorm, with an average hit rate of 3.2%, and the false alarm rate of short-term heavy rainfall forecast at the near moment is high. When short-term heavy rainfall is caused by the influence of systems such as the upper trough, the average hit rate of the model forecast is only 1.54%. In all types of rainfall processes, a comprehensive analysis of indicators such as the success rate of the forecast shows that the CMA-MESO model forecast effect is better than the other two models.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluation of High-Resolution Model Heavy Rainfall Forecast Under Different Weather Systems

  • 会议论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年

查看全文
  • Jing Liu,
  • Chuan Ren,
  • Ziqi Zhao,
  • Wei Dong,
  • Yue Wang,
  • Hao Ren,
  • Shuai Zhang

摘要

Using hourly rainfall data from the meteorological big data cloud platform and based on the fuzzy verification neighborhood method, the forecast performance of the China Meteorological Administration's North China Numerical Forecast Model System (CMA-BJ), China Meteorological Administration's Mesoscale Weather Numerical Forecast System (CMA-MESO), and the Ruitu Northeast Numerical Forecast Model System (CMA-DB) during the rainfall process in the main flood season of Liaoning from 2022 to 2023 was evaluated. The results show that without considering the bias within a radius of 40 km, the average hit rate of the model forecast during the Northeast cold vortex rainfall process is below 0.8%, with a maximum of 3.94%. The success rate of the three models’ 12-h forecasts is higher than 24 h, with an average of 14.5%. During the subtropical high rainfall process, the hit rate and TS score of all times of the three models are below 9%, with a maximum of 8.2%. The forecast effect during the typhoon rainstorm process is significantly better than the Northeast cold vortex and subtropical high rainstorm, with an average hit rate of 3.2%, and the false alarm rate of short-term heavy rainfall forecast at the near moment is high. When short-term heavy rainfall is caused by the influence of systems such as the upper trough, the average hit rate of the model forecast is only 1.54%. In all types of rainfall processes, a comprehensive analysis of indicators such as the success rate of the forecast shows that the CMA-MESO model forecast effect is better than the other two models.