Methane (CH₄) is a short-lived climate forcer (SLCF) with a high global warming potential; thus, reducing CH₄ emissions is expected to mitigate near-term global warming. Therefore, accurate emission inventories are essential for climate action; however, atmospheric observations suggest the presence of unaccounted-for CH₄ sources, particularly in urban areas. In Japan, urban CH₄ concentrations have increased faster than the global mean despite reported declines in emissions, highlighting discrepancies between observations and inventories for CH₄. This study compares long-term methane trends at urban and background monitoring sites in Japan to assess urban influences. Hourly CH₄ concentrations from National Atmospheric Monitoring Stations (NOMS; NIES, 2025) and 3 background sites (WDCGG, 2024) during the period 2010–2019 were analyzed. Long-term trends were derived from monthly means after removing seasonal variability, and stations were classified using hierarchical clustering. Gridded CH₄ flux datasets were employed for comparison, including anthropogenic emissions (EDGAR, 2025) and natural sources (WetCHARTs v1.3.3; GFED5, 2023; Ito, 2021). Background CH₄ concentrations showed a typical increasing trend (6.71 ppb yr^-1),which was moderately correlated with the trend in global emissions (R=0.36), while Japanese emissions were decreasing. This suggests that CH₄ concentrations at the Japanese background sites reflect emission changes at the global-scale. In contrast, CH₄ concentrations at multiple urban NOMS periods of rapid increase during this period and showed higher long-term growth rates higher than those observed background sites (urban means: 7.50 ppb yr^-1). These results suggest urban-specific CH₄ sources not fully captured by existing inventories. Ongoing analyses leverage denser urban monitoring networks (Atmospheric Environmental Regional Observation System) to examine spatial and temporal variability in CH₄ concentrations with co-emitted air pollutants to better identify local sources contributing to urban CH₄ changes. Preliminary results will be presented at the conference.

Carbon monoxide (CO) is a trace gas primarily emitted by incomplete combustion, with biomass burning as a major source, and is widely used as an indicator of atmospheric transport and oxidation chemistry. CO concentrations therefore exhibit nonlinear and nonstationary variability across a wide range of temporal scales. In this study, Holo–Hilbert Spectral Analysis (H-HSA), based on Masking Empirical Mode Decomposition (Masking-EMD), is applied to characterize multiscale CO variability at the Lulin Atmospheric Background Station (LABS, 2862 m) during 2006–2023. The mean CO concentration at LABS is 136.6 ppb, indicating influence from long-range transport in East and Southeast Asia. The decomposition suggests that short-term variability is weak; the diurnal oscillation (~1 day, IMF3) shows a modest amplitude (~4.6 ppb) and likely reflects episodic enhancements associated with terrain-induced local transport, particularly daytime upslope flows. Variability becomes more pronounced at longer timescales, with dominant seasonal and interannual signals. The semiannual (~0.5 year) and annual (~1 year) modes have amplitudes of 34.3 and 25.9 ppb, respectively, reflecting seasonal transport patterns and emission seasonality; these seasonal maxima are linked to biomass burning in continental Southeast Asia and southern China, which enhances dry-season outflow and elevates CO at LABS. During 2015–2017, the annual cycle weakened, producing an unusually flat seasonal variation. K-means clustering of 6,618 backward trajectories indicates a shift toward northeasterly transport during this period, reducing contributions from biomass-burning source regions and coinciding with ENSO-related circulation anomalies. Interannual oscillations spanning ~2–19 years are identified, including a distinct ~4.6-year mode consistent with ENSO modulation, and the Holo–Hilbert spectrum shows an interannual modulation band (>4 years). Overall, H-HSA provides a useful framework for resolving multiscale, nonstationary CO variability and for examining the combined influences of emissions, transport pathways, and climate variability at high-elevation background sites.

Car-DOAS enables high-resolution mapping of atmospheric trace gases around localized emission sources. In this study, Car-DOAS observations were operated along closed-circle routes surrounding the Yeosu–Gwangyang industrial complexes, and the retrieved NO2 columns were analyzed together with wind fields from ERA5 reanalysis. Most of the observations exhibited clear inflow–outflow patterns depending on wind direction, indicating well-defined transport of industrial plumes. In comparison with satellite observations, Car-DOAS was able to resolve plume structures at much finer spatial scales than the coarse satellite footprints.
By combining the Car-DOAS NO2 columns with wind information, we estimated NOx emission fluxes using a flux-integration approach that accounts for plume transport across the observation paths. Distinct and coherent NO2 plumes aligned with the prevailing wind fields were observed on meteorologically favorable days, providing suitable conditions for flux estimation.
Our results demonstrate that Car-DOAS observations provide valuable top-down information on NOx emissions at the scale of large industrial complexes. When combined with bottom-up emission inventories and satellite remote sensing products, this approach can substantially improve the interpretation of regional emission sources.

