Search Results (74)

Satellite-based monitoring of atmospheric column-averaged dry-air mole fraction (XCH₄) is essential for quantifying methane (CH₄) emissions, yet uncharacterized spatially varying biases in XCH₄ observations can cause misattribution in flux estimates. This study assesses the potential of the upcoming TanSat-2 satellite mission to estimate China’s CH₄ emission using a series of Observing System Simulation Experiments (OSSEs) based on an Ensemble Kalman Filter (EnKF) inversion framework coupled with GEOS-Chem on a 0.5° × 0.625° grid, alongside an evaluation of current TROPOMI-based products against Total Carbon Column Observing Network (TCCON) observations. Assuming a target precision of 8 ppb, TanSat-2 could achieve an annual national emission estimate accuracy of 2.9% ± 4.2%, reducing prior uncertainty by 84%, with regional deviations below 5.0% across Northeast, Central, East, and Southwest China. In contrast, limited coverage in South China due to persistent cloud cover leads to a 26.1% discrepancy—also evident in pseudo TROPOMI OSSEs—highlighting the need for complementary ground-based monitoring strategies. Sensitivity analyses show that satellite retrieval biases strongly affect inversion robustness, reducing the accuracy in China’s total emission estimates by 5.8% for every 1 ppb increase in bias level across scenarios, particularly in Northeast, Central and East China. We recommend expanding ground-based XCH₄ observations in these regions to support the correction of satellite-derived biases and improve the reliability of satellite-constrained inversion results. Full article

Atmospheric methane (CH₄) concentrations have increased to 2.5 times their pre-industrial levels, with a marked acceleration in recent decades. CH₄ is responsible for approximately 30% of the global temperature rise since the Industrial Revolution. This growing concentration contributes to environmental degradation, including ocean acidification, accelerated climate change, and a rise in natural disasters. The column-averaged dry-air mole fraction of methane (XCH₄) is a crucial indicator for assessing atmospheric CH₄ levels. In this study, the Sentinel-5P TROPOMI instrument was employed to monitor, map, and estimate CH₄ concentrations on both regional and global scales. However, TROPOMI data exhibits limitations such as spatial gaps and relatively coarse resolution, particularly at regional scales or over small areas. To mitigate these limitations, a novel Convolutional Neural Network Autoencoder (CNN-AE) model was developed. Validation was performed using the Total Carbon Column Observing Network (TCCON), providing a benchmark for evaluating the accuracy of various interpolation and prediction models. The CNN-AE model demonstrated the highest accuracy in regional-scale analysis, achieving a Mean Absolute Error (MAE) of 28.48 ppb and a Root Mean Square Error (RMSE) of 30.07 ppb. This was followed by the Random Forest (RF) regressor (MAE: 29.07 ppb; RMSE: 36.89 ppb), GridData Nearest Neighbor Interpolator (NNI) (MAE: 30.06 ppb; RMSE: 32.14 ppb), and the Radial Basis Function (RBF) Interpolator (MAE: 80.23 ppb; RMSE: 90.54 ppb). On a global scale, the CNN-AE again outperformed other methods, yielding the lowest MAE and RMSE (19.78 and 24.7 ppb, respectively), followed by RF (21.46 and 27.23 ppb), GridData NNI (25.3 and 32.62 ppb), and RBF (43.08 and 54.93 ppb). Full article

