On 10 October 2018 an intense storm, characterized by heavy rainfall, hit the Sardinia island, reaching a peak of 452 mm of rain measured in 24 h. Among others, two particularly intense phases were registered between 3 and 6 UTC (Universal Coordinated Time), and between 18 and 24 UTC. The forecast of this case study is challenging because the precipitation was heavy and localized. In particular, the meteorological model used in this paper, provides a good prediction only for the second period over the eastern part of the Sardinia island. In this work, we study the impact of lightning data assimilation and horizontal grid resolution on the Very Short-term Forecast (VSF, 3 and 1 h) for this challenging case, using the meteorological model. The comparison between the 3 h VSF control run and the simulations with lightning data assimilation shows the considerable improvement given by lightning data assimilation, especially for the precipitation that occurred in the eastern part of the island. Reducing the VSF range to 1 h, resulted in higher model performance with a good precipitation prediction over eastern and south-central Sardinia. In addition, the comparison between simulated and observed reflectivity shows an important improvement of simulations with lightning data assimilation compared to the control forecast. However, simulations assimilating lightning overestimated the precipitation in the last part of the day. The increasing of the horizontal resolution to 2 km grid spacing reduces the false alarms and improves the model performance.

中文翻译：

---

闪电数据同化在撒丁岛2018年10月10日案例研究中的应用

2018年10月10日，一场以暴雨为特征的强烈风暴袭击了撒丁岛，在24小时内达到452毫米的降雨峰值。其中，在3到6 UTC（通用协调时间）之间以及18到24 UTC之间记录了两个特别强烈的阶段。由于降水量大且局部化，因此本案例研究的预测具有挑战性。特别是，本文中使用的气象模型仅对撒丁岛岛东部的第二个时期提供了良好的预测。在这项工作中，我们使用气象模型研究了闪电数据同化和水平网格分辨率对这种极具挑战性的情况下的极短期预报（VSF，3和1 h）的影响。3 h VSF控制运行与雷电数据同化的模拟之间的比较表明，雷电数据同化可带来相当大的改善，尤其是对于岛屿东部发生的降水。将VSF范围减小到1 h，可实现较高的模型性能，并在撒丁岛中东部和南部具有良好的降水预测。此外，与控制预测相比，在模拟和观察到的反射率之间的比较显示，在闪电数据同化的情况下，模拟的重要改进。但是，模拟闪电会高估一天的最后一天的降水量。将水平分辨率提高到2 km网格间距可以减少误报并提高模型性能。