摘要
大气科学研究对地表气温观测精度有高达0.1℃甚至0.05℃的需求。然而,现有的地表气温观测仪器受到太阳直接辐射、下垫面反射辐射、长波辐射和散射辐射等影响,辐射误差可达1℃。本文设计了一种基于导流装置的地表气温观测仪器。首先,利用计算流体动力学(Computational Fluid Dynamics,CFD)方法量化该仪器在各种环境条件下的辐射误差;然后,在此基础上,利用极限学习机(Extreme Learning Machine,ELM)方法拟合可针对多变量变化的辐射误差订正方程;最后,为验证该仪器的观测精度,进行了外场比对实验。在实验过程中,以076B型强制通风气温观测仪器的测量值作为温度基准。实验结果表明,该仪器的平均辐射误差和最大辐射误差分别为0.07℃和0.15℃。该仪器辐射误差的实验测量值与订正方程提供的辐射误差订正值之间的平均偏移量、均方根误差和相关系数分别为0.033℃、0.028℃和0.703。
Atmospheric science research requires has a high demand of up to the accuracy of surface air temperature observation to be as high as 0.1℃or even 0.05℃for the observation accuracy of surface air temperature.However,due to the effects of various radiations including direct solar radiation,reflected radiation,long-wave radiation,diffuse radiation from the underlying surface,the radiation error of measured result of existing sensors is higher than the actual atmospheric temperature to the order of can reach 1℃.To improve the observation accuracy,a Surface temperature observation instrument with a novel structure is designed in this study.First,a computational fluid dynamics(CFD)method is used to quantify the radiation errors of this instrument under various environmental conditions.Then,on this basis,an extreme learning machine(ELM)method is employed to fit the radiation error correction equation.Finally,in order to verify the measurement accuracy of this new instrument,field observation of air temperature observation experiments are performed carried out.076B artificially ventilated temperature observation instrument serves as a temperature reference during the experiments.The experimental results demonstrate show that the mean average radiation error and the maximum radiation error are 0.07℃and 0.15℃,respectively.The mean absolute error offset,root mean square error and correlation coefficient between the radiation errors given by the correction equation and the radiation errors given by the experiments field observation are 0.033℃,0.028℃and 0.703,respectively.
作者
杨杰
葛祥建
张道远
刘清惓
丁仁惠
沈瑱
戴伟
YANG Jie;GE Xiangjian;ZHANG Daoyuan;LIU Qingquan;DING Renhui;SHEN Zhen;DAI Wei(Collaborative Innovation Center of Atmospheric Environment and Equipment Technology,Nanjing University of Information Science&Technology,Nanjing 210044,China;Jiangsu Meteorological Observation Center,Nanjing 210041,China;Key Laboratory of MEMS of the Ministry of Education,Nanjing 210096,China)
出处
《气象科学》
北大核心
2023年第2期177-185,共9页
Journal of the Meteorological Sciences
基金
国家自然科学基金资助项目(42275143,41905030)
江苏省产学研合作项目(BY2022544)
江苏省高校大学生创新创业训练计划项目(202210300011Z)
江苏省气象局青年基金项目(KQ202107)。
关键词
地表气温
辐射误差
气温观测仪器
计算流体动力学
极限学习机
surface air temperature
radiation error
air temperature observation instrument
computational fluid dynamics
extreme learning machine
