Atmospheric Research最新文献

{"title":"Evaluation of the near-surface climatology of centennial-scale reanalyses for Greenland","authors":"Yaxiao Ma ,&nbsp;Wuying Zhang","doi":"10.1016/j.atmosres.2026.108793","DOIUrl":"10.1016/j.atmosres.2026.108793","url":null,"abstract":"<div><div>Understanding the past climate variability of Greenland is vital for contextualizing its current and future changes, a purpose for which centennial-scale reanalyses provide critical support. However, the performance of centennial-scale reanalyses over Greenland remains an open question, as previous assessments have primarily focused on the post-1979 period. Therefore, we evaluated the performance on different climate variables of the Twentieth Century Reanalysis version 3 (20CR v3; 1806–2015) and the European Centre for Medium Range Weather Forecasts fifth generation reanalysis (ERA5; 1940–2015) over Greenland, based on in-situ observations and ice core records. For 2-m temperature (T_2m), both reanalyses exhibited small cold biases in summer, though with a failure to capture variation patterns during this season. 20CR v3 showed dry biases in summer and wet biases in winter, while ERA5 consistently overestimated precipitation among all the seasons. Significant dry bias (&lt;−2.5 mm) of 20CR v3 was found at southeastern coastal sites. We concluded that the elevation bias induced by the coarse resolution of 20CR v3 was one of the primary reasons for its cold and dry biases along the margin of the ice sheet. Notably, unlike ERA5, 20CR v3 failed to reproduce the observed increasing precipitation trend from 1940 to 2015 (−1.9 mm/dec vs 4.9 mm/dec, <em>p</em> &gt; 0.05). Although 20CR v3 performed slightly inferior than ERA5, their consistent patterns, coupled with 20CR v3's robust performance, support the reliability of 20CR v3 for centennial-scale climate studies in Greenland.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"335 ","pages":"Article 108793"},"PeriodicalIF":4.4,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146006415","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Source apportionment and health risk assessment of PM10-bound organic aerosols in urban South Korea: Insights from n-alkanes, PAHs, hopanes, and steranes","authors":"Md. Mozammel Haque ,&nbsp;Suresh K.R. Boreddy ,&nbsp;Jinsang Jung ,&nbsp;Kimitaka Kawamura","doi":"10.1016/j.atmosres.2026.108791","DOIUrl":"10.1016/j.atmosres.2026.108791","url":null,"abstract":"<div><div>Aerosols, one of the principal atmospheric pollutants, are receiving significant attention from the scientific community due to their impact on human health and climate. Despite the carcinogenicity of PAHs, the contributions of their sources to the urban atmosphere remain unclear. The present study examines the molecular distribution and source indices of various non-polar organic compounds, including n-alkanes, polycyclic aromatic hydrocarbons (PAHs), hopanes, and steranes, in PM₁₀ samples collected in Daejeon, South Korea, from 15 April to 22 May 2013 using the thermal desorption gas chromatography/mass spectrometry (TD-GC/MS). The molecular distribution of n-alkanes was characterized by the predominance of C₂₉, followed by C₂₇. The total concentrations of n-alkanes exhibited strong positive correlations with elemental carbon (EC) and potassium ion (K⁺), indicating a significant contribution from fossil fuel combustion and biomass-burning sources. Among the PAHs, indeno[1,2,3-<em>cd</em>]pyrene (IcdP) and dibenzo[<em>a</em>,<em>h</em>]anthracene (DahA) were the most prevalent compounds. The dominant hopanoid hydrocarbon was 17a(H),21b(H)-hopane (30ab), while 17b(H),21a(H)-hopane (30ba) was least abundant. During the sampling period, sterane compounds exhibited a higher concentration of C₂₇ compared to C₂₉. Based on the diagnostic ratios and source-specific markers, the present study revealed that coal combustion, biomass burning, and vehicular emissions were predominant sources of PAHs during spring. Positive matrix factorisation (PMF) analysis showed that in Daejeon, total PAHs were mainly from biomass and coal combustion (74%), followed by vehicle emissions (14%) and natural gas combustion (12%). Further, an inverse relationship has been found between PAH concentrations and air temperature; i.e., higher levels of PAHs correspond to lower temperatures, which could have significant implications for air quality and public health. The findings of this study offer valuable insights into the sources of organic compounds, which are highly carcinogenic, and can inform policymakers in developing effective air quality monitoring and mitigation strategies.