Planetary Boundary Layer Height (PBLH) is recognized as a key factor in atmospheric pollution dispersion, weather forecasting, and climate change studies, and it plays an important role in tropospheric processes such as aerosol distribution, convective activity, and cloud and fog formation. Estimation of PBLH is essential for improving atmospheric models, enhancing weather prediction, and monitoring air quality and climate. These PBLH estimates are mainly based on direct observational data such as radiosondes and lidars; however, studies relying on direct observations have inherent limitations in spatial and temporal resolution. Therefore, studies based on direct observations have inherent limitations in spatiotemporal resolution. In addition, it is difficult to define a single true value for currently estimated PBLH. Since multiple estimates of PBLH can be useful for research from various perspectives, it would be advantageous to derive many variables that have physical characteristics similar to PBLH.Accordingly, this study aims to alleviate the aforementioned limitations and to estimate Aerosol Scale Height (ASH), which can have physical properties similar to PBLH. For this purpose, Aerosol Optical Depth (AOD) data from the Geostationary Environment Monitoring Spectrometer (GEMS) and visibility data from the Automated Synoptic Observing System (ASOS) were used. GEMS provides AOD at hourly intervals, and ASOS also offers continuous observations, enabling the estimation of ASH on an hourly basis. Using these two datasets, the extinction coefficient was first calculated via the Koschmieder equation, and then ASH was derived by inversely solving the equation for AOD. The estimated ASH was compared with PBLH from a reanalysis dataset, and the results showed that the two exhibited similar tendencies in terms of their diurnal cycle.Acknowledgements : This work was supported by Korea Environment Industry &Technology Institute (KEITI) through "Climate Change R&D Project for New Climate Regime", funded by Korea Ministry of Environment (MOE) (2022003560006).

We compared GEMS AEH and HSRL extinction height obtained from HSRL aerosol extinction profile data during the ASIA-AQ campaign. Extinction profile data is categorized into three types based on peak count and altitude. Single low (SL) data is profile which has single peak with under 2 km, single high (SH) data is single with higher peak. And double (D) is for the others. All categories showed positive bias, the bias was approximately 0.5 km for single peak data. Although SH sample size is much smaller than SL, distribution of bias of SH shows more symmetric form. This is because SH profile is closer to the Gaussian distribution assumed for GEMS AEH retrieval algorithm. However, D type showed negative outliers, and this is estimated as coming from single layer assumption of GEMS AEH. HSRL extinction height could be positioned at upper layer for multi-layer profiles. These results show necessity for considering diverse conditions for GEMS AEH retrieval to improve its accuracy at more complex atmosphere.

In February 2026, the first field campaign of Geostationary Environment Monitoring Spectrometer-Air Quality (GEMS-AQ) will be conducted in Bangkok, Thailand. This campaign aims to investigate severe air pollution in the Bangkok Metropolitan Area and to validate GEMS satellite NO₂ retrievals using combined multi-platform observations, including in-situ, ground-based remote sensing, and tethered sonde observations. GEMS-AQ is distinguished by integrated high-resolution multi-platform observations, especially tethered sonde measurements that provide truth diurnal NO₂ vertical profiles within the boundary layer for GEMS validation. The tethered sonde system consists of a 16m³ Helikite, an electric winch, and a tether, with a total payload capacity of approximately 9 kg. For NO₂ vertical profile measurements, we deployed a Portable cavity-enhanced Absorption of Nitrogen Dioxide Analyzer (PANDA), based on incoherent broadband cavity-enhanced absorption spectroscopy (IBBCEAS). The altitude of the tethered-sonde system is monitored in real-time using an iMet-4RSB radiosonde, enabling repeated ascent and descent cycles during daytime to measure diurnal NO₂ vertical profiles up to ~1.3 km. Along with tethered sonde, in-situ, ground-based remote sensing instruments will be co-located, including a NO₂ analyzer (Teledyne T500U), Pandora (P28), and a Multi-Axis Differential Optical Absorption Spectroscopy (MAX-DOAS). The dataset collected from this multi-platform during GEMS-AQ is expected to improve the NO₂ retrieval algorithms of Pandora and MAX-DOAS, also contribute to the validation and improved accuracy of GEMS satellite products by updating a priori NO₂ vertical profiles.

