Research Of Surface Air Temperature Observation Instrument And Radiation Error Correction Method

The magnitude of surface air temperature change is on the order of 0.1 ? per decade.However,due to solar radiation effect,current air temperature sensor probe inside a thermometer screen or a naturally ventilated radiation shield may produce a radiation error that is on the order of 1 ?.The difference of air temperatures inside and around a thermometer screen or a radiation shield has become a bottleneck for improving the accuracy of surface air temperature observation.To tackle this problem,the actual temperature of the atmosphere in the periphery of a thermometer screen or a radiation shield needs to be obtained by radiation error correction.Furthermore,temperature sensor designs that feature small radiation error will be necessary as well.Since the physical environment of a temperature sensor is complex,the heat balance equation of the temperature sensor needs to be solved by coupling internal heat conduction,external forced-convection and radiation heat transfer.However,the complicated coupling heat transfer problem is still a difficult one to be solved by using some theoretical approaches.Therefore,accurately and quickly acquiring the heat transfer characteristic of the temperature sensor under a multi-physical field is the means of improving radiation error correction accuracy.It is also an emphasis and difficulty of this dissertation.In order to solve this problem,a multi-physics coupling model based on a computational fluid dynamics(CFD)method is implemented to obtain radiation errors under various environmental conditions.Then,a radiation error correction equation is obtained by fitting the CFD results using a neural net algorithm method.In order to verify the performance of the new temperature sensors proposed in this dissertation,and to verify the accuracy of the correction method,which is based on the CFD and neural net algorithm methods,some field intercomparisons are conducted.The main contents and conclusions of this dissertation are as follows:Based on the theory of convective heat and radiation heat transfer,a multi-physics model for the temperature sensor equipped with a thermometer screen,a naturally ventilated radiation shield or an aspirated radiation shield is set up.The multi-physics model is implemented to obtain radiation errors under various environmental conditions including different wind speeds,solar radiation intensities,upwelling long-wave radiation intensities,reflectivities of the underlying surfaces,reflectivities of the sensor surfaces,sun elevation angles and air densities.The CFD results show that the radiation errors of the temperature sensors inside the thermometer screen,inside the naturally ventilated radiation shield and inside the aspirated radiation shield are on the order of 0.5 ?,1 ? and 0.1 ?,respectively.To improve the air temperature observation accuracy,a naturally ventilated temperature sensor array based on tube-shaped shields and a naturally ventilated high reflectivity sphere-shaped temperature sensor are proposed in this dissertation to improve the air temperature observation accuracy.In order to overcome the air temperature rise problem of a temperature sensor inside a traditional aspirated radiation shield caused by heat pollution,a rotatable aspirated standard temperature sensor is designed in this dissertation.The aspirated standard temperature sensor under a aluminum plate with a mirror surface can be served as an air temperature reference.The heat transfer analysis of the above three new temperature sensorss are completed by using the CFD method.The CFD results show that the radiation errors of the temperature sensor array,the high reflectivity sphere-shaped temperature sensor and the aspirated standard temperature sensor without an aluminum plate are on the order of 0.05 ?,0.1 ? and 0.05 ?,respectively.Because the CFD method can only obtain limited radiation error results,a correction equation for radiation error may be needed.Several multivariable radiation error correction equations are fitted by using a neural net algorithm method.On this basis,a radiation error correction software is designed.The radiation error correction software can be used in weather station to correct the air temperature real-time and to correct the historical air temperature data.Then,the air temperature data after correction may serve for climate change research and other fields.In order to verify the accuracy of the correction method,and to verify the performance of three new temperature sensors proposed in this dissertation,some field intercomparisons are conducted.The experimental results show that the average radiation errors of the temperature sensors inside a reinforced plastics screen,inside a wooden screen,inside a HYTFZ02 naturally ventilated radiation shield,inside a DTR503A naturally ventilated radiation shield and inside a 41003 naturally ventilated radiation shield are 0.85 ?,0.5 ?,0.67 ?,0.36 ?and 0.23 ?,respectively.The average radiation errors of the temperature sensor array,the high reflectivity sphere-shaped temperature sensor and the aspirated standard temperature sensor without a aluminum plate are 0.03 ?,0.05 ? and 0.05 ?,respectively.The radiation errors given by the correction equations are agree with the radiation errors given by the intercomparison experiment with the reference of aspirated standard temperature sensor with a aluminum plate.The radiation error correction method proposed in this dissertation may be helpful to provide a relatively accurate air temperature measurement result.At last,the historical air temperature data of three national basic meteorological stations are corrected by using the radiation error correction method proposed in this dissertation.Then,a relatively accurate air temperature data may be provided.