资源环境科技发展态势分析平台

Due to solar radiation effect, air temperature sensor inside a reinforced plastic screen may produce a measurement error that is 0.8 degrees C or higher. To improve air temperature observation accuracy and correct historical temperature of weather stations, a radiation error correction method is proposed. The correction method is based on a computational fluid dynamics (CFD) method and a neural network method. The CFD method is implemented to obtain the radiation error of a reinforced plastic screen under various environmental conditions. Then, a radiation error correction equation is obtained by fitting the CFD results using the neural network method. To verify the performance of the radiation error correction equation, a reinforced plastic screen and an aspirated temperature measurement platform are characterized in the same environment to conduct the intercomparison. The aspirated temperature measurement platform serves as an air temperature reference. The average radiation error given by four sunny days intercomparison experiments is 0.85 degrees C. The corresponding average radiation error given by the correction equation is 0.83 degrees C. The mean absolute error, the root mean square error and the correlation coefficient between the radiation errors given by the correction equation and the radiation errors given by the experiments with the reference platform are 0.099, 0.109 and 0.713 degrees C, respectively.