Satellite observations of carbon monoxide (CO) and nitrogen dioxide (NO₂) provide critical information for monitoring global air quality and atmospheric composition. However, persistent data gaps caused by cloud cover, high solar zenith angles, surface reflectance limitations, and instrument constraints significantly hinder the generation of spatially and temporally continuous products, particularly over polluted and cloudy regions. Here, we develop a global, high-resolution machine learning–based gap-filling framework to reconstruct daily satellite CO and NO₂ fields with enhanced spatial completeness and accuracy. Multi-source datasets, including TROPOMI CO and NO₂ retrievals, meteorological variables from ERA5, chemical reanalysis from GEOS-CF, land surface characteristics, and auxiliary geospatial predictors, are integrated to train ensemble learning models. The framework is implemented in a high-performance computing environment, enabling global-scale processing at a spatial resolution of 0.05°. Cross-validation and independent validation using ground-based in-situ measurements demonstrate robust predictive performance, with substantial improvements in spatial coverage and reduced reconstruction uncertainties relative to conventional interpolation approaches. The resulting gap-filled global CO and NO₂ datasets exhibit continuous spatiotemporal patterns that capture both large-scale transport and fine-scale urban pollution features. Applications over major megacities reveal enhanced representation of emission hotspots and pollution gradients, facilitating improved characterization of emission sources, atmospheric transport processes, and pollution exposure. This dataset provides a valuable foundation for air quality assessment, emission inversion, chemical transport model evaluation, and data assimilation studies. Our results highlight the potential of machine learning–driven gap-filling to overcome intrinsic limitations of satellite observations, offering a scalable pathway toward generating long-term, high-resolution global atmospheric composition products for environmental monitoring and policy-relevant applications.

Light-absorbing carbonaceous aerosols play a key role in atmospheric radiative forcing, yet long-term variability in aerosol absorption optical depth (AAOD) remains poorly understood in East Asian background regions. This study examines long-term AAOD trends at the Anmyeondo Global Atmosphere Watch (GAW) station, influenced by East Asian continental outflow.
Ground-based observations from 2020 to 2024 were analyzed using aerosol absorption and scattering coefficients and AERONET A OD. While AOD₅₅₀ remained stable (0.227–0.260), AAOD₅₅₀ increased to 0.137 in 2024, with the AAOD/AOD ratio rising from ~17% to 54.6%. AAE showed a significant increasing trend (+0.066 yr⁻¹), suggesting enhanced wavelength-dependent absorption by BrC. Although Anmyeondo is a BC-dominant background site, optical classification showed decreasing BC-dominant and increasing BrC-related aerosol fractions. However, AAOD exhibited a weak negative correlation with BrC fraction (ρ = −0.22), demonstrating that increased absorption cannot be explained by BrC concentration dominance alone, but is more strongly associated with mixing-state-induced enhancement in carbonaceous aerosol absorption efficiency.
These findings demonstrate that long-term AAOD variability in East Asian background regions is associated with changes in internal mixing state of carbonaceous aerosols.This work was supported by the Korea Meteorological Administration Research and Development Program "Development of Asian Dust and Haze Monitoring and Prediction Technology" (KMA2018-00521) and the Institute for Information & Communications Technology Promotion (IITP) grant funded by the Korea Government (Ministry of Trade, Industry and Energy, MOTIE) (RS-2025-02263037).

