This study investigated the spatiotemporal variations of ozone (O₃) pollution and O₃ formation sensitivity in the Yangtze River Delta (YRD) with the air quality model CAMx and an optimized photochemical indicator (PH₂O₂/PHNO₃). The critical years of emission reduction strategies (2013, 2017, 2020, and 2030) were selected to simulate and predict changes in ozone concentration and formation sensitivity under different emission scenarios. The results revealed a north-south gradient in O₃ concentrations, with higher levels in northern areas and lower levels in southern areas. The ozone formation sensitivity regimes were divided by an Anqing-Nanjing-Suzhou-Jiaxing line. The northern aeras of the line were governed by a VOCs-limited regime (occupying 43.3% of the total area) and the southern areas were predominantly NO_x-limited (45.3%), with transitional zones covering the remaining area. Under fixed meteorological conditions, continuous reduction in anthropogenic emissions during 2013~2020led to a sustained decline in the daily maximum 8-hour average (MDA8) O₃ concentration, and a shift of the ozone formation regime from VOCs-limited to NO_x-limited. In 2030, O₃ levels in urban hotspots such as Shanghai are likely to increase, despite an overall reduction in regional O₃ levels. The shift towards the NO_x-limited region will continue, increasing by 30.2% from 2013 levels. Only specific cities along the Yangtze River and areas with heavy traffic emissions will remain VOCs-limited. In summary, excluding the impact of climate change, the emission mitigation strategies can effectively decrease the ozone levels in the YRD. Meanwhile it is still necessary to coordinate the reduction rates of NO_x and VOCs in VOCs-limited cities to prevent increases in O₃ levels.

In this study, ozone (O₃) formation sensitivity in Henan Province was assessed using the FNR indicator derived from the Community Multiscale Air Quality (CMAQ) model outputs, which was used to classify the province into distinct sensitivity zones. Observation data from national monitoring sites in Henan Province (2015~2022) were integrated with the Temperature method (Temp), Nighttime method (Night), 5_th percentile method, and CMAQ-Night method to evaluate regional background O₃ concentrations during both warm and cold seasons and their evolution trends across different zones. The results indicate that northern and eastern Henan are primarily VOC-limited, while the western and southern parts are predominantly NO_x-limited. In northern and eastern Henan, the planetary boundary layer exhibits stronger VOC-limitation than the troposphere. The results demonstrate a regime transition, where VOC-limited zones gradually shift to transition zones, and these transition zones evolve into NO_x-limited regimes. Background O₃ concentrations estimated by the Temp, Night, and 5_th percentile methods were 96.0, 103.2 and 88.4μg/m³, respectively, accounting for 64.9 69.7 and 59.9% of O₃-8h during the warm season. In the cold season, the corresponding estimates were 56.0, 73.6, and 40.0μg/m³, with the Temp and 5th percentile methods contributing 88.4 and 63.2% of O₃-8h, respectively. The result from the Night method exceeded O₃-8h, likely due to the significant influence of NO titration effects during the cold season. The annual increase in cold-season O₃ concentrations was attributed to the weakening of NO titration effects, driven by substantial reductions in NO_x emissions. The CMAQ-Night method revealed that, in June, background O₃ concentrations ranged from 73.1 to 151.4μg/m³ across Henan Province, accounting for 60% (in northern Henan) to 98% (in western Henan) of O₃-8h, with an average value of 103.3μg/m³ (71.6%). In VOC-limited zones of northern Henan, O₃ pollution was jointly driven by background contributions and local generation, while in the mountainous and hilly regions of western Henan, background contributions dominated. From 2015 to 2017, the increase in O₃ concentrations across the province was primarily driven by local generation (57%), whereas the decline from 2017 to 2020 was mainly attributed to reductions in background O₃ concentrations (58%). These findings highlight significant potential for further reducing locally generated O₃ through targeted emission control measures.

The typical compound pollution of PM_2.5 and O₃ in Baoding City was investigated in this study. Objective weather classification of sea-level pressure and 10m wind fields in the Beijing-Tianjin-Hebei region during the autumn-winter and spring-summer seasons was conducted using the Principal Component T-mode (PCT) method. Backward trajectory analysis (PSCF, CWT) and meteorological-air quality modeling system (WRF-CAMx) were applied to identify the source contributions and transport pathways of PM_2.5 and O₃ under pollution-related weather types. From 2019 to 2022, the annual average PM_2.5 concentration in Baoding showed a significant decline of 26.9%, whereas O₃ exhibited a fluctuating trend with a slight reduction of only 2.2%, indicating limited improvement. Monthly average concentrations of both pollutants demonstrated pronounced seasonal patterns. In autumn and winter, high-pressure systems dominated, and weather types T5 (eastern high-pressure rear, 11.6%), T6 (saddle-type, 11.6%), and T9 (isobaric field, 5.2%) were most frequently associated with PM_2.5 pollution events. In contrast, low-pressure systems prevailed in spring and summer, with weather types S3 (isobaric field, 12.9%) and S7 (northwest low-pressure rear, 16.8%) contributing notably to O₃ pollution. Local emissions were identified as the dominant source of PM_2.5 during pollution weather types. Under T5, additional contributions originated mainly from Handan and Shijiazhuang; under T6, from Shanxi Province and Shijiazhuang; and under T9, from Shanxi and Inner Mongolia. For O₃ pollution, local sources accounted for the largest share. Under S3, transport from surrounding cities played a secondary role, whereas under S7, contributions from southwestern regions, including Henan, Handan, Shanxi, and Shijiazhuang, were more prominent.