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"335 ","pages":"Article 108791"},"PeriodicalIF":4.4,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Influence of aerosols and clouds on diffuse radiation under complex sky conditions based on MFRSR","authors":"Huaping Li ,&nbsp;Yaxin Wang ,&nbsp;Ming Zhang ,&nbsp;Lunche Wang ,&nbsp;Chen Yuan","doi":"10.1016/j.atmosres.2026.108808","DOIUrl":"10.1016/j.atmosres.2026.108808","url":null,"abstract":"<div><div>Atmospheric components are critical for maintaining the radiative balance of the Earth–atmosphere system. Based on long-term observations from the Multi-Filter Rotating Shadowband Radiometer (MFRSR), this study classified clear-sky, thin cloud, and thick cloud conditions and analyzed the effects of aerosols and clouds on diffuse radiation (DR) at the Southern Great Plains (SGP) site under these different sky conditions. The results show that under clear skies, the interannual variability of DR is highly consistent with that of aerosol optical depth (AOD) (<em>r</em> = 0.88), while under cloudy conditions DR is more closely associated with cloud properties such as cloud optical depth (COD) and cloud fraction (CF). Specifically, COD exhibits a positive correlation with DR under thin clouds but a negative correlation under thick clouds. Radiative effect analysis indicates that the aerosol diffuse radiative effect (ADRE) reaches 53 W m⁻² under clear skies, decreases to 34 W m⁻² under thin clouds, and further shifts to −7 W m⁻² under thick clouds, reflecting the nonlinear response of aerosol forcing to different cloud backgrounds. With increasing COD, DR increases initially, then decreases, and eventually stabilizes. The cloud diffuse radiative effect (CDRE) decreases from 113 W m⁻² under thin clouds to 82 W m⁻² under thick clouds, revealing the shielding effect of cloud thickening on shortwave diffuse radiation. The diffuse radiative effect of water vapor (WVDRE) increases with cloud thickness, whereas that of the ozone diffuse radiative effect (ODRE) remains relatively small and stable. Feature importance analysis further shows that under clear skies, AOD accounts for the largest contribution (64%) to DR variability, with DR gradually increasing as AOD increases, followed by single scattering albedo (SSA) and the Ångström exponent (AE), while under cloudy conditions COD and CF dominate. This study highlights the regulatory role of atmospheric components in DR and provides valuable physical references for improving radiation parameterization and accurately assessing surface diffuse radiation under varying sky conditions.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"335 ","pages":"Article 108808"},"PeriodicalIF":4.4,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Uncertainty decomposition and quantification of seasonal precipitation forecasting based on Bayesian neural networks","authors":"Enzo Pinheiro,&nbsp;Taha B.M.J. Ouarda","doi":"10.1016/j.atmosres.2026.108815","DOIUrl":"10.1016/j.atmosres.2026.108815","url":null,"abstract":"<div><div>Seasonal climate forecasts are inherently uncertain due to the chaotic nature of the climate system and the limitations of predictive models. Quantifying and decomposing this uncertainty are essential for reliable forecasts. In this study, we investigate machine learning-based seasonal forecasting (MLSF) using a Bayesian adaptation of TelNet, a neural network model for precipitation prediction. We reformulate TelNet as a Bayesian Neural Network trained via variational inference, with alternative implementations based on Monte Carlo (MC) dropout. Both continuous (ensemble-based) and categorical (tercile-based) prediction strategies are assessed. Training is conducted with ERA5 reanalysis and historical simulations from 18 CMIP6 models, followed by validation and testing with ERA5 data over South America. We decompose predictive uncertainty into ML-aleatoric (data-related) and ML-epistemic (model-related) components. Results show that ML-aleatoric uncertainty dominates, being on average two orders of magnitude larger than ML-epistemic uncertainty. ERA5-trained models exhibit lower aleatoric uncertainty but are often overconfident and not well-calibrated, whereas CMIP6-trained models show higher aleatoric uncertainty but improved calibration. Models trained with continuous targets achieve lower uncertainty than categorical models, which lose information through discretization. Probabilistic accuracy, evaluated via Ranked Probability Score (RPS), indicates that CMIP6-based models trained with continuous targets and Continuous Ranked Probability Score (CRPS) loss perform best overall. Seasonal analysis reveals better calibration in austral summer (DJF) compared to winter (JJA), consistent with stronger teleconnection signals during summer. Finally, case studies of El Niño, La Niña, and neutral years demonstrate that forecast confidence and uncertainty strongly depend on the training dataset and target representation.