We investigated methane distribution over northeast Asia using the nadir viewing Greenhouse-gases Observing SATellite (GOSAT) and limb-viewing Atmospheric Chemistry Experiment Fourier Transform Spectrometer (ACE-FTS) satellite measurements. We found the ~25 ppbv difference of methane between East and Northeast China, implying that methane can have some regional difference in a same country. This feature is different from our typical idea, “Greenhouse gases usually do not have large regional difference”. From the comparison to the climate modeling results (Coupled Model Intercomparison Project - phase 6), it seems that climate modeling results used to underestimate methane, showing the necessity to fix emission part in the climate model. Next, we tried to see the vertical profile of methane in the Upper Troposphere and Lower Stratosphere (UTLS) using the ACE-FTS dataset. When we calculated correlation coefficients of methane between the surface measurements and upper atmospheric satellite measurements, it seems that correlations are not very strong. When we repeated same correlation studies for CO₂ instead of methane, however, correlation looks obvious between surface and upper atmosphere. These findings also reveal that methane is the greenhouse gas having large regional difference. So far, surface monitoring for greenhouse gases was sparse because the spatial homogeneity of greenhouse gas was typically considered. But our study means that we need to construct methane monitoring system more dense, to better understand the regional difference. Acknowledgement: This work was supported by Korea Environment Industry &Technology Institute (KEITI) through "Climate Change R&D Project for New Climate Regime", funded by Korea Ministry of Environment (MOE) (2022003560006).

Accurate identification of greenhouse gas emission sources depends on the quality of meteorological input fields, as atmospheric transport processes controlling the advection and dispersion of surface emissions are governed by wind fields. In Lagrangian particle dispersion models (LPDMs), small errors in wind speed or direction can lead to substantial uncertainties in backward trajectories and source attribution, particularly in regions with complex transport pathways. Such uncertainties may vary depending on whether mesoscale observations are assimilated into the meteorological input fields. This study evaluates how meteorological fields produced by assimilating synoptic- and mesoscale observations using an ensemble data assimilation approach improve LPDM-based greenhouse gas backward trajectories and emission source attribution.An ensemble data assimilation system based on the Local Ensemble Transform Kalman Filter (LETKF) is applied to assimilate Atmospheric Motion Vectors (AMVs), radiosonde, and wind profiler observations over East Asia, including the Korean Peninsula. High-concentration CO₂ events observed at Anmyeondo (AMY) are selected, and experiments with data assimilation (DA) are compared with those without data assimilation (NoDA). The resulting meteorological fields are used as input to the Stochastic Time-Inverted Lagrangian Transport (STILT) model to perform 48-h backward simulations and to derive two-dimensional footprints and surface emission contributions. Verification against radiosonde and wind profiler observations shows that the DA experiment reduces errors in low-level wind components and reproduces observed flow patterns during high-CO₂ inflow periods. In the STILT simulations, the NoDA experiment produces widely dispersed contribution regions, whereas the DA experiment yields more concentrated source regions, indicating reduced spatial uncertainty in emission source attribution.
Acknowledgments: This work was supported by the National Research Foundation (NRF) funded by the Korean government (MSIT) (Grant No. RS-2025-02242970) and Learning & Academic research institution for Master's⋅PhD students, and Postdocs (LAMP) Program of the NRF grant funded by the Ministry of Education (No. RS-2023-00301914).

This study examined daily ozone variability during the Asian summer monsoon using 62 daily ozonesonde measurements obtained from the pre-ACCLIP (2021) and ACCLIP (2022) campaigns. We also evaluated daily variability in vertical profiles of Geostationary Environment Monitoring Spectrometer (GEMS) ozone products. The daily ozone variability across the troposphere and lower stratosphere is closely linked to Asian monsoon circulations and synoptic weather events such as Rossby wave breaking and tropical cyclones. GEMS validation against the ozonesonde measurements reveals substantial quantitative improvements in GEMS version 3 (v3) compared to version 2 (v2). Regression slopes of GEMS versus ozonesonde total column ozone increased from 0.78 to 1.14 in 2021 and from 0.48 to 0.83 in 2022, indicating improved representation of daily variability. Layer-resolved evaluation shows that the largest improvement in v3 is most evident in the lower-troposphere. In this layer, correlations with ozonesonde increase from 0.56 to 0.74 in 2021 and from 0.18 to 0.66 in 2022. Consistent with these statistics, the lower-tropospheric ozone overestimation seen in v2 is reduced in v3 in the profile comparison. Process-based case studies during major synoptic events further support these improvements. GEMS v3 better captures ozonesonde-consistent signatures of a 2021 stratosphere–troposphere exchange (STE)-related high-ozone intrusion and 2022 typhoon-related low-ozone episodes that were poorly represented in v2. Spatially, GEMS v3 shows more realistic ozone patterns and reduced positive bias in the troposphere compared to v2. Overall, the improved v3 product more consistently reproduces observed daily ozone variability across key vertical layers, supporting its use for studies of transport processes and regional air-quality variability over East Asia.

