水平空间分辨率对于模式模拟影响的验证一直是大气模型重要的研究内容,提高分辨率能够细致描述地形、下垫面状况等区域特征.准确的下垫面信息对于区域大气模拟的精度具有显著的影响。本文对区域大气模型Weather Research and Forecasting Model(WRF)在黑河流域进行5km和1km的对比试验,检验水平分辨率对WRF模拟精度的影响。通过与黑河综合遥感联合试验中7个站点观测数据的比较,以平均误差、均方根误差和相关系数为指标,分析了WRF模式下垫面替换前后近地表气象要素的模拟精度。结果表明:1)水平分辨率为1km的WRF模式模拟的近地表气象要素刻画得比5km的WRF模式模拟更加精细,更加符合近地表下垫面的变化特征;2)采用1km水平分辨率,黑河综合遥感联合试验(Watershed Allied Telemetry Experimental Research,以下简称:WATER)站点所在的格点的高程信息更加准确;3)采用1km水平分辨率,WRF模式模拟的近地表气压与观测值的误差比5km的WRF模拟值的误差减少很多,尤其是在黑河流域上游区域。
The research of coupling WRF(Weather Research and Forecasting Model)with a land surface model is enhanced to explore the interaction of the atmosphere and land surface;however,regional applicability of WRF model is questioned.In order to do the validation of WRF model on simulating forcing data for the Heihe River Basin,daily meteorological observation data from 15 sta-tions of CMA(China Meteorological Administration)and hourly meteorological observation data from seven sites of WATER(Watershed Airborne Telemetry Experimental Research)are used to compare with WRF simulations,with a time range of a whole year for 2008.Results show that the average MBE(Mean Bias Error)of daily 2-m surface temperature,surface pressure,2-m rela-tive humidity and 10-m wind speed were-0.19 °C,-4.49 hPa,4.08% and 0.92 m/s,the average RMSE(Root Mean Square Error)of them were 2.11 °C,5.37 hPa,9.55% and 1.73 m/s,and the average R(correlation coefficient)of them were 0.99,0.98,0.80 and 0.55,respectively.The average MBE of hourly 2-m surface temperature,surface pressure,2-m relative humidity,10-m wind speed,downward shortwave radiation and downward longwave were-0.16 °C,-6.62 hPa,-5.14%,0.26 m/s,33.0 W/m2 and-6.44 W/m2,the average RMSE of them were 2.62 °C,17.10 hPa,20.71%,2.46 m/s,152.9 W/m2 and 53.5 W/m2,and the average R of them were 0.96,0.97,0.70,0.26,0.91 and 0.60,respectively.Thus,the following conclusions were obtained:(1)regardless of daily or hourly validation,WRF model simulations of 2-m surface temperature,surface pressure and relative humidity are more reliable,especially for 2-m surface air temperature and surface pressure,the values of MBE were small and R were more than 0.96;(2)the WRF simulating downward shortwave radiation was relatively good,the average R between WRF simulation and hourly observation data was above 0.9,and the average R of downward longwave radiation was 0.6;(3)both wind speed and rainfall simulated from WRF model did not agree well with observation data.
利用高精度的土地覆盖、土壤质地类型和地形高度值替换了天气研究和预报模式Weather Research and Forecasting Model(WRF)中的相关数据,通过数值模式试验检验了下垫面数据对WRF模拟精度的影响。同时,通过与黑河综合遥感联合试验中7个测站观测数据的比较,以平均误差、均方根误差和相关系数为指标,分析了WRF模式下垫面数据改变对近地表气象要素的模拟精度的影响。结果表明:(1)WRF模式本身的地形高度信息在黑河流域上游地区有较大误差,造成了一定的模拟误差。而使用高精度的下垫面数据可以提高WRF模式在黑河流域上游复杂区域的模拟能力;(2)2m气温除了随地形高度递减外,还受土壤质地和土地覆盖小幅度影响,而且进行地形订正后的2m气温与2m湿度的模拟在下垫面为水体的区域对比强烈,因此为模式提供准确的水体分布信息也至关重要;(3)2m气温和湿度等要素的模拟差异值与地形高度资料的差异呈负相关,而降雨量的差异与地形高度差异呈微弱的正相关,与土壤质地差异和土地覆盖差异的相关性也比较弱。
