Differential Absorption Lidar (DIAL) is capable of detecting temperature profiles within the atmospheric troposphere by measuring the variation of oxygen (O₂) absorption coefficient with altitude or temperature. However, in practical applications, the detection accuracy of the O₂ absorption coefficient and thus the temperature profile is influenced by various factors. Currently, the temperature detection error of O₂-DIAL is at 3-10 K. Therefore, it is necessary to establish theoretical models and systematically analyze various influence factors of the O₂-DIAL technique to improve the temperature detection accuracy. For this purpose, this paper focuses on the effects of noise (signal-to-noise ratio, SNR), Doppler broadening of molecule scattering, specific humidity, and laser wavelength stability on the temperature retrieval results for the O₂-DIAL technique. Consequently, this study provides important theoretical support and guidance for the design and implementation of an O₂-DIAL system and lays a crucial foundation for optimizing the retrieval algorithm of the temperature profile.Based on the O₂ absorption spectrum (Fig.1) and the atmospheric model (Fig.2-3), a simulation model of the O₂-DIAL technique operating at 770 nm has been developed. The on-resonance wavelength and the off-resonance wavelength are selected to be 769.7958 nm (λ_on) and 769.8156 nm (λ_off), respectively. The impacts of noise (SNR), Doppler broadening of molecular scattering, specific humidity, and laser wavelength stability on the retrieved temperature profile have been investigated based on the Monte Carlo method and the O₂-DIAL model. Lidar signals added by random noise with different SNRs are used for the retrieval of the O₂ absorption coefficient and thus the temperature profile based on an iterative approach. The Doppler broadening effect of molecular scattering has also been added into the simulation model, while it is neglected in the retrieval process to evaluate its influence on the retrieved temperature profile. In addition, the temperature retrieval results of three atmospheric models with different aerosol distributions are also compared to illustrate the influence of the aerosol gradient. Besides, the temperature retrieval results under different specific humidities are simulated to show the influence of specific humidity. Finally, the influence of laser wavelength stability (including wavelength shift and fluctuation), has also been investigated based on Monte Carlo method.Simulation studies have shown that these factors have different impacts on the retrieval accuracy of the temperature profile. As the increasing of the measurement altitude, the SNR decreases (Fig.4), and the temperature deviation will significantly increase. In addition, the larger the segmented fitting distance, the smaller the retrieval error (Fig.5). Therefore, in order to accurately retrieve the temperature profile, it is crucial to improve the SNR. At the same time, in low SNR situations, the segmented fitting distance can be prolonged to reduce retrieval errors. If the influence of the Doppler broadening effect is neglected during the retrieval process, the temperature retrieval error (Fig.8) may increase significantly (up to 12 K) especially at the altitude with large gradients in aerosol load. Therefore, the influence of the Doppler broadening effect should be carefully considered during the retrieval process for high accuracy retrieval of the temperature profile. When the deviation of specific humidity is less than 0.02, the temperature retrieval deviation will be less than 1 K (Fig.11). Therefore, to accurately retrieve the temperature profile, it is necessary to obtain an accurate specific humidity profile. If the frequency shift and fluctuation of the laser are controlled within 50-100 MHz, the retrieval deviation of temperature profile will be less than 1 K (Fig.15).According to the above discussion, the noise (SNR), Doppler broadening of molecular scattering, specific humidity, and laser wavelength stability are important factors affecting the retrieval accuracy of temperature profile. In practical measurements, if the frequency shift and fluctuation of the laser source should be controlled within 50-100 MHz, the corresponding temperature retrieval deviation will be less than 1 K or even negligible. The influence of specific humidity on the retrieved temperature results is relatively small. In actual measurements, by using the specific humidity profile data from radiosondes, the measurement error caused by the uncertainty of specific humidity can be effectively reduced. If the influence of aforementioned factors can be reduced to a negligible level, the noise and the Doppler broadening effect become the main factors influencing the retrieval accuracy of the temperature profile, which should be carefully considered in practical measurements.