墨脱的降水变化与印度洋、 孟加拉湾的水汽向内陆的输送、 夏季风及我国东部雨带的推进过程联系紧密。它的独特地形与西南气流的相互作用在当地形成了年均降水超过2000 mm的降水带。但墨脱的复杂地形与有限的电力交通条件造成了降水观测的困难。2019年， 第二次青藏高原综合科学考察研究任务专题项目在墨脱设置了一部X波段双偏振相控阵雷达， 实现了对墨脱云降水的连续观测， 对当地生态环境和云水资源研究以及高原东南水汽通道特征研究等具有重要的意义。文中使用墨脱X波段双偏振相控阵雷达2020年6 -8月的观测数据， 使用统计手段筛选降水回波片段， 整合后构造得到符合墨脱复杂地形的混合仰角； 利用2019年墨脱雨滴谱数据计算雷达参量、 拟合得到适合墨脱地区降水特点的X波段天气雷达定量降水估测（QPE）公式。在此基础上选取2020年7 -8月三次累积降水量、 平均雨强、 持续时间各不相同的降水过程， 使用线性规划方案计算Φ _DP（差分传播相移）、 最小二乘拟合计算K _DP（差分传播相移率）、 “ZPHI”降水廓线算法订正Z _H（反射率因子）； 以Z _H和K _DP为阈值， 使用R（Z _H）与R（K _DP）分段估测降水， 与单独使用R（Z _H）和R（K _DP）的估测结果进行对比； 并基于降水估测结果和雷达参量的水平分布探讨降水分布与地形的关系。本文经过质量控制的Z _H、 K _DP数据质量有明显改善； 使用统计方法构造的混合仰角有效回波面积大于使用SRTM1 v3.0地形数据构造的混合仰角， 三次降水的典型时刻参量与过程累积降水量和平均Z _H、 Z _DR能初步反映降水变化与地形的关系； 本文的降水估测公式和分段估测降水方法整体结果表现良好， 各评估参数均具有较好的表现； 结果表明： （1）墨脱雷达周边地形变化剧烈， 根据现有地形数据构造的混合仰角在降水估测中表现不如本文统计筛选雷达观测数据得到的混合仰角数据； （2）本文采用的数据质控方法、 定量降水估测公式与分段估测方法（PEM）在降水反演中表现良好， 与单一的R（Z _H）、 R（K _DP）方法相比， 得到的结果在CC和RMSE无明显变化的情况下， 估测误差明显减小； （3）墨脱降水的发生发展或与西南气流被河谷偏北方山坡抬升有关， 较大雨强、 较小粒子直径的云位于地势平缓的谷底。

The variation in precipitation in Motuo is closely associated with the transport of water vapor from the India Ocean and the Bay of Bengal to inland China， the Asia summer monsoon， and the progression of the rain band in eastern China.A rain band with an average annual precipitation of more than 2000 mm has been formed due to the interaction of the distinctive topography of Motuo and the southwest air flow.However， it has been challenging to observe precipitation in Motuo due to the peculiar topography， insufficient electricity， and inadequate traffic conditions.An X-band dual-polarization phased array radar （XPAR） is installed during the Second Qinghai-Xizang Plateau Scientific Expedition and Research Program.The advanced dual-polarization phased array radar is used to continuously observe the precipitation in Motuo and is of great significance to the research of the local ecological environment and cloud water resources， and the impact on downstream.Based on XPAR data collected for the period from June to August 2020， we selected the radial precipitation echoes using a statistical method， and after integrating the precipitation echoes， we created the hybrid elevation angle that tallied with the topography of Motuo.Using DSD data from Motuo in 2019， we calculated radar parameters and obtained the localized QPE （Quantitative Precipitation Estimation） formulas of X-band weather radar.We selected three processes of precipitation with different cumulative precipitation， duration， and average rain rate from July to August 2020， calculated Φ _DP （Differential Propagation Phase Shift） using the linear programming method， K _DP （Differential Propagation Phase Shift Rate） using the ordinary least square method， and corrected Z _H （Reflectivity Factor） using the “ZPHI” rain profiling algorithm.Using Z _H and K _DP as thresholds， we estimated precipitation piecewise utilizing R（Z _H） and R（K _DP）， and contrasted results with the QPE results from the R（Z _H） and R（K _DP） methods， respectively.And explored the relationship between precipitation distribution and the topography of Motuo based on the QPE results.The quality of Z _H and K _DP had been significantly improved.The hybrid elevation angle constructed using a statistical method had a larger echo area than the hybrid elevation angle constructed using STRM1 v3.0 data.The PPI （Plane Position Indicator） diagrams of three precipitation （radar parameters at typical times， and average radar parameters） processes can show the relationship between the variation in precipitation and the topography of Motuo.The QPE formulas and Piecewise Estimation Method （PEM） used in the research functioned well， and each evaluation parameter performed well.The following conclusions are drawn： （1） The hybrid elevation angle constructed using the statistical method outperformed the hybrid elevation angle constructed using topography data in QPE， due to the drastic variation of the topography around the XPAR in Motuo； （2） The data quality control method， the QPE formulas， and the PEM utilized in the research all performed well at precipitation retrieval.The estimation error of the QPE results was significantly reduced， without causing significant changes in CC and RMSE； （3） Precipitation in Motuo may occur and grow as a result of the lifting of the southwest air flow by northern slopes of the valley.Clouds with higher rain rates and smaller drops are found near the level bottom of the valley.