基于数值模拟与加密观测的青藏高原东侧地区大气要素对比分析

Analysis of Atmospheric Elements Based on Numerical Simulation and Intensive Observation at the East Region of the Qinghai- Tibet Plateau

本文将高原东坡及其下游盆地区域加密探空观测的低层大气物理要素场与WRF模式结果进行对比分析,得到如下结论:1)川西高海拔地区,模式格点与站点海拔差异非常大,模式地形普遍偏高,最大差值超过上千米。低海拔地区,模式格点与站点海拔比较接近。2)在高海拔地区,差异主要体现在近地层大气中;00时的比湿差异最小;06时的比湿差异最为显著,模拟的低层大气的比湿比探空观测值大。06时模拟的温度高于探空观测,其它12、18、00时3个时次则略低于探空观测。除了初始场,模拟的低层大气的水平风速普遍比探空观测的值大。3)在低海拔地区,模式初始场给出的低层大气比湿、温度与探空观测差异较小;06、12、18时,模拟的大气比湿通常比探空观测偏湿,温度也显著偏高,4个时次中,正午时分低层大气的温湿偏差最显著。同一时次,积分时长越短模拟的风速越小,低层大气中常常存在一个风速的大值区。4)模式比较稳定,没有随着模拟时长的增加,误差明显增长。模拟的低层大气比湿、温度、水平风速逐日波动形态与观测基本一致。

Using intensive sounding data in the eastern Tibetan Plateau and its downstream basin, the lower - layer atmos- pheric physics element fields and WRF modeling results are compared and analyzed. The conclusions are as follows ： （ 1 ） In the high altitude area of Western Sichuan, altitude height simulated by model is generally bigger than actually station elevation and the maximum deviation can exceed almost 1,000 meters. While in the lower altitude area, the two are nearly similar. （2） In the high altitude area atmospheric physics elements are different mainly in the surface layer. The difference between specific humidity simulated by model and the sounding data is the smallest at 0：00 and the biggest at 06：00 when the former is bigger than the latter. Air temperature simulated by model is higher at 06：00 and is slightly smaller at the else 3 o＇ clock than sound- ing data. Horizontal wind speed simulated by model is generally bigger than sounding data except for initial field. （3） In the lower altitude area, the differences of specific humidity and air temperature are much small between model initial fields and sounding data in the lower layer. Simulated specific humidity is mostly wetter and simulated temperature is obviously higher than sounding data at 06： 00, 12：00 and 18： 00. Especially at 12： 00, the differences are the most remarkable. The shorter the integral duration is, the smaller the wind speed is. There is usually a strong wind - speed center in the lower layer. （4） WRF model is relatively stable： the error does not grow with the increasing of simulation duration. The daily fluctuation patterns of Specific humidity, temperature and horizontal wind simulated by model are basically consistent with the observation.