The 2020 wildfire season in the Western U.S. was historic in its intensity and impact on the land and atmosphere. This study aims to characterize satellite retrievals of carbon monoxide (CO), a tracer of combustion and signature of those fires, from two key satellite instruments: the Cross-track Infrared Sounder (CrIS) and the Tropospheric Monitoring Instrument (TROPOMI). We evaluate them during this event and assess their synergies. These two retrievals are matched temporally, as the host satellites are in tandem orbit and spatially by aggregating TROPOMI to the CrIS resolution. Both instruments show that the Western U.S. displayed significantly higher daily average CO columns compared to the Central and Eastern U.S. during the wildfires. TROPOMI showed up to a factor of two larger daily averages than CrIS during the most intense fire period, likely due to differences in the vertical sensitivity of the two instruments and representative of near-surface CO abundance near the fires. On the other hand, there was excellent agreement between the instruments in downwind free tropospheric plumes (scatter plot slopes of 0.96–0.99), consistent with their vertical sensitivities and indicative of mostly lofted smoke. Temporally, TROPOMI CO column peaks were delayed relative to the Fire Radiative Power (FRP), and CrIS peaks were delayed with respect to TROPOMI, particularly during the intense initial weeks of September, suggesting boundary layer buildup and ventilation. Satellite retrievals were evaluated using ground-based CO column estimates from the Network for the Detection of Atmospheric Composition Change (NDACC) and the Total Carbon Column Observing Network (TCCON), showing Normalized Mean Errors (NMEs) for CrIS and TROPOMI below 32% and 24%, respectively, when compared to all stations studied. While Normalized Mean Bias (NMB) was typically low (absolute value below 15%), there were larger negative biases at Pasadena, likely associated with sharp spatial gradients due to topography and proximity to a large city, which is consistent with previous research. In situ CO profiles from AirCore showed an elevated smoke plume for 15 September 2020, highlighted consistency between TROPOMI and CrIS CO columns for lofted plumes. This study demonstrates that both CrIS and TROPOMI provide complementary information on CO distribution. CrIS’s sensitivity in the middle and lower free troposphere, coupled with TROPOMI’s effectiveness at capturing total columns, offers a more comprehensive view of CO distribution during the wildfires than either retrieval alone. By combining data from both satellites as a ratio, more detailed information about the vertical location of the plumes can potentially be extracted. This approach can enhance air quality models, improve vertical estimation accuracy, and establish a new method for assessing lower tropospheric CO concentrations during significant wildfire events. Full article

Global warming caused by the increase in the atmospheric CO₂ content has become a focal environmental issue of common concern to the international community. As a key resource support for achieving the “dual carbon” goals in Western China, Sichuan Province requires a deep analysis of its carbon sources, carbon sinks, and its characteristics in terms of atmospheric environmental capacity, which is of great significance for formulating effective regional sustainable development strategies and responding to global climate change. In view of the unique geographical and climatic conditions in Sichuan Province and the current situation of a low and uneven distribution of atmospheric environmental capacity, this paper uses three forms of multi-source satellite data, OCO-2, OCO-3, and GOSAT, combined with other auxiliary data, to generate a daily XCO₂ concentration dataset with a spatial resolution of a 1km grid in Sichuan Province from 2015 to 2022. Based on the Optuna optimization method with 10-fold cross-validation, the optimal hyperparameter configuration of the four base learners of Stacking, random forest, gradient boosting decision tree, extreme gradient boosting, and the K nearest neighbor algorithm is searched for; finally, the logistic regression algorithm is used as the second-layer meta-learner to effectively improve the prediction accuracy and generalization ability of the Stacking ensemble learning model. According to the comparison of the performance of each model by cross-validation and TCCON site verification, the Stacking model significantly improved in accuracy, with an R², RMSE, and MAE of 0.983, 0.87 ppm and 0.19 ppm, respectively, which is better than those of traditional models such as RF, KNN, XGBoost, and GBRT. The accuracy verification of the atmospheric XCO₂ data estimated by the model based on the observation data of the two TCCON stations in Xianghe and Hefei showed that the correlation coefficients were 0.96 and 0.98, and the MAEs were 0.657 ppm and 0.639 ppm, respectively, further verifying the high accuracy and reliability of the model. At the same time, the fusion of multi-source satellite data significantly improved the spatial coverage of XCO₂ concentration data in Sichuan Province, effectively filling the gap in single satellite observation data. Based on the reconstructed XCO₂ dataset of Sichuan Province, the study revealed that there are significant regional and seasonal differences in the XCO₂ concentrations in the region, showing seasonal variation characteristics of being higher in spring and winter and lower in summer and autumn; in terms of the spatial distribution, the overall spatial distribution characteristics are high in the east and low in the west. This study helps to deepen our understanding of the carbon cycle and climate change, and can provide a scientific basis and risk assessment methods for policy formulation, effect evaluation, and international cooperation. Full article

