Aerosol particles are a crucial component of Earth’s atmosphere, significantly affecting the solar radiation balance and driving both climate and environmental changes, making aerosol research a priority in atmospheric science. In addition, greenhouse gases such as carbon dioxide (CO₂) and methane (CH₄), primarily emitted by human activities, are the main drivers of global warming. Controlling these anthropogenic carbon emissions is essential to mitigating global temperature rise. Through the Paris Climate Conference, countries have implemented a series of policies to address the influences of climate change, providing critical support to mitigate the negative effects of the greenhouse effect. Satellite remote sensing, with its advantages of objectivity, continuity, stability, broad coverage, and repeatability, has become an indispensable tool for monitoring global atmospheric greenhouse gas volume fraction. It is also emerging as the next-generation, internationally recognized approach for global carbon verification. The United States, European Union, Japan, Canada, and China have successively launched satellites equipped for atmospheric greenhouse gas monitoring, with increasingly advanced technology and enhanced detection precision. To meet the high-precision demands of carbon accounting, developing high-accuracy algorithms and data products is also essential. Thus, research into high-precision retrieval methods for greenhouse gases using fully physical algorithms in satellite remote sensing is of great significance. Satellite remote sensing has become an essential tool for monitoring aerosols and greenhouse gases (CO₂ and CH₄). The combination of polarization and hyperspectral techniques can reduce aerosol-induced XCO₂ retrieval errors by more than twofold. To improve the retrieval accuracy of CO₂ and CH₄ in the shortwave infrared (SWIR) band, it is crucial to effectively correct aerosol influences to meet the high-precision requirements for satellite-based greenhouse gas retrievals.

In this paper, we utilize a two-pronged technical approach: forward model construction and joint retrieval of aerosols and greenhouse gases. Observational data, integrating spectral, polarization, and spatial information, are acquired using polarized hyperspectral imaging. A vector radiative transfer model is used for forward simulations, incorporating both spectral radiance measurement covariance and state vector covariance to quantify uncertainties. Retrieval is performed using an optimal estimation method, iteratively adjusting the state vector by comparing actual observations with simulated data to ensure convergence. Once the retrieval process is complete, an error analysis is conducted to assess the reliability of the results, ultimately yielding retrieved parameters for aerosols and greenhouse gases.

In this paper, we integrate observations from multi-band polarimetric sensors and hyperspectral instruments to investigate the influence of different spectral band combinations on the retrieval of aerosol parameters and greenhouse gas column volume fraction, using a radiative transfer model and full-physics inversion method. The results indicate that adding shortwave infrared bands (1610 nm and 2250 nm) to the polarimetric sensor significantly improves the retrieval accuracy of aerosol optical depth (AOD). Specifically, the bias for AOD decreases from 0.078 to 0.024, and the root mean square error (RMSE) is reduced from 0.212 to 0.161 (Fig. 6). The improvement is more pronounced for coarse-mode aerosols, with an RMSE difference of 0.109 and a bias difference of 0.056 compared to fine-mode aerosols (Table 5). In addition, the combined retrieval results from polarimetric and hyperspectral data show high accuracy, with the AOD bias at 1600 nm reaching 0.015 and an RMSE of 0.040 (Fig. 7). The column volume fraction of methane (XCH₄) reaches 4.644×10^-9, and XCO₂ reaches 0.990×10^-6 (Fig. 9). For fine-mode aerosols, the percentage change in error for XCO₂ and XCH₄ is mostly around 0.01, whereas the error percentage distribution for coarse-mode aerosols is more uniform (Fig. 11). Moreover, simulated band tests based on the BK-1 satellite further validate the importance of the 1610 nm band in the polarimetric sensor for greenhouse gas retrievals (Fig. 12).

In response to the strategic demands of global carbon accounting and China’s dual-carbon goals, there is an urgent need to develop high-precision retrieval algorithms for aerosols and greenhouse gases (GHGs). Addressing the significant influence of aerosols on GHG retrieval processes, we conduct a collaborative simulation of aerosol and GHG satellite remote sensing based on a full-physics algorithm. Utilizing observation modes of multispectral polarization sensors and hyperspectral instruments, the LINTRAN vector radiative transfer model is employed to simulate multispectral single-angle polarization and radiance measurements under various scenarios, followed by aerosol-synchronized correction for GHG retrieval. Based on these simulations, we analyze trends in aerosol parameters and major GHG parameters across different payloads and spectral band combinations, leading to the following conclusions. When retrieving aerosol parameters using polarized payloads, the addition of SWIR vector information at 1610 nm and 2250 nm significantly enhances the retrieval of AOD, with notable improvements for coarse-mode aerosol variations. In the joint retrieval of aerosol and greenhouse gas parameters based on polarization and hyperspectral data, aerosol parameter retrievals exhibit high stability. The incorporation of hyperspectral bands effectively reduces the variability of coarse-mode aerosols, while the addition of SWIR bands improves retrieval accuracy for XCH₄ and XCO₂. Comparing retrieval results across three aerosol modes reveals that mode 1 (fine-mode aerosols) achieves better retrieval accuracy, while mode 2 (dust-type aerosols) proves more challenging, resulting in lower precision. The analysis of the influence of XCH₄ and XCO₂ variability across different aerosol modes indicates that fine-mode aerosols tend to cluster at lower values, while the density distribution of coarse-mode aerosols is relatively uniform. As the scattering angle increases, errors in XCH₄ and XCO₂ remain within the range of -0.2% to 0. Validation using band retrievals from the BK-1 satellite further confirms the importance of the 1610 nm band for improving GHG retrieval accuracy. In summary, future research should incorporate multi-angle, multi-band polarized observations to enhance GHG retrieval capabilities, thus providing more precise scientific support for achieving China’s carbon peak and carbon neutrality objectives.