大气科学  2018, Vol. 42 Issue (6): 1327-1343 PDF

1 中国气象科学研究院灾害天气国家重点实验室, 北京 100081
2 中国气象科学研究院云雾物理环境重点实验室, 北京 100081
3 南京信息工程大学气象灾害预报预警预评估协同创新中心, 南京 210044

Cloud Microphysics and Regional Water Budget of a Summer Precipitation Process at Naqu over the Tibetan Plateau
TANG Jie1,2, GUO Xueliang1,2,3, CHANG Yi1,2
1 State Key Laboratory of Severe Weather(LaSW), Chinese Academy of Meteorological Sciences, Beijing 100081
2 Key Laboratory for Cloud Physics, Chinese Academy of Meteorological Sciences, Beijing 100081
3 Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing 210044
Abstract: Intensive observational studies on clouds and precipitation have been conducted in the project of the Third Tibetan Plateau Atmospheric Scientific Experiment. In order to further reveal cloud microphysical structure, water transformation in clouds and regional water budget properties over the plateau, a relatively typical convective precipitation process on July 5-6, 2014 in the Naqu region is investigated using mesoscale numerical prediction model (WRF) combined with observational data collected during the experiment. The results indicate that WRF model could reproduce the general characteristics of cloud process and diurnal variation of precipitation over the plateau. The modeling results show that there were large amounts of supercooled cloud water and graupel particles in the summer convective clouds over the plateau. The ice process played a critical role in the development of clouds and precipitation over the plateau. The surface precipitation mainly formed by the melting of graupel particles. Although the warm cloud microphysical process had small direct contribution to the formation of surface precipitation, it had an important contribution to the formation of graupel embryos. The accretion transformation process of ice crystal with supercooled raindrops contributed to the production of graupel particles and its growth mainly relied on the riming process for supercooled cloud water. The net water vapor budget was positive at Naqu over the plateau and the mean daily conversion rate from vapor to precipitation was as high as 20.75%, which is close to that in the downstream of the Yangtze river and higher than that in northern and northwestern China. The contribution of daily mean surface evaporation to precipitation, i.e., precipitation recycling, was 10.92%, indicating that 90% of the rainfall was from the conversion of water vapor outside, although local evaporation of water vapor had a certain contribution to the water vapor source of the rainfall over the plateau.
Keywords: Tibetan Plateau    Numerical simulation    Cloud microphysics    Water transformation and budget    Precipitation recycling
1 引言

2 模式设置与观测数据处理

 图 1 三层嵌套模拟区域（彩色阴影：海拔高度；黑色曲线：青藏高原的边界线；“+”：那曲站位置） Figure 1 Triple-nested model domain (color shadings indicate the terrain height, black curves indicate the boundary of the Tibetan Plateau, and symbol + indicates the location of Naqu station)

3 模拟结果检验 3.1 降水过程简介

2014年7月5日12:00至6日12:00那曲地区出现了较强的降水过程，那曲站点24 h累积降水量为21.4 mm，达中雨量级。5日18时200 hPa、500 hPa的高度场和风场显示，18:00青藏高原对流层上层（200 hPa）受南亚高压控制，低层（500 hPa）受弱脊控制，说明高原上空并没有明显的大尺度天气系统。因此，5~6日高原那曲地区的对流降水过程主要与高原特有的热力作用相关（Zhu and Chen, 2003），属于局地热对流过程发展和演变产生的降水过程。

 图 2 FY-2E卫星的TBB随时间演变情况：（a–c）分别为5日12:00（当地时间，协调世界时UTC+6 h，下同）、18:00、6日00:00TBB分布（“+”：那曲站位置） Figure 2 Time series of TBB derived from FY-2E satellite infrared cloud image: (a), (b) and (c) are at 1200 LT (Local time, UTC+6 h), 1800 LT 5 July and 0000 LT 6 July, respectively (symbol + indicates the location of Naqu station)