Polluted air masses reaching high-altitude environments often carry chemical signatures shaped not only by their source regions but also by cloud and fog processing during transport. Here we examine the chemical evolution of polluted air masses arriving at a high-altitude mountain site in Taiwan, using a combination of in-situ aerosol and fog-water measurements with Lagrangian transport analysis. Observations include particulate-matter composition and aqueous-phase chemistry from fog samples, enabling direct comparison between dry and multiphase processing regimes. Air-mass histories are reconstructed using FLEXPART driven by ERA5 reanalysis, allowing identification of source regions, transport pathways, and meteorological conditions encountered en route. A key focus is quantifying cloud processing time prior to arrival at the site and distinguishing between events dominated by locally formed fog and those associated with clouds advected to the mountain. Reanalysis-derived cloud liquid water content, relative humidity, and temperature sampled along trajectories are used to estimate the duration of aqueous-phase exposure before the air masses reached the site. The air masses analyzed originate primarily from biomass burning regions in Peninsular Southeast Asia, as well as mixed anthropogenic source regions in East Asia. Clear differences in chemical composition are observed between PM and fog samples, consistent with varying degrees of cloud processing during transport. These results highlight the importance of multiphase processing in shaping the chemical properties of polluted air masses at high-altitude sites and demonstrate the value of combining trajectory-based diagnostics with coordinated aerosol and fog measurements to better understand regional pollution transport in East Asia.

The Geostationary Environment Monitoring Spectrometer (GEMS) is a geostationary Earth orbit (GEO) satellite instrument launched by Korea in 2020, providing hourly daytime observations of atmospheric pollutants, aerosols, and trace gases over Asia. Among the GEMS products, nitrogen dioxide (NO₂) plays a critical role in photochemical ozone (O₃) formation and exhibits strong hourly variability driven by changes in anthropogenic emissions and atmospheric chemistry. Therefore, investigating the temporal and spatial variations of NO₂ using GEMS observations is essential for improving our understanding of photochemical processes and their controlling factors. However, accurate interpretation of geostationary NO₂ observations requires careful validation by comparing with ground-based measurements, particularly to assess uncertainties associated with diurnal variability and vertical representativeness. In this study, we evaluate the accuracy of GEMS v4.0 NO₂ tropospheric vertical column density (NO₂ TropVCD) through comprehensive comparisons with MAX-DOAS observations at both urban and rural sites in Korea. We assess the consistency, bias, and correlation of NO₂ TropVCD and analyze the diurnal, seasonal, and vertical characteristics of the satellite–ground agreement.

The Airborne and Satellite Investigation of Asian Air Quality (ASIA-AQ) campaign was held from February to ­March 2024 to investigate air quality in Asia. During the ASIA-AQ period, various observations were carried out, including ground-based, airborne, and remote sensing measurements. In this study, vertical profiles of trace gases and aerosols were retrieved from Multi-Axis Differential Optical Absorption Spectroscopy (MAX-DOAS) measurements and evaluated through comparisons with other observations obtained during ASIA-AQ. Nitrogen dioxide (NO₂) surface concentrations from MAX-DOAS showed moderate to strong correlation coefficients of 0.63–0.88 with in-situ measurements, but have negative biases from –15% up to –48%. These negative biases reflect spatial heterogeneity of NO₂ concentrations associated with differences in the locations of MAX-DOAS and in-situ instruments. MAX-DOAS NO₂ concentrations below 1 km were consistent, within uncertainty ranges, with airborne in-situ instruments (NASA GSFC CANOE and NCAR NO_x and O₃ Chemiluminescence) onboard NASA’s DC-8 aircraft. NO₂ vertical column densities from MAX-DOAS showed very good agreement with those from an airborne spectrometer (GCAS), with a correlation coefficient of ~0.96. MAX-DOAS aerosol optical depth (AOD) exhibited high correlations with AERONET measurements (0.77–0.81), and aerosol extinction profiles were also in good agreement with airborne lidar measurements. In addition, preliminary analyses of formaldehyde (HCHO) and nitrous acid (HONO) from MAX-DOAS will be presented by comparing to available supporting observations, highlighting the capability of MAX-DOAS to characterize key reactive species.

