Abstract: Water vapor is one of the fundamental thermodynamic variables that define the state of the atmosphere. It is not only of great significance to the radiation and energy balance of the Earth's atmosphere, but also closely related to environmental issues such as the greenhouse effect. Therefore, detecting the spatiotemporal distribution of atmospheric water vapor content is of great significance to global or regional meteorological and climate research, weather forecasting, etc. Differential absorption lidar (DIAL) technology mainly utilizes the differential absorption effect of water vapor at two neighboring wavelengths, one wavelength with strong absorption (λon) and the other with much weak absorption (λoff), to detect water vapor profiles. It has the characteristics of strong echo and superior daytime performance. DIAL has been considered as an important candidate for remote sensing of atmospheric water vapor profiles, but so far, there is still a lack of systematic quantitative analysis on the influencing factors in the retrieval of the water vapor profile for 828 nm H2O-DIAL. This article establishes a theoretical model for an 828 nm narrowband H2O-DIAL system based on differential absorption spectroscopy of water vapor and atmospheric lidar theory. And based on Monte Carlo method, the influence of molecular Doppler broadening effect, atmospheric temperature profile, frequency stability of the light source, signal-to-noise ratio, and other factors on the retrieval profile of water vapor have been carefully evaluated. It has been found out that the maximum retrieval error of water vapor concentration profile caused by the Doppler broadening effect can reach up to 7%, the error of retrieving the water vapor profile using the average temperature profile obtained from local radiosonde is generally less than 2.5%, and when the frequency fluctuation of the laser is less than 400 MHz, its impact on the retrieved water vapor profile can be ignored. In addition, it is shown that the noise of echo signal has a significant impact on the inversion result. To maintain an inversion error within 10% at 10 km (fitting distance of 300 m), the signal-to-noise ratio of the differential absorption curve at 10 km should be greater than 30 dB. The evaluation result in this work indicates that the uncertainty of temperature profile and the Doppler broadening effect of molecular scattering should be carefully considered for high precision measurements (< 5%). This work will provide important guidance for the design and optimization of near-infrared 828 nm H2O-DIAL system.

Key words: differential absorption lidar, water vapor, retrieval error, Doppler broadening