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"335 ","pages":"Article 108815"},"PeriodicalIF":4.4,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The evolution of cloud-to-ground lightning and radar reflectivity in mesoscale convective systems over Central-Eastern China","authors":"Yanan Fu ,&nbsp;Jianhua Sun ,&nbsp;Yuanchun Zhang ,&nbsp;Shenming Fu ,&nbsp;Xiaofang Wang ,&nbsp;Chunguang Cui","doi":"10.1016/j.atmosres.2026.108818","DOIUrl":"10.1016/j.atmosres.2026.108818","url":null,"abstract":"<div><div>By combining geostationary satellite observations, cloud-to-ground (CG) lightning detections, and radar composite reflectivity (CR) mosaics, this study investigates the intensity, evolution, and distribution of CG lightning and CR associated with mesoscale convective systems (MCSs). A total of 551 MCSs are identified in central-eastern China during May–September of 2018–2021 using Fengyun-4 A satellite brightness temperature (BT) data and subsequently matched with CG lightning and CR mosaics. The geographical analysis of CG lightning associated with MCSs reveals that the highest occurrences are most frequent in the transition zone between the Poyang Lake Plain and the surrounding mountains. The proportion of negative-polarity flashes is relatively lower in MCS-related CG lightning than in total CG lightning. An overall negative correlation between BT and the peak current of CG lightning is observed with seasonal variations. The positive correlations detected in May and September may be attributed to the advection of ice particles from convective cores into adjacent stratiform regions, driven by stronger mid-to-upper-level winds. The peak occurrence of CG lightning precedes the maximum MCS extent by 0–2 h. As the MCS develops, the proportion of convective clouds decreases, while that of non-precipitating anvil increases, and the stratiform regions consistently account for 50%–60% of the total MCS extent, dominating throughout its lifespan. The stratiform region is predominantly located in the southern part of the MCS, whereas convective clouds are mainly situated in the northern part, possibly influenced by the Meiyu front.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"335 ","pages":"Article 108818"},"PeriodicalIF":4.4,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Nighttime cloud cover from multi-site observations and its relationships with ground- and satellite-based measurements","authors":"Nofel Lagrosas ,&nbsp;Sheng-Hsiang Wang ,&nbsp;Bing-Qian Wu ,&nbsp;Neng-Huei Lin ,&nbsp;Yotsumi Yoshii ,&nbsp;Masanori Yabuki ,&nbsp;Moe Ishii ,&nbsp;Kein Itou ,&nbsp;Tatsuo Shiina ,&nbsp;Hitoshi Irie ,&nbsp;Hiroaki Kuze ,&nbsp;Atsushi Shimizu","doi":"10.1016/j.atmosres.2026.108823","DOIUrl":"10.1016/j.atmosres.2026.108823","url":null,"abstract":"<div><div>Reliable characterization of nighttime cloud cover (CC) remains challenging due to the limited availability of long-term ground-based observations and the lack of coordinated multi-instrument measurements. This study presents a coordinated nighttime cloud observation framework that integrates standardized ground-based camera systems with collocated meteorological measurements, lidar-derived cloud base height (CBH), pyrgeometer-based net longwave radiation, and Himawari satellite’s CC across four sites in East Asia — Chiba, Toyama, Shiga (Japan), and Taoyuan (Taiwan). A continuous six-year nighttime CC record is available at the Chiba site that can be used to assess of seasonal variability and cloud-environment relationships. Unlike previous studies that rely on single instruments or isolated locations, this work systematically analyzes nighttime CC in relation to near-surface temperature, relative humidity (RH), cloud base height, and longwave radiation using a unified methodology. In persistently humid conditions such as Taoyuan, enhanced atmospheric water vapor suppresses the increase of CC with near-ground temperature. At the same site, net longwave radiation increases by 1.778 W m⁻² per 1% increase in CC. In Chiba, a pronounced decrease in CC with increasing CBH (approximately −4.931% km⁻¹). These results exhibit the value of coordinated multi-site nighttime observations using standardized ground-based instrumentation. The study provides a methodological foundation for expanding nighttime cloud monitoring networks and offers observational constraints useful for validating satellite-based nighttime CC and improving the representation of nighttime clouds in atmospheric models.