Fine particulate matter (PM2.5) concentrations in Southeast Asia frequently exceed World Health Organization (WHO) air quality guidelines, resulting in serious air pollution. In Thailand, elevated levels of fine particulate matter are particularly pronounced during the dry season, driven by the combined effects of biomass burning and dense urban traffic emissions. As part of the preparatory phase for the upcoming GEMS-AQ field campaign, this study presents a preliminary evaluation of the WRF-Chem PM2.5 simulations over Southeast Asia, with a specific focus on the Bangkok metropolitan region. A one-way nested modeling framework was employed, consisting of an outer domain covering Southeast Asia at a 20 km horizontal resolution and an inner domain centered on Bangkok at a 4 km resolution. The inner domain was designed to represent better local atmospheric circulation and the impacts of biomass burning. Anthropogenic and biomass burning emissions were obtained from EDGAR version 8.1 and FINN version 1.5, respectively, and were speciated according to the chemical mechanisms implemented in WRF-Chem. Simulations were conducted for March 2022, corresponding to the dry season and the EDGAR base year, during which active fire occurrences are most frequent. The WRF-Chem simulation results were compared with GEMS aerosol products from the geostationary satellite GEO-KOMPSAT2B and with PM2.5 over Thailand. The preliminary evaluation of the WRF-Chem modeling system set up in this study is expected to provide preliminary information for analyzing the integrated observational data and WRF-Chem PM2.5 simulation results planned for the future GEMS-AQ campaign. Further details on the aerosol performance of the WRF-Chem model over the Thailand region will be discussed in an upcoming session.

Particulate matter (PM) pollution remains a critical environmental concern across Southeast Asia, driven by rapid urbanization and widespread fossil fuel and biomass combustion. Carbonaceous aerosols, particularly black carbon (BC) and brown carbon (BrC), contribute substantially to regional radiative forcing due to their light-absorbing properties. However, large uncertainties persist in quantifying their optical characteristics, atmospheric processing, and source-dependent variability, limiting the accurate interpretation and validation of satellite-based aerosol observations.As part of the GEMS-AQ initiative, this study aims to provide in-situ constraints on the optical and chemical properties of carbonaceous aerosols in a densely populated urban environment. A three-week intensive observation campaign in January-February 2026, will be conducted in Bangkok, Thailand, integrating continuous ground-based measurements of BC absorption along with size-resolved PM sampling. Black carbon light absorption will be measured using an aethalometer, and additionally with a thermodenuder, to assess the influence of atmospheric aging on light absorption. Size-segregated PM samples will be collected using high-volume samplers for offline chemical and optical analysis. Filter extracts will be analysed to quantify water-soluble and methanol-soluble organic carbon (WSOC and MSOC), enabling the characterization of BrC absorption. These observations aim to link aerosol optical properties with dominant combustion sources and its enhancement due to aging in Southeast Asia. The primary objective of this work is to reduce uncertainties in carbonaceous aerosol absorption and improve the integration of ground-based observations with satellite retrievals and modelling efforts under GEMS-AQ. Results from this campaign will be presented at the meeting.

There have been many studies for performing aerosol type classifications using sun-photometer measurements based on Aerosol Robotic Network (AERONET) platform. Through these studies, several methods were suggested, and they some interesting features were found about the dominant type of regional aerosols. But comparison of results from different methods was not much conducted, therefore it is not clear if there is consistency among different methods of aerosol type classification. In this context, we designed this study, which is composed of (1) comparison of different aerosol type classification, and (2) evaluation its performance by comparing to the aerosol composition of in-situ aircraft measurements. In February to March, 2024, ASIA Air Quality (hereafter, ASIA-AQ) campaign was performed in Philippines, South Korea, Taiwan, and Thailand. During the ASIA-AQ campaign period, we classified aerosol types using AERONET data in these countries, and then assessed their features by comparing the aerosol composition obtained from DC-8 aircraft measurements. Here we used three classification methods: (1) based on absorbing and scattering Angstrom exponent, (2) based on single scattering albedo and fine mode fraction, and (3) based on aerosol optical depth from 440 and 1020 nm. We found that these 3 methods reflect the regional aerosol types similarly in the angle of mean pattern, but also found that there are some critical differences among three methods. Comparison to aircraft measurements revealed that there is a large sensitivity of classified aerosol type to the portion of sea-salt in total aerosol amount.  Acknowledgement: This work was supported by Korea Environment Industry &Technology Institute (KEITI) through "Climate Change R&D Project for New Climate Regime", funded by Korea Ministry of Environment (MOE) (2022003560006).

