Multiphysical Field Analysis of a Temperature Sensor for Meteorological Observation

Abstract

Due to the influence of solar radiation, the observed values of existing meteorological temperature sensors may differ from the free air temperatures up to the order of 1 °C. This article proposed a temperature sensor consisting of a sensor probe, an airflow deflector, and two aluminum plates. The airflow deflector can effectively guide the airflow to the sensor probe and reduce radiation error. The two silver-plated mirror aluminum plates have up to 98% reflectivity. They can effectively block direct solar radiation, reflected radiation, long-wave radiation, and so on. The radiation protection and airflow guiding ability of the sensor are analyzed by the computational fluid dynamics (CFD) method. Then, the CFD approach obtains the radiation errors of the sensor under different environmental conditions. Next, the neural network algorithm fits the simulation data to form a high-accuracy radiation error correction approach. Finally, the 076B artificially ventilated temperature sensor is used as the temperature reference during experiments. The experimental results show that the mean absolute error (MAE) and the root-mean-square error (RMSE) between the radiation errors provided by the experiments and the radiation errors given by the neural network are 0.031 °C and 0.026 °C, respectively. After correction, the maximum, minimum, and average radiation errors of the new sensor are 0.095 °C, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$-0.074\,\,^{\circ }\text{C}$ </tex-math></inline-formula> , and 0.01 °C, respectively. The correlation coefficient between the temperature results of the new sensor after correction and the reference temperature results is 0.999. These results show that this new sensor might reduce the measurement error to within 0.1 °C.