For atmospheric Carbon dioxide (CO₂) satellite monitoring, high precision is the most important. However, because of the high similarity of spectra between plant chlorophyll fluorescence and atmospheric scattering effects, the atmospheric scattering effects were always misinterpreted and CO₂ retrieval errors were caused. The minor radiance and high spectral similarity of the plant chlorophyll fluorescence made it hard to retrieve and correct. Considering the non-negligible effects of plant chlorophyll fluorescence on CO₂ retrieval, we designed a correction method based on atmospheric scattering effects parameterization using three spectral bands of O₂-Aband, 1.6 and 2.06 μm CO₂ band to retrieve chlorophyll fluorescence, photonpath-length probability density function (PPDF) parameters and CO₂ simultaneously. Firstly, atmospheric scattering effects were parameterized using the PPDF model to reduce its spectral similarity with the plant chlorophyll fluorescence. Secondly, the prior information database with 5 km resolution of plant chlorophyll fluorescence was built using plant chlorophyll fluorescence products from the Orbiting Carbon Observatory(OCO-2), which can strength the prior constraints on chlorophyll fluorescence retrievals and improves its retrieval precision. Through disentangling and quantifying the atmospheric scattering and plant chlorophyll fluorescence, CO₂ will be retrieved more accurate. Park Falls station covered with dense plant location in 45.945°N and 90.273°W of Total Carbon Column Observing Network (TCCON) was selected for validation. CO₂ retrievals of Greenhouse Gases Observing Satellite (GOSAT) without chlorophyll fluorescence correction are about 2 ppm under bias, which was corrected obviously through the method of this paper, as the largest bias reduces from 2.58 ppm to 1.49 ppm. In the same time, the standard deviations are also improved. From the validation results we can see that the effect of plant chlorophyll fluorescence on CO₂ retrieval is corrected effectively.