Accurate retrieval of column-averaged dry-air mole fraction of methane (XCH₄) in the atmosphere is important for greenhouse gas emission management. Traditional XCH₄ retrieval methods are complex, while machine learning can be used to model nonlinear relationships by analyzing large datasets, providing an efficient alternative. This study proposes an XGBoost algorithm-based retrieval method to improve the efficiency of atmospheric XCH₄ retrieval. First, the key wavelengths affecting XCH₄ retrieval were determined using a radiative transfer model. The TROPOspheric Monitoring Instrument (TROPOMI) L1B satellite data, L2 XCH₄ products, and auxiliary data were matched to construct the dataset. The dataset constructed was used to train the XGBoost model and obtain the TRO_XGB_XCH₄ model. Finally, the accuracy of the proposed model was evaluated using various parameter values and validated against XCH₄ products and Total Carbon Column Observing Network (TCCON) ground-based observations. The results showed that the proposed TRO_XGB_XCH₄ model had a tenfold cross-validation accuracy R of 0.978, a ground-based validation R of 0.749, and a temporal extension accuracy R of 0.863. Therefore, the accuracy of the TRO_XGB_XCH₄ retrieval model is comparable to that of the official TROPOMI L2 product. Full article

Accurately assessing the impact of anthropogenic carbon dioxide (CO₂) emissions on CO₂ concentrations is essential for understanding regional climate change, particularly in high-emission countries like China. This study employed the GEOS-Chem chemical transport model to simulate and compare the spatiotemporal distributions of XCO₂ of three anthropogenic CO₂ emission inventories in mainland China for the 2018–2020 period and analyzed the effects of emission variations on atmospheric CO₂ concentrations. In eastern China, particularly in the Yangtze River Delta (YRD) and Beijing-Tianjin-Hebei (BTH) regions, column-averaged dry air mole fractions of CO₂ (XCO₂) can exceed 420 ppm during peak periods, with emissions from these areas contributing significantly to the national total. The simulation results were validated by comparing them with OCO-2 satellite observations and ground-based monitoring data, showing that more than 70% of the monitoring stations exhibited a correlation coefficient greater than 0.7 between simulated and observed data. The average bias relative to satellite observations was less than 1 ppm, with the Emissions Database for Global Atmospheric Research (EDGAR) showing the highest degree of agreement with both satellite and ground-based observations. During the study period, anthropogenic CO₂ emissions resulted in an increase in XCO₂ exceeding 10 ppm, particularly in the North China Plain and the YRD. In scenarios where emissions from either the BTH or YRD regions were reduced by 50%, a corresponding decrease of 1 ppm in XCO₂ was observed in the study area and its surrounding regions. These findings underscore the critical role of emission control policies in mitigating the rise in atmospheric CO₂ concentrations in densely populated and industrialized areas. This research elucidates the impacts of variations in anthropogenic emissions on the spatiotemporal distribution of atmospheric CO₂ and emphasizes the need for improved accuracy of CO₂ emission inventories. Full article

As carbon dioxide (CO₂) concentrations continue to rise, climate change, characterized by global warming, presents a significant challenge to global sustainable development. Currently, most global shortwave infrared CO₂ retrievals rely on fully physical retrieval algorithms, for which complex calculations are necessary. This paper proposes a method to predict the concentration of column-averaged CO₂ (XCO₂) from shortwave infrared hyperspectral satellite data, using machine learning to avoid the iterative computations of the physical method. The training dataset is constructed using the Orbiting Carbon Observatory-2 (OCO-2) spectral data, XCO₂ retrievals from OCO-2, surface albedo data, and aerosol optical depth (AOD) measurements for 2019. This study employed a variety of machine learning algorithms, including Random Forest, XGBoost, and LightGBM, for the analysis. The results showed that Random Forest outperforms the other models, achieving a correlation of 0.933 with satellite products, a mean absolute error (MAE) of 0.713 ppm, and a root mean square error (RMSE) of 1.147 ppm. This model was then applied to retrieve CO₂ column concentrations for 2020. The results showed a correlation of 0.760 with Total Carbon Column Observing Network (TCCON) measurements, which is higher than the correlation of 0.739 with satellite product data, verifying the effectiveness of the retrieval method. Full article

