中尺度天气预报模式边界层参数化方案以及近地层方案对苏州东山冬季近地层气象要素模拟的影响

运用WRF3. 9模式并选取四种常用的边界层参数化方案(YSU、ACM2、MYJ和BL)和两种近地层方案(Eta和MM5)模拟了2015年1月16～25日苏州东山的近地层气象要素,并与东山气象站观测实验数据进行对比,评估了四种边界层参数化方案对近地层气象要素的模拟能力。同时设置了边界层参数化方案与近地层方案耦合的敏感性试验,分析两类方案的耦合对近地层气象要素模拟的影响。结果表明:①白天四种边界层方案对2 m温度的模拟差异较小,两种近地层方案的模拟结果有差异,对流混合较弱时Eta方案模拟较好,对流混合较强时MM5方案较好;夜间四种边界层方案和近地层方案均有影响,但是整个观测期间四种方案的模拟结果统计量差异较小;②无论白天还是夜间,四种边界层参数化方案模拟的2 m相对湿度均高于观测值,BL方案的模拟效果最佳,MYJ方案的模拟偏差最大;③无论白天还是夜间,四种边界层参数化方案对10 m风速的模拟均存在一定程度的高估,MYJ方案的模拟效果最好,四种方案对风向的模拟均优于对风速的模拟,白天的模拟效果整体优于夜间;④选择不同的近地层方案对风速和风向的模拟结果没有明显影响,对2 m气温模拟结果的影响小于对2 m相对湿度模拟结果的影响,BL边界层方案与MM5近地层方案耦合对近地层气象要素2 m气温和2 m相对湿度的模拟效果最好。

The Effect of Planetary Boundary Layer Parameterization Schemes and Their Coupling with Surface Layer Schemes on the Simulation of Surface Layer Meteorological Factors at Dongshan, Suzhou in Winter by WRF

In this study, four commonly used planetary boundary layer (PBL) parameterization schemes (YSU, ACM2, MYJ and BL) in the WRF model were utilized together with the observed data at Dongshan meteorological station in Suzhou city, China, from January 16 to 25, 2015. The effects of different PBL schemes and their coupling with different surface layer schemes on the simulation of surface layer meteorological factors were compared and evaluated. Results revealed that the four boundary layer schemes have little effect on the simulation of 2 m temperature, but the simulation results of two surface layer schemes are different during the daytime. The Eta scheme is better when the convection mixing is weaker, and the MM5 scheme is better when the convection mixing is stronger. Although the boundary layer schemes and the surface layer schemes have effects on the 2 m temperature, the statistical differences of the four schemes are small during the whole observation period. The simulated values of 2 m relative humidity were mostly higher than the observed values, day or night; the simulation using BL scheme did the best and MYJ scheme had the maximum deviation. Simulations using the four PBL schemes all overestimated wind speed at 10 m level to a certain degree, MYJ scheme did the best during the daytime and nighttime. The simulations of wind direction are better, and the simulations of daytime are better than that in nighttime. The selection of different surface layer schemes had no obvious influence on 10 m wind speed and wind direction, but it had a different degrees of influence on 2 m air temperature and 2 m relative humidity; the coupling of BL PBL scheme and MM5 surface layer scheme did the best simulation for air temperature and relative humidity at 2 m level.