Optimized retrieval method for atmospheric temperature profiling based on rotational Raman lidar

Abstract

Aimed at addressing the disadvantages of restricted retrieval height caused by signal-to-noise ratio (SNR) differences between high-quantum-number and low-quantum-number pure rotational Raman scattering signals (PRRSs) obtained with the traditional retrieval method, an optimized retrieval method is proposed for atmospheric temperature profiling based on rotational Raman lidar. This method allows independent alternating solutions to high- and low-quantum-number PRRSs, where high-quantum-number PRRS lidar returns are used to solve the channel constant, and low-quantum-number PRRS returns with a high SNR are used for retrieving temperature profiles. The system sensitivity, SNR, and statistical error in temperature measurements by the two methods are first simulated and discussed, and the results are then compared to show that a higher SNR and stable sensitivity can be attributed to stable statistical errors with the optimized method. A further assessment is demonstrated by three sets of lidar data from a multifunctional Raman-Mie lidar system at the Xi'an University of Technology (34.233°N, 108.911°E). The retrieved atmospheric temperature profiles under different weather conditions are compared with radiosonde data; then, the temperature deviations are further evaluated, and a correlation analysis is performed to evaluate the reliability and correctness of the temperature data obtained by the optimized retrieval method. The results show that the effective temperature retrieval height can be greatly improved from 17 to 25 km under clear weather conditions, and a high correlation >0.99 and stable relative deviations of less than 5 K can be obtained up to 25 km. Additionally, the retrieval height can be extended from 8 to 16 km in cloudy weather, and the existence of an inversion layer can be successfully captured as well. It is evident that the proposed optimized method will provide a new and reliable retrieval theorem for atmospheric temperature profiles, and the proposed method is propitious for retrieving temperature profiles over a larger height range, even up to the lower stratosphere. It is also deduced that the proposed algorithm can favorably simplify the spectroscopic system for temperature detection in the future when the channel constant is determined in advance.