Research On Retrieval Of Lower Atmospheric Profiles By Ground-based Microwave Remote Sensing And Radiosonde Data

The remote sensing of temperature,water vapor and refractivity of lower atmosphere which is the main environment for people's production and life is important to the weather forecasting,natural disaster prevention,meteorological research and other fields.In this thesis,the principle and techniques of remote sensing atmospheric parameters based on ground-based multi-channel microwave radiometer have been studied.Main works of this thesis are as follows:1.According to the principle of atmospheric measurement by ground-based microwave radiometer,the weighting functions of the radiation brightness temperature at different frequencies,different zenith angles and different time are simulated,and the peak heights are obtained.The brightness temperatures at different zenith angles of nine oxygen frequency bands are calculated by the historical radiosonde data of Qingdao.Based on the distribution of the weighting functions,the combination of nine frequencies and zenith angles of every 5° from 0° to 85° is selected.Linear regression algorithm and neural network algorithm are used to retrieve lower atmospheric temperature profiles of Qingdao.The simulation results show that the accuracy of temperature profiles retrieved by the combined brightness temperatures compared with zenith brightness temperatures is improved by 5%~30%.2.The brightness temperatures at eight K-band frequencies,four zenith angles(0°,45°,60°,75°)are simulated by the historical radiosonde data.Linear regression algorithm and neural network algorithm are used to retrieve lower atmospheric water vapor profiles of Qingdao.The simulation results indicate that the accuracy of humidity retrieval is slightly improved using the multiangle brightness temperatures compared with zenith brightness temperatures.3.Based on the observation data measured by multi-channel microwave radiometer in Qingdao and radiosonde data measured at Qingdao meteorological station,the temperature and water vapor profiles have been retrieved.The results of the experiment verifies the superiority of retrieving temperature profiles by combined brightness temperatures contrasted with zenith brightness temperatures and slant path brightness temperatures.However,the accuracy of retrieving water vapor profiles by the multiangle brightness temperatures contrasted with zenith brightness temperatures is slightly improved.4.Based on the historical radiosonde data of Qingdao,the occurrence frequencies of temperature inversion,intense humidity gradient and atmospheric duct are statisticed and the characteristics of temperature inversion are analyzed.The temperature inversions are retrieved by neural network algorithm with combined brightness temperatures.The simulation results show that the retrieval accuracy of temperature inversion is more than 76%,and the characteristics of retrieving temperature inversion are agreed with sounding data.The intense humidity gradient and atmospheric duct have also been retrieved.The simulation results show that the ground-based microwave radiometer is feasible to detect the abnormal structures in the lower atmosphere.