Understanding the mesoscale spatial and temporal variability of precipitation (of the order of 1 to 100 km, and of minutes to hours, respectively) is essential for larger-scale studies, especially in highly heterogeneous areas, such as coastal regions, mountains and urban areas. The adequate capturing of the precipitation patterns in both diagnostic and prognostic terms remains an open challenge. On the one hand, commonly used ground measuring devices, such as rain gauges, disdrometers and other point instruments, exhibit an ineffective sampling for the whole range of scales and areas covered with precipitation. On the other hand, weather radars provide a much greater spatial and temporal resolution which, however, must be exploited with care in rain rate estimations, due to complicated and non-unique transfer functions used in the processing procedures. This study aims at bridging, to at least some extent, this scale gap and to demonstrate the role of fine-scale precipitation estimates derived from weather radar data, and to illustrate their impacts on larger scale weather patterns. The focus is on the precipitable water content and its changes during a Mediterranean cyclone evolution.

In this paper, the first results with the regional radar signal processing chain that provides the radar data assimilation (RDA) in the Harmonie convection permitting numerical model are described. The numerical experiments exhibit effective simulation of the precipitable water at different stages of the frontal depression which affected Cyprus in the Eastern Mediterranean, during 31 December 2018–2 January 2019. In particular, it was found that the major volume of precipitable water was related to the cyclone frontal zone (which is in turn related to heavier precipitation), followed by a weaker field of precipitable water behind the frontal zone (and weaker precipitation). During the whole severe weather period, the precipitable water minimum was maintained over the island's central mountain range, reflecting the importance of interactions and feedbacks between the air flow and orography.

The radar data assimilation influences the results of the simulations, depending on the particular area and cyclone stage. The largest innovations due to RDA occur with the approach of the cyclone's frontal zone. Enhancement of precipitable water due to RDA is observed over most of the whole area, but this is more pronounced in weaker precipitation locations. The impact is associated with a shift of the precipitable water maximum towards higher values in the precipitable water distribution, which is accompanied by enlarged sizes of areas with greater precipitable water. During the cyclone's mature stage and at the rear zone, the integrated RDA impact over the whole domain is almost negligible but takes the form of mesoscale cells of opposite signs; at this stage, radar measurements yield an alteration of both the positions and rates of spatial precipitable water fields, which result in intermittent mesoscale changes.

了解降水的中尺度时空变化（分别为1至100 km的数量级，以及几分钟至几小时的数量级）对于大规模研究至关重要，尤其是在高度异质性地区，例如沿海地区，山区和城市地区。在诊断和预后方面充分捕捉降水模式仍然是一个公开挑战。一方面，常用的地面测量设备（例如雨量计，测速计和其他点式仪器）无法在整个范围的降雨和覆盖降水的区域进行有效采样。另一方面，天气雷达提供了更高的空间和时间分辨率，但是在估算雨率时必须谨慎使用，由于处理过程中使用了复杂且非唯一的传递函数。这项研究旨在至少在一定程度上弥合这一尺度鸿沟，并证明从天气雷达数据中得出的小尺度降水估计的作用，并说明它们对更大尺度天气模式的影响。重点是在地中海气旋演变过程中可沉淀的水含量及其变化。

在本文中，描述了在Harmonie对流允许数值模型中提供雷达数据同化（RDA）的区域雷达信号处理链的第一个结果。数值实验显示了在2018年12月31日至2019年1月2日期间影响地中海东部塞浦路斯的额陷的不同阶段的可降水量的有效模拟。特别是，发现主要的可降水量与旋风锋区（这又与更大的降水有关），然后是锋区后面的较弱的可沉淀水域（和较弱的降水）。在整个恶劣的天气期间，该岛中央山脉的最低可降水量保持不变，

雷达数据的同化影响模拟结果，具体取决于特定区域和旋风分离器的阶段。RDA带来的最大创新是旋风分离器的额叶区的接近。在整个区域的大部分地区都观察到由于RDA引起的可沉淀水的增加，但是在降水量较弱的地区更为明显。该影响与可沉淀水最大值向可沉淀水分布中的较高值的移动相关，这伴随着具有更大可沉淀水的区域的扩大。在旋风的成熟阶段和后部区域，RDA对整个区域的综合影响几乎可以忽略不计，但采取相反符号的中尺度细胞的形式。在这个阶段，