A detailed assessment is carried out in relation to the influence of the uncertainties associated with the input auxiliary atmospheric state parameters on retrieving aerosol optical properties from high-spectral-resolution lidar (HSRL) observations. The study starts from a review of the main spectral structure of the Rayleigh backscattering followed by evaluating the temperature effects on a backscattering cross section of atmospheric molecules based on numerical simulation. It shows that the transmittance of the background interference filter should be taken into account, depending on the full width at half maximum, although overall temperature dependence is negligible. Based on the Taylor expansion of the Tenti S6 model, the systematic errors arising from input temperature and pressure profiles are analyzed. It is demonstrated that the atmospheric pressure profiles have limited effects on the inversion results of aerosol optical parameters, as the atmospheric pressure is usually quite stable. The relative errors of the aerosol backscatter coefficient mainly stem from temperature profile errors and highly depend on the aerosol concentration. Quantitatively, the aerosol backscatter coefficient error could be larger than 5% with a 3 K deviation of temperature when the backscatter ratio is larger than 1.1. The accuracy of aerosol extinction coefficient retrieval is affected not only by the error in temperature, but also by the error in temperature lapse rate; the retrieval accuracy is more sensitive to the latter than the former. Further analysis based on the sounding temperature data shows that the variation of the temperature inversion layer during the night could induce a bias larger than 0.04 km-1 on the aerosol extinction coefficient retrieval. Therefore, the time resolution of temperature measurement from sounding balloons twice per day is too low to obtain an accurate retrieval of the aerosol optical properties from the HSRL.