 图 4 2014年7月5日12:00至6日12:00（a）模拟与观测多站点平均的逐小时降水率对比，（b）模拟与观测d04区域总降水量的对比，（c）第三层嵌套模拟区域24 h累积降水量（彩色阴影）与12个站点观测的24 h累积降水量（白圆圈表示站点位置，数字代表其累积降水） Figure 4 (a) The mean hourly precipitation rates observed and simulated at 12 stations, (b) total rainfall observed and simulated over d04 area, (c) 24-hour accumulated rainfall simulated over the third nested simulation domain (color shadings) and observed at 12 stations (white circles indicate the locations of stations and the numbers indicate their 24-hour accumulated rainfall) from 1200 LT 5 July 2014 to 1200 LT 6 July 2014
3.2 云和降水过程的验证

 图 3 2014年7月5日12:00至6日12:00（a）C波段连续波雷达回波、（b）WRF模拟的d04区域强回波（＞20 dBZ）频数、（c）WRF模拟的d04区域最大上升速度的高度—时间分布情况、（d）7月5日12:00 WRF模拟的组合反射率 Figure 3 Time–height distributions of (a) C-band continuous wave radar echo, (b) strong echo frequency (> 20 dBZ), (c) horizontal maximum updrafts in the d04 area simulated by WRF model from 1200 LT 5 July 2014 to 1200 LT 6 July 2014, and (d) composite reflectivity at 1200 LT 5 July 2014 simulated by WRF

 图 5 2014年7月6日04:00那曲站水平组合反射率和垂直反射率分布：（a，c）C波段多普勒雷达观测；（b，d）WRF模拟。图a和b中黑色实线AB为垂直截面位置。（e）两个时刻（5日14:00、6日06:00）回波强度平均廓线的对比（实线为C波段多普勒雷达，虚线为WRF模拟） Figure 5 Horizontal distributions of composite reflectivity and corresponding vertical distributions of reflectivity at 0400 LT 6 July 2014: (a, c) C-band Doppler radar observations; (b, d) WRF simulation. The black lines AB in Figs. a and b stand for the cross-section positions. (e) Averaged profiles of echo intensity at 1400 LT 5 July and at 0600 LT 6 July (solid lines stand for C-band Doppler radar, dotted lines stand for WRF simulation)

4 云微物理结构及水分转化特征

 图 6 2014年7月5日12:00至6日12:00 WRF模拟的d04区域水平平均的云中粒子比含水量高度—时间分布：（a）云水qc；（b）冰晶qi；（c）雨水qr；（d）雪qs；（e）霰qg。图中虚线为温度（单位：℃） Figure 6 Time–height distributions of horizontally averaged hydrometeors mixing ratios in the d04 area simulated by WRF model from 1200 LT 5 July 2014 to 1200 LT 6 July 2014: (a) Cloud water qc; (b) cloud ice qi; (c) rain water qr; (d) snow qs; (e) graupel qg. Horizontal dashed lines are temperature (units: ℃)

 图 7 2014年7月5日14:00 B1点（31.43°N，92.26°E）（a）环境温度、（b）垂直速度、（c）水凝物混合比含水量的垂直廓线以及（d）雨水源项、（e）雪源项、（f）霰初生源项、（g）霰增长源项的微物理过程转化率的垂直廓线 Figure 7 Vertical profiles of (a) environmental temperature, (b) vertical velocity, (c) hydrometeors mixing ratio as well as microphysics conversion rate of (d) rain water source terms, (e) snow source terms, (f) graupel formation source terms, (g) graupel growth source terms at location B1(31.43°N, 92.26°E) at 1400 LT 5 July 2014

 图 8 2014年7月5日12:00至6日12:00 d04区域降水性粒子微物理过程源项最大转化率的时间变化：（a）雨水；（b）雪；（c）霰胚；（d）霰 Figure 8 Temporal evolutions of maximum conversion rate of cloud microphysical source terms in domain d04 from 1200 LT 5 July 2014 to 1200 LT 6 July 2014: (a) Rain water; (b) snow; (c) graupel embryos; (d) graupel growth
5 那曲区域水分收支特征