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"335 ","pages":"Article 108823"},"PeriodicalIF":4.4,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mechanism of the quasi-biweekly precipitation variability over the Mongolian Plateau and its simulations in CMIP6 models","authors":"Yepeng Xiao ,&nbsp;Yingxia Gao ,&nbsp;Rong Li ,&nbsp;Yitian Qian ,&nbsp;Xiaodong Liu","doi":"10.1016/j.atmosres.2026.108767","DOIUrl":"10.1016/j.atmosres.2026.108767","url":null,"abstract":"<div><div>The Mongolian Plateau (MP), a typical arid and semi-arid region, hosts fragile ecosystems that are highly sensitive to variations in precipitation. Within a summer season, precipitation in the MP exhibits significant 10–30-day variability, accounting for ∼40% of the total precipitation variability. This study investigates the characteristics of quasi-biweekly precipitation in the MP and evaluates its representation in 20 CMIP6 climate models. Surprisingly, all models underestimate the intensity of the quasi-biweekly precipitation (only 35–80%). Further analysis shows that this deficiency mainly results from the insufficiency of dynamic conditions associated with ascending motion rather than the thermodynamic conditions related to moisture. Diagnosis of the scale-decomposed moisture budget equation reveals that the anomalous water vapor in the MP mostly comes from horizontal moisture advection, in which both anomalous southeasterlies transporting the background moisture and background westerlies transporting the moisture perturbations into the MP play positive roles. CMIP6 models reasonably capture these moisture processes, but they poorly simulate the mid-latitude wave-train system that is critical for generating ascending motion. The pronounced positive vorticity advection observed on the lee-side of the Altai Mountains is not captured in the CMIP6 simulations. Consequently, compared to the ascending motion appearing over the entire MP in the observations, the simulated upward motion is only present in the eastern plateau. The negative biases of the circulation anomalies greatly suppress the quasi-biweekly rainfall intensity, highlighting dynamical deficiencies in current climate models.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"334 ","pages":"Article 108767"},"PeriodicalIF":4.4,"publicationDate":"2026-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145961778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MCF-XCO2: A cross-mission consistency and fusion framework for integrating multi-satellite XCO2 observations","authors":"Yutang Yu ,&nbsp;Wenjie Tian ,&nbsp;Lili Zhang ,&nbsp;Tao Yu ,&nbsp;Yu Wu ,&nbsp;Tianhai Cheng","doi":"10.1016/j.atmosres.2026.108747","DOIUrl":"10.1016/j.atmosres.2026.108747","url":null,"abstract":"<div><div>Carbon satellites, as an essential means of obtaining atmospheric XCO₂ concentration, play a key role in monitoring the global carbon cycle. However, the differences in observation platforms, resolutions, and inversion algorithms among different satellites lead to apparent inconsistencies among multi-source XCO₂ data, which limits the joint application and comprehensive analysis of the data. In this paper, we develop a framework named <strong>MCF-XCO</strong>_{<strong>2</strong>} (<strong>M</strong>ulti-source <strong>C</strong>onsistency <strong>F</strong>usion of <strong>XCO</strong>_{<strong>2</strong>}) for correcting multi-source satellite XCO₂ observations and performing uncertainty-weighted fusion. The method leverages high-precision satellite products as references, while minimizing the need for direct ground-based correction, to enhance the consistency and overall accuracy of multi-source observations. Based on this framework, multi-source satellite data, including GOSAT, GOSAT-2, OCO-2, and OCO-3, were integrated to construct a sparsely gridded global XCO₂ fusion dataset at 0.01° × 0.02° nominal spatial resolution and nominal daily sampling, reflecting available observations. The findings indicate that the fused dataset shows improved coverage in grids with available observations compared to individual satellite products, improved accuracy, and better consistency over time. Independent validation against TCCON ground-based observations further confirms the method's effectiveness, with R² = 0.91, RMSE = 1.09 ppm, bias = 0.07 ppm, and MRE = 0.2 %. The spatial and temporal dynamic analysis of the fused dataset reveals the typical spatial structure and seasonal variation of global carbon concentration, demonstrating the potential application of this dataset in studying the carbon cycle. The MCF-XCO₂ framework is also designed to accommodate future satellite missions, supporting timely updates and extended temporal coverage.