To enhance the independence and controllability of regional atmospheric environment numerical prediction and its capability to support air quality forecast and air pollution warning, a CMA-GFS driving numerical prediction system (RAEMS-CMA) was established and operationally implemented. This system was based on the NCEP-GFS driving Regional Atmospheric Environmental Modeling System for eastern China (RAEMS-NCEP). The forecast effects evaluation on the major air pollutants over the YRD region from July 4^th, 2023, to July 3^rd, 2024, was executed. Evaluation results showed that RAEMS-CMA had good performance in forecasting the major air pollutants, e.g. correlation coefficients were generally over 0.7 and root mean square errors were 30%~70% of corresponding observation means. The performance was consistent in different forecast lengths of 24h ~ 96h. Further analysis indicated that the performance differs from air pollutants, with better skills in forecasting PM_2.5 and O₃. The prediction of PM₁₀ was underestimated, while the others were overestimated. It also showed better forecast capability over the north region than over the south. The intercomparison indicated that RAEMS-CMA had slightly superior performance to that of RAEMS-EC (which was driven by ECMWF) and accordant to RAEMS-NCEP, while their differences were usually less than 5%. These results proved the applicability of CMA-GFS in numerical air quality prediction over the YRD region. At the same time, the divergent performance in pollutants and regions suggests further efforts on technological improvements and application approaches to numerical air quality prediction.

To investigate concentration variations, source characteristics, and relationships with environmental factors of biogenic volatile organic compounds (BVOCs) in urban areas, we conducted measurements of BVOCs at 100meters using a proton transfer reaction mass spectrometer (PTR-ToF-MS) at the Shenzhen Meteorological Gradient Tower in the spring and summer of 2024. During the observation period, the concentrations of isoprene, Methacrolein+methyl vinyl ketone (MVK+MACR), and monoterpenes were measured at (0.24±0.29)×10^-9, (0.20±0.22)×10^-9, and (0.11±0.14)×10^-9, respectively. These results indicate the isoprene concentrations were lower than those reported in Beijing during the same seasons, likely due to differences in environmental conditions and oxidation during vertical transport. Temporally, BVOC concentrations peaked at midday under high temperatures and strong sunlight. Spatially, higher concentrations were observed in the upwind direction towards the south, where vegetation is lush. The initial isoprene concentrations, calculated from the product concentrations, were significantly higher than the observed concentrations, suggesting a high degree of air mass aging. Source apportionment based on tracers revealed that the average contribution of anthropogenic sources to monoterpene concentrations was only 9.9%, but this contribution could reach 21.7% at night. Temperature was identified as a significant environmental factor influencing BVOC concentrations. The study found that the empirical coefficients of initial isoprene concentrations with temperature were more consistent with those of MVK+MACR than with observed isoprene concentrations, indicating that initial isoprene better represents the temperature dependence of plant emissions.

To investigate the vertical distribution and source characteristics of formic acid in the atmospheric boundary layer, online measurements of formic acid and related air pollutants were conducted at six heights from the ground level to 335m in autumn 2021 using iodide chemical ionization mass spectrometry (I-CIMS) coupled with a multi-inlet tubing system at the Shiyan Meteorological Gradient Tower in Shenzhen. Results revealed that the average ground-level formic acid concentration during the campaign was 0.7 ppb, lower than previously reported values from other regions globally. Vertically, formic acid concentrations exhibited a consistent height-dependent pattern with an initial increase followed by a decrease, peaking at 120m during both daytime and nighttime. This vertical maximum was particularly pronounced in daylight hours. Correlation analyses with co-measured pollutants (ozone, NO_x, PM_2.5) demonstrated that the sources of formic acid were influenced by both primary emissions and secondary formation. A vertical stratification of sources was identified: contributions below 120m predominantly originated from direct emissions of vehicle exhausts, while photochemical secondary formation became the dominant source above 120m.

