Fig. 1. Water vapor absorption cross-sections at different temperatures^[39]

Fig. 2. Emission spectra, scattering spectra, and water vapor absorption cross-section (simulated temperature 296 K, simulated pressure 101.325 kPa). (a) Emission spectrum and water vapor absorption cross-section, emission spectrum linewidth 0.15 pm (65.6 MHz); (b) scattering spectrum and water vapor absorption cross-section, the Doppler broadening of the scattering spectrum is 1.8 pm (787.3 MHz)

Fig. 3. Simulated profiles of the atmospheric aerosol and molecular backscattering coefficients. (a) Atmospheric aerosol with high concentration; (b) atmospheric aerosol with low concentration

Fig. 4. Radiosonde data measured on July 1, 2024. (a) Temperature profile; (b) pressure profile; (c) relative humidity profile;(d) calculated water vapor mass concentration profile

Fig. 5. Lidar profiles and the corresponding DIAL ratio curve. (a) Lidar profiles in atmospheric model with high aerosol concentration; (b) the DIAL ratio curve in atmospheric model with high aerosol concentration; (c) lidar profiles in atmospheric model with low aerosol concentration; (d) the DIAL ratio curve in atmospheric model with low aerosol concentration

Fig. 6. Lidar profiles (a) and the relative errors of water vapor concentration inversion (b) with and without considering Doppler effect into the atmospheric model with high aerosol concentration and lidar profiles (c) and the relative errors of water vapor concentration inversion (d) with and without considering Doppler effect into the atmospheric model with low aerosol concentration

Fig. 7. Average atmospheric parameters measured by radiosonde at 19:15 every night from June 1 to June 25, 2024. (a) Average temperature profile measured by the radiosonde; (b) average pressure profile measured by the radiosonde; (c) average relative humidity profile measured by the radiosonde; (d) average water vapor mass concentration profile

Fig. 8. Comparison (a) and deviations (b) between actual temperature profile on July 1, 2024 and local historical average temperature profile

Fig. 9. The relative errors based on the radiosonde temperature profile on July 1, 2024 and the local historical average temperature profile, in comparison with the actual water vapor mass profile

Fig. 10. The inversion error of the water vapor mass concentration with different temperature deviations (The temperature deviation along x-axis is the value along the whole measurement altitude between 0–10 km)

Fig. 11. DIAL ratio curves under different frequency fluctuation ranges. (a) 100 MHz; (b) 500 MHz; (c) 1000 MHz

Fig. 12. The inversion error of the water vapor mass concentration with different frequency fluctuations

Fig. 13. Signals with random noise. (a) Lidar profiles with random noise; (b) DIAL ratio curve with random noise;(c) the signal-to-noise ratio of the DIAL ratio curve; (d) the retrieved water vapor profile

Fig. 14. Inversion error of water vapor mass concentration at different altitudes under different SNRs at 10 km with four different fitting ranges. (a) 50 m fitting range; (b) 100 m fitting range; (c) 200 m fitting range; (d) 300 m fitting range

Fig. 15. The relative error of the water vapor profile considering three main factors: the Doppler broadening effect, random noise and the frequency fluctuation. (a) Relative inversion error in atmospheric model with high aerosol concentration; (b) relative inversion error in atmospheric model with low aerosol concentration