中文摘要：

藏东南地区是青藏高原山地复杂下垫面的典型代表,其边界层大气过程异常复杂,给数值模拟和预报带来较大困难。大气边界层参数化方案的选取关系到能否正确模拟和预报局地大气过程。本研究采用中尺度模式WRF（Weather Research and Forecasting Model）对藏东南林芝地区对流和稳定边界层大气过程进行模拟,与2013年夏季＂藏东南地区复杂下垫面地气交换观测实验＂资料对比,研究ACM2,Boulac,M YJ,QNSE和YSU5种边界层参数化方案在青藏高原复杂下垫面的适用性。结果表明：对于水汽混合比垂直结构的模拟,Boulac和MYJ方案分别在模拟对流边界层和稳定边界层时能力最优。ACM 2方案最适宜藏东南复杂下垫面条件下的位温和风速垂直分布的模拟。各边界层参数化方案模拟对流边界层高度均较实际观测偏低,其中,QNSE方案模拟的边界层高度最接近观测。同一种边界层参数化方案对于夜间稳定边界层和正午对流边界层的模拟能力也不相同。该地区边界层风场受地形影响显著,风速较小,模拟的近地层风场较观测偏弱,MYJ和QNSE方案对近地层风场的模拟效果较好。

英文摘要：

Southeast Tibet is the typical representative of complex underlying surfaces of the Qinghai-Tibetan Plateau,the boundary layer processes above this region are quite complex,which greatly challenge modeling and numerical weather forecasting. Accurate simulation and forecasting of local atmospheric processes depend on the choice of boundary layer parameterization scheme. In this study,the Weather Research and Forecasting model（ WRF） was employed to simulate the convective and stable atmospheric boundary layer processes of Nyingchi district,Tibet. By comparing with the results from comprehensive campaign ＂Observation on the surface-to-air Exchange Processes in the Southeast Himalayas＂（ OSEP） carried out during the summer of 2013,the 5 boundary layer schemes ACM2,Boulac,MYJ,QNSE,and YSU from WRF model were evaluated for their applicability above complicated underlying surface in southeast Tibet. The results indicate that Boulac scheme and MYJ scheme perform the best on convective and stable boundary layer vapor mixing ratio simulation,respectively. ACM2 scheme is a more preferable parameterization scheme on potential temperature and wind filed vertical structure simulation over the complex underlying surface of southeast Tibet. Convective boundary layer height simulated by all schemes are lower than observations; While QNSE simulation shows the minimum deviation. Simulation ability of the same boundary layer parameterization scheme on convective boundary layer and stable boundary layer are different.Boundary layer wind field in this region is greatly weakened by orography. Near surface wind speed simulated are weaker than observation,MYJ and QNSE parameterization scheme shows the best result.