Using observations from the GOSAT-GW/TANSO-3 satellite, launched in June 2025, we aim to detect CO₂ emissions from large sources, including megacities and power plants, by combining NO₂ observations as a tracer from combustion sources. Our measurements in Yokosuka (35.32°N, 139.65°E), located south of Tokyo, demonstrate and characterize their co-variation patterns. We used Pandora, MAX-DOAS, and EM27/SUN to observe CO₂ and NO₂ columns, and LI-7810 and CAPS instruments to observe surface concentrations during winters of 2024–2026. Using an updated retrieval algorithm (JMnet2) that reduced the depths of the first two layers near the ground, the MAX-DOAS profiles showed better agreement with the tropospheric NO₂ vertical column density (TropNO₂VCD) observed with Pandora146, as well as with the surface NO₂ observations. This provides a solid basis for satellite validation and further analysis. The column-averaged dry-air mole fractions (XCO₂) from the EM27/SUN at a 10 min resolution showed preliminarily positive correlations with the TropNO₂VCD from Pandora over 40 days, suggesting the capabilities to detect local and combustion-related emissions. On average, the slopes of the daily regression lines, ΔXCO₂/ΔTropNO₂VCD, were 2.8 (1.5–3.6) ×10³ mol/mol (N=10 with favorable conditions). They were higher than the surface values of ΔCO₂/ΔNO₂ on the same days, 1.9 (0.41–4.3) × 10³ mol/mol, which agreed reasonably well with the bottom-up CO₂/NOx emission ratios ranging from 0.09-0.38 × 10³ (road and other transportation) to 1.95 × 10³ (power plants for the Kanto region), considering that the observed surface NOx/NO₂ values were typically ~1.16. The higher column-based ratios could be due to non-local sources or oxidation of NOx to NOz. Caution is warranted when characterizing local emissions quantitatively from satellite observations with respect to such perturbations, and the spatial extent of the analysis.

Nitrogen dioxide (NO₂) is a well-known tracer of pollution emitted from combustion sources. The GOSAT-GW/TANSO-3 satellite observes NO₂ to improve GHG source identification and quantification. However, the NO₂ satellite observations require validation using independent measurements, such as airborne observations. Here, we developed an airborne imaging spectrometer for mapping atmospheric NO₂ for a joint aircraft observation campaign (Tokyo-FC) with NASA  scheduled for March 2026. The instrument has a 54.5° of field of view using a low-distortion wide-angle lens mounted above a bottom window of the aircraft to collect sun light reflected from the ground. The collected light is introduced to a spectrometer (Isoplane160, Teledyne Princeton Instruments) via a 48-core optical fiber bundle with 100 µm core diameters. The across-track resolution is estimated to be 189 m at flight altitudes of approximately 8800m. Spectra in the 429–491 nm wavelength range are recorded using a 600 lines/mm grating and a CCD detector (ProEM-HS:512BX3). The temperature of the spectrometer is stabilized at 25±0.2°C. Ground-based tests were conducted on the rooftop of a building at JAMSTEC in Yokosuka, Japan (35.32°N, 139.65°E), at an altitude of 27 m. The lens was mounted on a pan-tilt unit to vertically scan the line of sight from the ground to the sky. Reference spectra were acquired from zenith measurements. Exposure time and integration times were optimized to be 200 ms and 5 times, respectively. This configuration corresponded to an along-track resolution of 70–100 m. From the ground-based experiments, NO₂ was successfully retrieved with the DOAS algorithm with spectral-fitting residuals down to 1.5×10^-3, when measuring sun-glint positions on the sea surface and shaded ground areas. In this presentation, we introduce the developed NO₂ instrument in detail, together with the preliminary results.