The COllaborative Carbon Column Observing Network has become a reliable source of high-quality ground-based remote sensing network data that provide column-averaged dry-air mole fractions of carbon dioxide (XCO₂), methane (XCH₄), and carbon monoxide (XCO). The fiducial reference measurements of these gases from the COCCON complement the TCCON and NDACC-IRWG data. This study shows the application of COCCON data for the validation of existing greenhouse gas satellite products. This study includes the validation of XCH₄ and XCO products from the European Copernicus Sentinel-5 Precursor (S5P) mission, XCO₂ products from the American Orbiting Carbon Observatory-2 (OCO-2) mission, and XCO₂ and XCH₄ products from the Japanese Greenhouse gases Observing SATellite (GOSAT). A total of 27 datasets contributed to this study; some of these were collected in the framework of campaign activities and covered only a short time period. In addition, several permanent stations provided long-term observations. The random uncertainties in the validation results, specifically for S5P with a lot of coincidences pairs, are found to be similar to the comparison with the TCCON. The comparison results of OCO-2 land nadir and land glint observation modes to the COCCON on a global scale, despite limited coincidences, are very promising. The stations can, therefore, expand on the coverage of the already existing ground-based reference remote sensing sites from the TCCON and the NDACC network. The COCCON data can be used for future satellite and model validation studies and carbon cycle studies. Full article

A good quantitative knowledge of regional sources and sinks of atmospheric carbon dioxide (CO₂) is essential for understanding the global carbon cycle. It is also a key prerequisite for elaborating cost-effective national strategies to achieve the goals of the Paris Agreement. However, available estimates of CO₂ fluxes for many regions of the world remain uncertain, despite significant recent progress in the remote sensing of terrestrial vegetation and atmospheric CO₂. In this study, we investigate the feasibility of inferring reliable regional estimates of the net ecosystem exchange (NEE) using column-averaged dry-air mole fractions of CO₂ (XCO₂) retrieved from Orbiting Carbon Observatory-2 (OCO-2) observations as constraints on parameters of the widely used Vegetation Photosynthesis and Respiration model (VPRM), which predicts ecosystem fluxes based on vegetation indices derived from multispectral satellite imagery. We developed a regional-scale inverse modeling system that applies a Bayesian variational optimization algorithm to optimize parameters of VPRM coupled to the CHIMERE chemistry transport model and which involves a preliminary transformation of the input XCO₂ data that reduces the impact of the CHIMERE boundary conditions on inversion results. We investigated the potential of our inversion system by applying it to a European region (that includes, in particular, the EU countries and the UK) for the warm season (May–September) of 2021. The inversion of the OCO-2 observations resulted in a major (more than threefold) reduction of the prior uncertainty in the regional NEE estimate. The posterior NEE estimate agrees with independent estimates provided by the CarbonTracker Europe High-Resolution (CTE-HR) system and the ensemble of the v10 OCO-2 model intercomparison (MIP) global inversions. We also found that the inversion improves the agreement of our simulations of XCO₂ with retrievals from the Total Carbon Column Observing Network (TCCON). Our sensitivity test experiments using synthetic XCO₂ data indicate that the posterior NEE estimate would remain reliable even if the actual regional CO₂ fluxes drastically differed from their prior values. Furthermore, the posterior NEE estimate is found to be robust to strong biases and random uncertainties in the CHIMERE boundary conditions. Overall, this study suggests that our approach offers a reliable and relatively simple way to derive robust estimates of CO₂ ecosystem fluxes from satellite XCO₂ observations while enhancing the applicability of VPRM in regions where eddy covariance measurements of CO₂ fluxes are scarce. Full article

Carbon dioxide (CO₂) is a key driver of global climate change. Since the Industrial Revolution, the rapid rise in atmospheric CO₂ levels has significantly intensified global warming and climate-related issues. To accurately and promptly monitor changes in CO₂ concentrations and to support the development of climate policies, this study proposes a method based on random forest models to generate a continuous monthly dataset of CO₂ column concentration (XCO₂) across the entire Chinese region from 2004 to 2023. The study integrates XCO₂ satellite observations from SCIAMACHY, GOSAT, OCO-2, and GF-5B, alongside nighttime light remote sensing data, meteorological parameters, vegetation indices, and CO₂ profile data. Using the random forest algorithm, a complex relationship model was established between XCO₂ concentrations and various environmental variables. The goal of this model is to provide XCO₂ estimates with enhanced spatial coverage and accuracy. The XCO₂ concentrations predicted by the model show a high level of consistency with satellite observations, achieving a correlation coefficient (R-value) of 0.9959 and a root mean square error (RMSE) of 1.1631 ppm. This indicates that the model offers strong predictive accuracy and generalization ability. Additionally, ground-based validation further confirmed the model’s effectiveness, with a correlation coefficient (R-value) of 0.956 when compared with TCCON site observation data. Full article

