中文摘要：

对于水平网格距小于10 km的高分辨率非静力平衡的显式云分辨模式,云微物理量的初始化以及物理量之间的相互协调是十分重要的,并且一直是云分析领域的一个难题。考虑在云雨处于定常状态的前提下,根据暖云过程,可得到云中水成物之间相互转化以及与垂直速度的约束关系,而雨水含量跟雷达反射率因子（Z）有关。因此,采用合肥新一代天气雷达2003年7月5日02时（北京时）的观测资料,针对雷达探测的特点对雷达原始数据进行了坐标转换,并进行了插值处理和简单的质量控制,然后依据Z-qr关系和定常暖云方案,反演了雨水混合比（qr）、云水混合比（qc）、水汽混合比（qv）和垂直速度（w）。结果表明,由此得到的雷达回波主要特征及量值分布与雷达站的CAPPI图像基本一致;水汽、云水、雨水和垂直速度量值大小的分布与雷达回波强度的分布也是吻合的,体现出了云中水成物和垂直速度的三维分布结构;各物理量的量值也符合梅雨锋暴雨的特点,梅雨锋积层混合云系中层状云和对流云的差别十分明显。在较强的回波区,雨水在6 km以下含量较大,最大值位于4 km左右,可超过3.0 g/kg;上升速度在5 km左右最大,最大值超过5 m/s;云水含量大值区位于5 km以上,在7 km高度上达到最大,超过3.0 g/kg;雨滴末速度虽然上下比较一致,一般几米每秒,但在5 km左右为大值区。

英文摘要：

For non-hydrostatic cloud resolving models with a grid spacing less than 10 km, the incorporation of microphysical elements and the coherence of physical variables in the initial field of numerical weather prediction are very important, however it has remained a difficult problem in the cloud analysis field for a long time. It is known that warm cloud microphysical processes link the transform of hydrometeors with vertical motion and the radar echo intensity has close relation with hydrometeors. So, it is possible to acquire microphysical variables in clouds from radar echo under the condition that only acceptable precision losses are paid. Under the conditions of warm cloud microphysical processes and hypothesizing the cloud in a stationary state at a given time, the water vapor mixing ratio（qv）, cloud water mixing ratio（qc） and vertical velocity（w） can be obtained through rain water mixing ratio（qr）, after the qr has been retrieved from the radar reflectivity factor（Z） through the Z-qr relation. A retrieval technique has been developed in this paper following the above thinking clue, and to illustrate its performance several retrieval experiments were performed using the Doppler radar observations of Hefei NEXRAD, Anhui province at 02：00 BST 5 July 2003. After some preprocessings such as the simple quality control, coordinate conversion, interpolation, and spatial smooth of the raw radar volume-scanning data, q r, q v, qc, and w etc were retrieved. Results show that the CAPPI image at the level of 4 km obtained using the coordinate transform method is just the same as the output using radar station＇s technology. The patterns of mixing ratios of water vapor, cloud water, and rain water, and vertical velocity retrieved from the radar reflectivity ac- cord well with those of radar echo intensity, and reflect satisfactorily the three-dimensional spatial structure of hydrometeors and vertical motion in the convective cloud. Echo characteristics of heavy rain on the meiyu front is very obvious