Effective monitoring approaches for carbon dioxide (CO₂) have become critical as the impact of increasing atmospheric CO₂ concentration on the global climate system intensifies. Satellite remote sensing technology is the prevailing method for CO₂ monitoring, and the key to successful retrieval lies in constructing forward models. Traditional forward modeling software, although capable of simulating atmospheric radiative transfer processes, suffers from limitations such as low resolution, poor computational efficiency, the neglect of scattering effects, and the inability to integrate real-time measurement data into simulations. To address these issues, this study employs the Line-By-Line (LBL) calculation method and Mie theory to calculate the spectral properties of various atmospheric components, selecting ideal bands for CO₂ retrieval. Furthermore, a forward model for CO₂ radiative transfer was developed based on the Discrete Ordinate Radiative Transfer (DISORT) method. The forward model accounts for multiple scattering, achieves high resolution, and is capable of integrating real-time environmental observation data into radiative transfer simulations. To address challenges such as the uncertainty of boundary conditions, physical parameters, and the unknown sensitivity of environmental factors, the model was used to analyze the impact of different environmental parameters on the spectral radiance of CO₂-sensitive bands. These findings provide a theoretical basis for the development of atmospheric CO₂ concentration retrieval algorithms, the selection of environmentally sensitive parameters, and the analysis of retrieval errors.Accurate gas absorption coefficients in the atmosphere are first calculated using the LBL method, followed by the computation of aerosol spectral property parameters via Mie theory. High-resolution solar spectra, underlying surface types, and atmospheric models are selected, with their results incorporated into the atmospheric radiative transfer equation. The equation is then solved using the DISORT method to obtain radiance results under arbitrary solar zenith and azimuth angles. The forward simulation results are convolved with the instrument response function to produce the final forward model outputs. After identifying CO₂-sensitive spectral bands, the simulated results of the model are compared with GOSAT-2 satellite observations to validate its accuracy. Finally, the model is used to analyze the impact of environmental parameters, such as surface types, solar zenith angles, aerosol types, and Aerosol Optical Depth (AOD), on the spectral radiance within CO₂-sensitive bands.The results indicate that CO₂ in the 6300–6400 cm^-1 band is minimally affected by other gases, with moderate absorption, making it highly suitable for CO₂ retrieval. The normalized simulation results of the model within this band exhibit a consistent trend with the wavelength-dependent variation trend of the normalized detection results from the GOSAT-2 satellite L1B product (Fig.5), demonstrating the validity of the model. Sensitivity analysis reveals that an increase in surface albedo results in a corresponding rise in reflected radiance, thereby enhancing radiance as surface albedo varies across different surface types (Fig.6). When the surface albedo difference reaches 0.14, the difference in the average rate of relative radiance change is 123.29% (Tab.2). As the solar zenith angle increases, the optical path length grows, resulting in a decay in radiance (Fig.7). The relative radiance change exhibits bimodal characteristics (Fig.8). Aerosols, due to their varying compositions, significantly impact radiance. Urban aerosols, which include strongly absorbing components, cause substantial radiance attenuation (Fig.9), with the average relative radiance change reaching -37.66% (Tab.3). An increase in AOD leads to distinct radiance outcomes for different aerosol types (Fig.10). Urban aerosols show high sensitivity to radiance changes, with radiance rapidly decreasing as AOD increases. In contrast, maritime aerosols, characterized by strong scattering properties, result in a slight enhancement of radiance. The average rate of relative radiance change for maritime and rural aerosols remains within ±5% (Fig.11).A high-resolution forward model was developed to simulate the spectral radiance of CO₂-sensitive bands, incorporating scattering effects and real-time environmental observation data. The results demonstrate that the selection of environmental parameters has a significant impact on forward modeling in the regional atmospheric CO₂ retrieval based on satellite data. During the retrieval process, it is recommended to use surface albedo data derived from MODIS satellite observations that are spatiotemporally matched with carbon-monitoring satellites. Additionally, high signal-to-noise ratio observations with smaller zenith angles should be utilized to achieve more accurate retrievals. Moreover, the multiple scattering and absorption effects caused by aerosols cannot be ignored, particularly when retrieving atmospheric CO₂ concentrations over urban areas. To minimize uncertainties caused by aerosols, prioritizing data with lower AOD is recommended. These findings provide a theoretical foundation and model basis for the development of atmospheric CO₂ retrieval algorithms, the selection of environmentally sensitive parameters, and the analysis of retrieval errors.