This paper presents the results of the integration, commissioning, and in-orbit calibration of the ARGUS 2000 SWIR wavelength spectrometer onboard the Colombian FACSAT-2 satellite, Colombia’s first satellite used for the measurement of greenhouse gases (GHG). The satellite has been in orbit for approximately one year, gathering spectral signatures in order to characterize the data and perform calibration. The calibration represents a certain behavior following a second-order adjustment. For data analysis, retrieval algorithms have been developed to obtain synthetic spectra using the Genspect 1.2 software code. These synthetic spectra were obtained from the spectroscopic data associated with atmospheric gases (H₂O, O₂, CO, and CO₂), which are stored in the HITRAN database. Attempting to achieve a more accurate simulation of the experimental spectrum, certain instrumental parameters have been incorporated into the synthetic spectrum, including the resolution of the spectrometer, the field of view (FOV) angle of observation, the limited quantum efficiency of detection, and the slit function. As a result, six slit functions were tested, with the Gaussian and the diffraction functions proving to be the most effective. Finally, a profile of Total Carbon Column Observing Network (TCCON) concentrations was used for comparison with a spectral signature acquired by FACSAT-2, resulting in a close match between the synthetic spectrum and the measured spectrum. Full article

The Total Carbon Column Observing Network (TCCON) and the Infrared Working Group of the Network for the Detection of Atmospheric Composition Change (NDACC-IRWG) are two ground-based networks that provide the retrieved concentrations of up to 30 atmospheric trace gases, using solar absorption spectrometry. Both networks provide reference measurements for the validation of satellites and models. TCCON concentrates on long-lived greenhouse gases (GHGs) for carbon cycle studies and validation. The number of sites is limited, and the geographical coverage is uneven, covering mainly Europe and the USA. A better distribution of stations is desired to improve the representativeness of the data for various atmospheric conditions and surface conditions and to cover a large latitudinal distribution. The two successive Fiducial Reference Measurements for Greenhouse Gases European Space Agency projects (FRM4GHG and FRM4GHG2) aim at the assessment of several low-cost portable instruments for precise measurements of GHGs to complement the existing ground-based sites. Several types of low spectral resolution Fourier transform infrared (FTIR) spectrometers manufactured by Bruker, namely an EM27/SUN, a Vertex70, a fiber-coupled IRCube, and a Laser Heterodyne spectro-Radiometer (LHR) developed by UK Rutherford Appleton Laboratory are the participating instruments to achieve the Fiducial Reference Measurements (FRMs) status. Intensive side-by-side measurements were performed using all four instruments next to the Bruker IFS 125HR high spectral resolution FTIR, performing measurements in the NIR (TCCON configuration) and MIR (NDACC configuration) spectral range. The remote sensing measurements were complemented by AirCore launches, which provided in situ vertical profiles of target gases traceable to the World Meteorological Organization (WMO) reference scale. The results of the intercomparisons are shown and discussed. Except for the EM27/SUN, all other instruments, including the reference TCCON spectrometer, needed modifications during the campaign period. The EM27/SUN and the Vertex70 provided stable and precise measurements of the target gases during the campaign with quantified small biases. As part of the FRM4GHG project, one EM27/SUN is now used as a travel standard for the verification of column-integrated GHG measurements. The extension of the Vertex70 to the MIR provides the opportunity to retrieve additional concentrations of N₂O, CH₄, HCHO, and OCS. These MIR data products are comparable to the retrieval results from the high-resolution IFS 125HR spectrometer as operated by the NDACC. Our studies show the potential for such types of spectrometers to be used as a travel standard for the MIR species. An enclosure system with a compact solar tracker and meteorological station has been developed to house the low spectral resolution portable FTIR systems for performing solar absorption measurements. This helps the spectrometers to be mobile and enables autonomous operation, which will help to complement the TCCON and NDACC networks by extending the observational capabilities at new sites for the observation of GHGs and additional air quality gases. The development of the retrieval software allows comparable processing of the Vertex70 type of spectra as the EM27/SUN ones, therefore bringing them under the umbrella of the COllaborative Carbon Column Observing Network (COCCON). A self-assessment following the CEOS-FRM Maturity Matrix shows that the COCCON is able to provide GHG data products of FRM quality and can be used for either short-term campaigns or long-term measurements to complement the high-resolution FTIR networks. Full article