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"334 ","pages":"Article 108747"},"PeriodicalIF":4.4,"publicationDate":"2026-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145902909","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Understanding the spatial heterogeneity of black carbon variation drivers in China: Views from explainable machine learning","authors":"Huang Zheng ,&nbsp;Weiwei Chen ,&nbsp;Xiaohui Bi ,&nbsp;Nan Chen ,&nbsp;Cheng Wu ,&nbsp;Mingming Zheng ,&nbsp;Shaofei Kong","doi":"10.1016/j.atmosres.2026.108749","DOIUrl":"10.1016/j.atmosres.2026.108749","url":null,"abstract":"<div><div>Differences in black carbon (BC) emissions and diverse climate zones in China result in significant spatial heterogeneity of BC levels. Understanding the roles of emissions, meteorological conditions, and other factors in BC variation is important for mitigating its adverse effects. This study conducted synchronous observations of BC in Changchun (CC), Tianjin (TJ), Wuhan (WH), and Guangzhou (GZ) during winter. Results showed that BC levels were highest in WH compared to the other cities (<em>p</em> &lt; 0.001). Simulations using the Flexible Particle Dispersion (FLEXPART) model and BC emission inventories quantified the contributions from local transport and different sectors. Results indicated that local emissions were the dominant geographical source of observed BC levels at all four sites. BC from residential, commercial, and other sectors was the dominant source in TJ, WH, and GZ. Using the SHapley Additive exPlanations (SHAP) method, emissions were identified as the predominant factor influencing BC variation at the four sites. While changes in meteorological conditions were the most influential factor contributing to BC concentration increases as air quality worsened from clean to polluted in CC, TJ, and WH. Regarding the impacts of features on model outputs, emissions showed a linear relationship with simulated BC, while the effects of transport and meteorological conditions exhibited spatial heterogeneity. This study highlights the need for continuous reduction of BC emissions to decrease ambient BC levels, and this recommendation can be extended to other parts of the country.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"334 ","pages":"Article 108749"},"PeriodicalIF":4.4,"publicationDate":"2026-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145903481","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Atmospheric CO₂ concentration gradients at the coastal region of the Yangtze River Delta: Patterns and drivers","authors":"Yi Lin ,&nbsp;Honggang Lv ,&nbsp;Kehan Chen ,&nbsp;Peng Guo ,&nbsp;Yifei Jiang ,&nbsp;Lin Xiao ,&nbsp;Tienan Zhao ,&nbsp;Haiyan Wang ,&nbsp;Yuanyuan Chen ,&nbsp;Kunpeng Zang ,&nbsp;Shuangxi Fang","doi":"10.1016/j.atmosres.2026.108746","DOIUrl":"10.1016/j.atmosres.2026.108746","url":null,"abstract":"<div><div>Atmospheric carbon dioxide (CO₂), the primary anthropogenic greenhouse gas, is a critical tracer for understanding carbon-cycle–climate feedback. Despite intense industrialization and population density, the Yangtze River Delta (YRD) lacks high-frequency in situ CO₂ observations in its coastal and marine-influenced zones. In this study, the continuous atmospheric CO₂ measurements collected between December 2020 and December 2022 at an urban tower in Shanghai and a coastal background site in Shengsi were analyzed. Results reveal a distinct bimodal diurnal cycle in Shanghai, primarily driven by local anthropogenic emissions, whereas Shengsi exhibits a unimodal pattern more closely coupled to biospheric processes and marine dilution. Nighttime data from Shanghai station reliably represents urban background concentrations, as the stable concentration performance, less planetary boundary layer (PBL) and anthropogenic impacts. Wind and trajectory analyses link CO₂ enhancement at Shanghai to emissions from northern and northwestern sectors, whereas Shengsi concentrations are modulated by ocean-atmosphere carbon exchange—as evidenced by correlations with sea surface temperature, salinity, and pressure, alongside stronger negative links to normalized difference vegetation index (NDVI) and air temperature compared to Shanghai. Furthermore, partial least squares regression (PLSR) further highlights the dominance of local emissions at Shanghai (&lt; 200 km scale) and oceanic/biospheric drivers at Shengsi, with regional transport amplifying variability across scales. These findings advance understanding of CO₂ spatiotemporal variability in coastal megaregions and provide an empirical basis for improving top-down carbon flux estimates and informing targeted climate mitigation strategies in eastern China.</div></div>","PeriodicalId":8600,"journal":{"name":"Atmospheric Research","volume":"334 ","pages":"Article 108746"},"PeriodicalIF":4.4,"publicationDate":"2026-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145903501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}