 $\mathit{\boldsymbol{Q}}={{Q}_{\lambda }}\mathit{\boldsymbol{i}}+{{Q}_{\varphi }}\mathit{\boldsymbol{j, }}$ (1)
 ${{Q}_{\lambda }}=-\int_{{{p}_{\text{s}}}}^{{{p}_{\text{t}}}}{\frac{qu}{g}}\text{d}p$ (2)
 ${{Q}_{\varphi }}=-\int_{{{p}_{\text{s}}}}^{{{p}_{\text{t}}}}{\frac{qv}{g}}\text{d}p$ (3)

 图 10 降水再循环模型 Figure 10 Precipitation recycling model

 $\text{PW}=-\int_{{{p}_{\text{s}}}}^{{{p}_{\text{t}}}}{\frac{q}{g}}\text{d}p$ (4)

 图 9 2014年7月5日12:00至6日12:00日平均水汽通量分布。箭头代表水汽通量，右上角方框里的箭头长度代表 100 kg m-1 s-1，作为标准刻度，并非真实值；彩色阴影代表可降水量 Figure 9 Distribution of water vapor flux averaged from 1200 LT 5 July 2014 to 1200 LT 6 July 2014. Arrows stand for water vapor flux, the arrow at the upper right corner of the panel represents 100 kg m-1 s-1 which acts as a standard scale and is not the real value; color shadings stand for precipitable water vapor

Dominguez et al.（2006）建立的降水再循环模型适用于计算短时间尺度的降水再循环率，如日时间尺度，故我们利用该模型计算WRF第三层嵌套模拟区域日尺度上（2014年7月5日12:00至6日12:00）的降水再循环率。图 10是按照Dominguez et al.（2006）建立的降水再循环模型画出的示意图。其中，六面体表示研究区域中任一格点i的水汽输送框架，格点i的水平面面积为AQin表示外界输入的水汽；Qout表示从区域中输出的水汽；E表示地表蒸发量；Pa表示来自外界输送水汽转化的那部分降水；Pm表示来自局地蒸发水汽转化的降水（P=Pa+Pm）；Wa表示来自外界输送水汽的可降水量；Wm表示来自局地蒸发水汽的可降水量（W=Wa+Wm）。

 ${{\rho }_{i}}=\frac{{{P}_{\text{m}i}}}{{{P}_{i}}}$ (5)

 ${{\rho }_{i}}=\frac{{{P}_{\text{m}i}}}{{{P}_{i}}}=\frac{{{W}_{\text{m}i}}}{{{W}_{i}}}$ (6)

 $r=\frac{{{P}_{\text{m}}}}{P}=\frac{\sum\limits_{i=1}^{n}{{{\rho }_{i}}{{P}_{i}}{{A}_{i}}}}{\sum\limits_{i=1}^{n}{{{P}_{i}}{{A}_{i}}}}$ (7)

 $\frac{\partial W}{\partial t}+\frac{\partial \left(Wu \right)}{\partial x}+\frac{\partial \left(Wv \right)}{\partial y}=E-P,$ (8)

 图 11 气块的后向轨迹（小方框为研究区域范围） Figure 11 The backward trajectories followed by the moisture parcels (small rectangle stands for the study area)

6 结论

（1）WRF模式能够基本再现高原夏季对流云的发展演变过程以及降水的日变化特征。

（2）高原夏季对流云云底温度较低，云中含有丰富的过冷云水和霰粒子，冰相过程在高原云的发展和降水产生过程中具有十分重要的作用，霰粒子的融化过程是地面雨水的主要来源。暖雨过程对降水的直接贡献很小，但对云中霰胚形成过程十分重要，这一点与我国东部平原地区有明显差异。霰粒子胚胎的形成主要依赖冰晶与过冷雨水的撞冻转化过程，雪粒子和过冷雨水的碰冻转化及过冷雨水的均质冻结过程贡献较小。霰粒子的增长主要依靠对过冷云水的凇附过程，其次是其对雪粒子的聚并过程。

（3）那曲地区净水汽收支为正，水汽充沛，为对流活动的发展提供了条件。日平均降水转化率可达20.75%，接近长江下游地区，高于华北、西北地区。该地区的日降水再循环率为10.92%，说明局地蒸发的水汽对高原降水的水汽来源具有一定的贡献，但高原降水的90%仍然由外界输入的水汽转化形成。