基于WRF模式的北方地区逆温层特征模拟研究

Modelling Study of Temperature Inversion Layer Characteristics over Northern China Based on WRF Model

摘要: 为分析我国北方地区的逆温层特征，本研究选取北方地区十个代表性站点使用WRF模式进行数值模拟，并结合2019年相关站点临近区域的探空及地面观测数据进行对比分析。WRF模拟结果显示：各站点中发生逆温的最高频率为64%，最低频率为23%，秋冬季发生的频率高于春夏季；大部分站点逆温现象的日变化呈现出了U形分布特征，发生逆温的高频时段集中在世界时21~00时；逆温厚度方面，多数站点呈现出秋冬季高于春夏季的特征，冬季最厚者居多，悬浮逆温厚度在多数站点中均为最厚；逆温强度在所有模拟站点中都是贴地逆温强于悬浮逆温，从不同季节上来看，冬季最强，春秋季相当，夏季最弱；逆温底高和顶高方面，北京、山西和白城悬浮逆温的相对位置较高且接近本地边界层高度。对比分析模式模拟与实际观测结果则表明，世界时0时各逆温指数平均偏差为17%，世界时12时平均偏差为29%，模拟与观测的逆温时次的月变化趋势基本一致。总的来说，WRF模式能比较好的刻画出模拟地区边界层内气温的垂直变化特征。因此，可借助模式模拟弥补外场观测的局限性，为我国北方地区大气污染物的排放和扩散提供研究基础，继而为污染防治措施提供理论支撑。

Abstract: To better understand the characteristics of inversion layer in northern China, ten representative stations in northern China are selected for numerical simulation using WRF model, and compared with the radiosonde and ground observation data in the vicinity of relevant stations in 2019. WRF simulation results show that: the highest frequency of temperature inversion is 64%, minimum frequency is 23%, inversion in autumn and winter occurs more frequently than in spring and summer; temperature inversion at most simulated sites have a daily U-shaped distribution feature, the high-frequency period of temperature inversion happened between 21 UTC and 00 UTC; in terms of the temperature inversion thickness, most sites have thicker inversion layer in autumn and winter than spring and summer, most of the thickest inversion layer happen in winter, the above surface temperature inversion layer thickness is the thickest in most sites; temperature inversion intensity of from surface temperature inversion layer is stronger than above surface temperature inversion layer in all simulation sites, from the different seasons, winter has the highest inversion layer intensity, spring and autumn are equivalent, summer is the weakest; Inversion temperature bottom height and top height aspects, the relative position of the above surface temperature inversion layer in Beijing, Shanxi and Baicheng is high and close to the height of the local boundary layer. Comparative analysis of the model simulation and actual observation results shows that the average deviation of each inversion index at 00 UTC is 17%, and the average deviation at 12 UTC is 29%.The monthly variation trend of the inversion time between the simulation and observation is basically consistent. In general, the WRF model well described the vertical variation characteristics of temperature in the boundary layer over simulated region. Therefore, model simulation can make up for the limitations of field observation, provide a research basis for the emission and diffusion of air pollutants in northern China, and then provide theoretical support for pollution control measures in these areas.