CO₂ is one of the primary greenhouse gases impacting global climate change, making it crucial to understand the spatiotemporal variations of CO₂. Currently, commonly used satellites serve as the primary means of CO₂ observation, but they often suffer from striping issues and fail to achieve complete coverage. This paper proposes a method for constructing a comprehensive high-spatiotemporal-resolution XCO₂ dataset based on multiple auxiliary data sources and satellite observations, utilizing multiple simple deep neural network (DNN) models. Global validation results against ground-based TCCON data demonstrate the excellent accuracy of the constructed XCO₂ dataset (R is 0.94, RMSE is 0.98 ppm). Using this method, we analyze the spatiotemporal variations of CO₂ in China and its surroundings (region: 0°–60° N, 70°–140° E) from 2019 to 2020. The gapless and fine-scale CO₂ generation method enhances people’s understanding of CO₂ spatiotemporal variations, supporting carbon-related research. Full article

The Arctic and boreal regions are experiencing a rapid increase in temperature, resulting in a changing cryosphere, increasing human activity, and potentially increasing high-latitude methane emissions. Satellite observations from Sentinel-5P TROPOMI provide an unprecedented coverage of a column-averaged dry-air mole fraction of methane (XCH₄) in the Arctic, compared to previous missions or in situ measurements. The purpose of this study is to support and enhance the data used for high-latitude research through presenting a systematic evaluation of TROPOMI methane products derived from two different processing algorithms: the operational product (OPER) and the scientific product (WFMD), including the comparison of recent version changes of the products (OPER, OPER rpro, WFMD v1.2, and WFMD v1.8). One finding is that OPER rpro yields lower XCH₄ than WFMD v1.8, the difference increasing towards the highest latitudes. TROPOMI product differences were evaluated with respect to ground-based high-latitude references, including four Fourier Transform Spectrometer in the Total Carbon Column Observing Network (TCCON) and five EM27/SUN instruments in the Collaborative Carbon Column Observing Network (COCCON). The mean TROPOMI–TCCON GGG2020 daily median XCH₄ difference was site-dependent and varied for OPER rpro from −0.47 ppb to 22.4 ppb, and for WFMD v1.8 from 1.2 ppb to 19.4 ppb with standard deviations between 13.0 and 20.4 ppb and 12.5–15.0 ppb, respectively. The TROPOMI–COCCON daily median XCH₄ difference varied from −26.5 ppb to 5.6 ppb for OPER rpro, with a standard deviation of 14.0–28.7 ppb, and from −5.0 ppb to 17.2 ppb for WFMD v1.8, with a standard deviation of 11.5–13.0 ppb. Although the accuracy and precision of both TROPOMI products are, on average, good compared to the TCCON and COCCON, a persistent seasonal bias in TROPOMI XCH₄ (high values in spring; low values in autumn) is found for OPER rpro and is reflected in the higher standard deviation values. A systematic decrease of about 7 ppb was found between TCCON GGG2014 and GGG2020 product update highlighting the importance of also ensuring the reliability of ground-based retrievals. Comparisons to atmospheric profile measurements with AirCore carried out in Sodankylä, Northern Finland, resulted in XCH₄ differences comparable to or smaller than those from ground-based remote sensing. Full article

Although atmospheric CO₂ concentrations collected by satellites play a crucial role in understanding global greenhouse gases, the sparse geographic distribution greatly affects their widespread application. In this paper, a hybrid CNN and spatiotemporal Kriging (CNN-STK) model is proposed to generate a monthly spatiotemporal continuous XCO2 dataset over China at 0.25° grid-scale from 2015 to 2020, utilizing OCO-2 XCO2 and geographic covariates. The validations against observation samples, CAMS XCO2 and TCCON measurements indicate the CNN-STK model is effective, robust, and reliable with high accuracy (validation set metrics: R² = 0.936, RMSE = 1.3 ppm, MAE = 0.946 ppm; compared with TCCON: R² = 0.954, RMSE = 0.898 ppm and MAE = 0.741 ppm). The accuracy of CNN-STK XCO2 exhibits spatial inhomogeneity, with higher accuracy in northern China during spring, autumn, and winter and lower accuracy in northeast China during summer. XCO2 in low-value-clustering areas is notably influenced by biological activities. Moreover, relatively high uncertainties are observed in the Qinghai-Tibet Plateau and Sichuan Basin. This study innovatively integrates deep learning with the geostatistical method, providing a stable and cost-effective approach for other countries and regions to obtain regional scales of atmospheric CO₂ concentrations, thereby supporting policy formulation and actions to address climate change. Full article