Accurate estimation of CO₂ emissions at the regional scale, particularly in large urban agglomerations, is essential for effective greenhouse gas mitigation. In this study, the WRF-STILT model, combined with three prior anthropogenic CO₂ emission inventories (EDGAR v6.0, an improved version of EDGAR v6.0 integrated with GCG v1.0, and the ODIAC inventory) was used to simulate atmospheric CO₂ concentrations in the Yangtze River Delta during winter 2018. Observational data from a 70-meter tower in Quanjiao, Anhui province, served as the reference. A scale factor Bayesian inversion method was applied to optimize the simulation results, allowing estimation of anthropogenic CO₂ emission fluxes in the region. The results indicate that the CO₂ concentrations simulated by the WRF-STILT model effectively reflect the emission characteristics of the Yangtze River Delta. In winter, simulations based on the improved inventory aligned more closely with observations than those using EDGAR v6.0alone. The posterior CO₂ emission fluxes estimated using the EDGAR and improved inventories were (0.199 ± 0.005) and (0.200 ± 0.007) mg/(m²·s), respectively—representing decreases of 0.02 and 0.01mg/(m²·s) compared to the prior flux estimates. The Bayesian inversion method significantly optimized the prior emissions from the EDGAR inventory. When calculating total CO₂ emissions for the region using the improved inventory, power generation and industrial sectors were identified as the primary sources of uncertainty. Additionally, due to the low nocturnal boundary layer height, the model tended to overestimate the contribution of elevated emission sources near the observation site. Future simulations should ensure the WRF model accurately captures hourly boundary layer heights, and that vertical emission profiles of point sources are incorporated into emission inventories.

To clarify the variation characteristics and causes of urban atmospheric CO₂ concentrations during heatwave periods, this study investigates Hangzhou—a typical city in the Yangtze River Delta—using the WRF-STILT atmospheric transport model, the EDGAR anthropogenic emission inventory, high-precision laser spectrometer observations, and a Bayesian optimization inversion algorithm. Results show that heatwaves significantly affect atmospheric CO₂ concentrations: daytime concentrations are lower, while nighttime concentrations are higher compared to non-heatwave periods. The spatial representativeness of measurement sites and changes in CO₂ emissions are the main factors influencing atmospheric CO₂ levels in Hangzhou. Heatwaves also lead to notable changes in spatial representativeness: during heatwaves, local emissions account for 80.5% of the observed CO₂ compared to only 51.6% during non-heatwave periods, and representative spatial range is smaller. In addition, CO₂ emissions from various anthropogenic sectors respond differently to heatwaves: emission from the manufacturing, energy, and refining and conversion sectors increase, while emissions from transportation decreases.

This study systematically analyzes and compares the spatiotemporal coverage and distribution patterns of the Level-2 atmospheric column CO₂ concentration (XCO₂) products from OCO-3, OCO-2, and GOSAT satellites over China. The results showed that the annual average spatial coverage of the Level-2 XCO₂ products in China were 43.38%, 21.19% and 0.40% for OCO-3, OCO-2, and GOSAT, respectively, with corresponding monthly average coverages were 4.77%, 2.43% and 0.16%. Notably, OCO-3 exhibited a significant lower monthly average coverage in odd-numbered months (2.23%) compared to even-numbered months (7.31%). There were significant spatial distribution differences in the coverage of the three satellite products. The coverage of OCO-2 and OCO-3 in southern and central China was significantly lower than that in northeastern, northwestern and northern China. In contrast, GOSAT's observations were extremely sparse, mainly concentrated in eastern China. In addition, OCO-3 identified over 20 hotspot observation areas over provincial capitals and industrial cities, which showed markedly higher coverage and observation frequency. From 2019 to 2022, all three satellite XCO₂ products reveal consistent increasing trends in annual average XCO₂, albeit with declining annual increment. Seasonal pattern showed consistently higher XCO₂ in spring and winter than in summer and autumn. Spatially, annual average XCO₂ showed higher concentrations in the eastern than western China, and in southern than northern China. However, the location of XCO₂ extreme value varied across satellites, and significant differences in XCO₂ concentrations were observed among different products within the same sub-region.

This study investigated the effect of ozone and photolysis rate on formic acid formation mechanisms from oxidation of benzene across five major Chinese cities (Beijing, Shanghai, Guangzhou, Xi'an, and Chengdu) between 2014~2023. Combining the spatiotemporal patterns of ozone (maximum 8-hr average from China's National Environmental Monitoring Center), J_(NO2) (derived from solar zenith angle calculations) and the oxidation mechanism of benzene, the effect of ozone and J_(NO2) on formic acid production from the atmospheric oxidation of benzene were modelled. Key findings revealed that formic acid yields from benzene exhibited: (1) strong seasonal correlation with ozone levels across all cities (R² = 0.96); (2) consistent trends with that of J_(NO2) in Beijing, Shanghai, Xi'an, and Chengdu (R² = 0.80~0.94), contrasting with Guangzhou's weak correlation (R² = 0.27); (3) distinct seasonal variations with significant lower production in winter than other seasons in all cities, consistent peak production of formic acid in summer in Beijing, Xi'an, and Chengdu, alternating maxima between spring and summer in Shanghai, and peak formation in autumn in Guangzhou.