摘要
利用中尺度数值模式WRFV3对2009年1、4、8和10月进行了约4个月的逐日(0~36小时)到达地表短波辐射模拟,将输出结果转换成逐小时太阳总辐射,并以武汉站为例,与地表实测逐小时总辐射值进行对比分析。结果表明:(1)模拟结果与观测结果有较好的对应关系,4个月的日间模拟和观测值的相关系数均在0.8以上,均通过了置信度0.01的显著性检验。(2)日间各时刻模拟值与实况相比,平均预报误差早晚小,中午大;平均绝对误差变化趋势与平均预报误差相同,平均绝对百分比误差则一般早晚大,09—15时较小且变化幅度小,可信度高于其他时刻。(3)WRFV3模式对太阳总辐射,在各种气象条件下均有一定预报能力,尤其以晴天表现最佳。距直接使用模式输出的地表短波辐射结果作为光伏发电系统初值用于发电量预报还有一些误差,可以尝试采用误差订正等方法修订,提高数值预报结果的可用性。
In this paper,the 0—36 h downward shortwave flux at ground surface(SWDOWN) forecasting has been simulated by WRFV3 continuously for approximately four months;January,April,August and October in 2009.The hourly total solar radiation(HTSR) forecasting was converted by the 12—36 h SWDOWN output from WRF simulation.And HTSR forecast was compared with the observation of Wuhan Station.The results show that:(1) the simulation of HTSR is interrelated well with observation.Correlation of every month is larger than 0.8 and corresponds to 0.01 significance levels.(2) Each time HTSR mean forecast errors(MFE) in the morning and evening are smaller than in other time,but the maxima of mean absolute percentage errors(MAPE) appear in the morning and evening.MAPEs from 9 AM to 3 PM change a little and their reliability is higher than in other time.(3) WRFV3 model has ability to forecast the total solar radiation in clear,cloudy or overcast weather conditions,especially good performance in clear sky.However,this simulation is not a perfect initial field for power prediction of photovoltaic systems. Our future work is to use some methods for correcting HTSR forecast and improve their practical application.
出处
《气象》
CSCD
北大核心
2012年第5期585-592,共8页
Meteorological Monthly
基金
公益性行业(气象)科研专项(GYHY201006036)
华中区域气象中心科技发展基金重点项目(QY-Z-201010)
中国气象局(气预函[2010]76号和气预函[2011]35号)共同资助
关键词
地表短波辐射
太阳总辐射
WRF模式
太阳能
downward shortwave radiation
the total solar radiation
WRF model
solar energy