Estimation of combustion efficiency (CE) remains challenging and is a major source of uncertainty in emission inventories, particularly in downwind regions where transported urban and industrial emissions strongly influence atmospheric composition. The nitrogen dioxide (NO₂) to carbon monoxide (CO) ratio (NO₂/CO) is widely used as an important proxy for combustion conditions and emission characterization. While previous studies have largely relied on space-borne sensors, such observations are typically limited to a single snapshot per day. In contrast, ground-based UV–VIS and Fourier Transform Spectroscopy (FTS) instruments provide more detailed and continuous observations. In this study, we use co-located column and surface measurements of NO₂ and CO, retrieved using UV–VIS (PANDORA) and FTS (EM27/SUN) techniques, to improve the quantification of CE and emission factors (EFs). The analysis is based on co-located observations from Pandora, EM27/SUN, CAPS-NO₂ and Picarro CRDS-CO instruments at Tsukuba (36.05° N, 140.12° E; 31 m a.m.s.l.), a semi-urban site in the northeastern Tokyo metropolitan area, from September 2021 to April 2025. Total column CO from EM27/SUN show positive correlations (r² > 0.5) with total column NO₂ from Pandora on 87 days out of 350 days, with a daily mean regression slope of 0.49 x 10^-1 mole mole^-1) (0.05–1.72 x 10^-1 mole mole^-1 ) with r² > 0.5. In comparison, surface-based hourly observations show strong positive correlations between surface CO and surface NO₂ for about 250 days, with a mean slope of 0.92 x 10^-1 mole mole⁻¹ (0.12–5.70 x 10^-1 mole mole⁻¹) based on at least 5 hourly data points and r² > 0.5. Furthermore, these results are utilised to evaluate the performance of emission inventories. More detailed will be presented.

Nitrogen oxides (NOx) are crucial components in understanding tropospheric photochemical processes, making the clarification of their variations essential. While satellite-based nitrogen dioxide (NO₂) observations have evolved toward urban-scale resolutions, geostationary satellites have also recently begun capturing their diurnal variations. Fukuoka City, a major urban area in the westernmost part of Japan, is influenced by both transboundary transport from the East Asian continent and significant local emissions. With NOx sources concentrated in the city center and dominant land-sea breeze circulations, this area provides a unique opportunity to distinguish between physical and chemical processes. This study validates satellite NO₂ observations, such as GEMS (Geostationary Environment Monitoring Spectrometer) and TROPOMI (Tropospheric Monitoring Instrument), using ground-based remote sensing (MAX-DOAS and Pandora). We focus specifically on characterizing the spatial inhomogeneity over the Fukuoka urban area.  Multi-azimuth ground-based observations during July–August 2025 effectively captured spatial inhomogeneity along with diurnal variations, and the GEMS satellite (L2, v3.0.0) successfully captured these variations. Nonetheless, the systematic underestimation in satellite observations remains a critical issue for future refinement.

MAX-DOAS systems and mini-DOAS instruments have been built in Taiwan as part of a joint project with IUP Bremen. Since summer 2024, one MAX-DOAS has been operating continuously at Cape Fuguei (the northern tip of Taiwan), providing a long-term record of background inflow and regional transport. We have also developed an on-road mobile MAX-DOAS platform to map NO₂ spatial gradients and plume structures across Taiwan.This study investigates NO₂ variability by combining fixed-site MAX-DOAS observations at Cape Fuguei, which constrain background conditions and transport regimes, with repeated on-road MAX-DOAS transects that capture near-field heterogeneity and identify enhanced NO₂ features. Nearby EPA in-situ NO₂ measurements are used to relate column observations to surface concentration variability under evolving boundary-layer conditions. By jointly analyzing fixed-site, mobile, satellite, and in-situ observations, we quantify NO₂ spatial heterogeneity from background conditions to urban outflow (e.g., Taipei City) and cross-validate satellite NO₂ products. The proposed framework provides a practical pathway to improve NO₂ source characterization and observational consistency across platforms in northern Taiwan.

The concentration of major greenhouse gases (GHGs) in the atmosphere is steadily increasing due to emissions from anthropogenic activities. Total and/or partial column measurements of GHGs using remote-sensing techniques from ground-based and space-based platforms provide a useful tool for the quantification of GHGs in the whole atmosphere and emission source apportionment. Well-calibrated ground-based measurements are further used to calibrate and validate satellite-based measurements and to establish an indirect traceability to the WMO scale. However, to do a full evaluation of the satellite data a dense distribution of such reference measurements is required around the source-area of interest and covering an extensive range of the measurand space in terms of influencing parameters. In this presentation, we will show some of the applications of Fiducial Reference Measurements highlighting the benefits of the selected set of instruments. Furthermore, we will also show the status of the implementation of the first Total Carbon Column Observing Network and Network for the Detection of Atmospheric Composition Change – Infrared Working Group (NDACC-IRWG) station and COllaborative Carbon Column Observing Network (COCCON) stations in India.