Fig. 1. Simulation results of lidar echo signals. (a) Ideal echo signal and signal with noise at the altitude range of 30‒70 km; (b) variation in SNR of echo signal with altitude

Fig. 2. Measured data by Rayleigh lidar. (a) Echo signal of lidar; (b) variation in SNR of echo signal with altitude

Fig. 3. Comparison of denoise results by different methods on the measured echo signal, illustrations are enlarged for selected parts. (a) MA; (b) Hann; (c) WTs; (d) EEMD; (e) WT-EEMD-LOWESS; (f) EEMD-VMD-IMWOA

Fig. 4. Characteristic analysis of denoising results by different methods in the high-frequency range of the frequency-domain. (a) Comparison among six methods; (b) fine comparison of MA, Hann, EEMD, and EEMD-VMD-IMWOA

Fig. 5. Effects comparison of different methods on atmospheric fluctuation information. (a) MA; (b) Hann; (c) WTs; (d) EEMD; (e) WT-EEMD-LOWESS; (f) EEMD-VMD-IMWOA

Fig. 6. Atmospheric temperature retrieval results and uncertainties of lidar measured data by different methods. (a) Atmospheric temperature retrieval result and (b) uncertainty of profile 1; (c) atmospheric temperature retrieval result and (d) uncertainty of profile 2; (e) atmospheric temperature retrieval result and (e) uncertainty of profile 3

Fig. 7. Variation in SABER temperature data with altitude

Fig. 8. Deviation between the temperature retrieval results obtained by different denoising methods and the SABER temperature data, the two ends of red line segment are the maximum and minimum deviations. (a) Profile 1; (b) profile 2; (c) profile 3