Korean Journal of Remote Sensing (대한원격탐사학회지)

Korean Society of Remote Sensing (대한원격탐사학회)
권호 표지
DOI QR코드

DOI QR Code

Evaluation of GEMS NO2 Retrieval Algorithm Version 2.0 and 3.0 Using TROPOMI and Pandora Observations

  • Yeonjin Jung (Major of Spatial Information Engineering, Division of Earth and Environmental System Sciences, Pukyong National University) ;
  • Junsung Park (Harvard-Smithsonian Center for Astrophysics) ;
  • Hyunkee Hong (Environmental Satellite Center, National Institute of Environmental Research) ;
  • Hanlim Lee (Major of Spatial Information Engineering, Division of Earth and Environmental System Sciences, Pukyong National University) ;
  • Ukkyo Jeong (Major of Spatial Information Engineering, Division of Earth and Environmental System Sciences, Pukyong National University)
  • Received : 2025.10.10
  • Accepted : 2025.10.27
  • Published : 2025.10.31
Download PDF
⟨ Previous Next ⟩

Abstract

The Geostationary Environment Monitoring Spectrometer(GEMS), onboard the Geostationary Korea Multi-Purpose Satellite-2B (GEO-KOMPSAT-2B) satellite, provides hourly measurements of nitrogen dioxide (NO2) over East Asia, enabling continuous monitoring of air quality and pollution transport. To ensure data reliability, the GEMS NO2retrieval algorithm has undergone iterative improvements. This study presents a comprehensive comparison and evaluation of the GEMS NO2 Version 2.0 (V2.0) and Version 3.0 (V3.0) data products using independent satellite (TROPOspheric Monitoring Instrument, TROPOMI) and ground-based Pandora observations. V3.0 introduces key updates, including revised a priori NO2 vertical profiles, revised tropopause pressure,recalculated air mass factors(AMFs), and updated auxiliary inputs such as aerosol optical depth, cloud, ozone, and surface reflectance. Pixel-by-pixel intercomparison between V2.0 and V3.0 shows strong overall correlation, while regression slopes below unity indicate reduced total NO2 columns in V3.0 due to the removal of the overestimated tropospheric signal. Stratospheric columnsincreased substantially, correcting the unrealistic underestimation observed in V2.0 and yielding more physically consistent vertical distributions. Comparison with TROPOMI demonstrates that V3.0 reduces systematic overestimation over industrial regions such as northeastern China and the Korean Peninsula, achieving improved spatial agreement and higher correlations. Validation with Pandora measurements across six East Asian sites(Beijing, Seoul, Busan, Bangkok, Tsukuba, and Tokyo) confirms that V3.0 decreases regional biases and root-mean-square errors(RMSE) by approximately 20%, with enhanced correlation coefficients. Overall, the algorithmic refinements in GEMS NO2 V3.0 lead to improved accuracy, stability, and physical realism in both tropospheric and stratospheric retrievals. The results demonstrate the enhanced capability of GEMS for diurnal and seasonal monitoring of air quality across East Asia, supporting its use for emission estimation, pollution transport analysis, and long-term trend assessment.

Keywords

1. Introduction

Nitrogen dioxide (NO2) is a major atmospheric pollutant with significant impacts on air quality, human health, and the environment. As a precursor to tropospheric ozone (O3) and secondary aerosols, NO2 plays a central role in the formation of photochemical smog (Crutzen, 1979; Beirle et al., 2003; Hong et al., 2017). Short-term exposure to elevated NO2 levels has been linked to respiratory and cardiovascular diseases, while long-term exposure contributes to chronic health risks (Richter et al., 2011; Yang et al., 2014). Because of its short atmospheric lifetime—ranging from several hours to a few days—NO2 exhibits strong spatiotemporal variability, with elevated concentrations near emission sources such as urban and industrial areas (Beirle et al., 2011). For these reasons, continuous monitoring of NO2 is essential for air quality assessment, emission inventory adjustments, and the formulation of environmental policies.

Satellite remote sensing has become a powerful tool for monitoring atmospheric NO2 at both global and regional scales. Instruments aboard low Earth orbit (LEO) satellites, such as the Ozone Monitoring Instrument (OMI) (Levelt et al., 2006) launched in 2004 aboard the Aura satellite and the TROPOspheric Monitoring Instrument (TROPOMI) (Veefkind et al., 2012) launched in 2017 aboard Sentinel-5 Precursor satellite, have provided valuable long-term global datasets. However, these sensors have limited temporal coverage of LEO satellites—typically one or two overpasses per day—which makes it difficult to capture short-term variations in emissions and atmospheric chemistry. To address this limitation, geostationary Earth orbit(GEO) satellites have been developed to provide high temporal resolution observations over certain regions.

The Geostationary Environment Monitoring Spectrometer (GEMS), launched in February 2020 onboard the Geostationary Korea Multi-Purpose Satellite-2B (GEO-KOMPSAT-2B) satellite, is the world’s first geostationary spectrometer dedicated to air quality monitoring (Kim et al., 2020). Covering East Asia, GEMS provides hourly measurements with a spatial resolution of approximately 7 × 8 km at nadir and a spectral range of 300–500 nm. These observations enable the monitoring of diurnal variations in air pollutants, the evaluation of transboundary pollution transport, and the improvement of emission estimates across East Asia. To ensure the reliability of GEMS data products, the retrieval algorithms have been continuously refined.

Version 2.0 (V2.0), the first operational product released in late 2021, incorporated a priori NO2 vertical profiles from Goddard Earth Observing System-Chem (GEOS-Chem) model by replacing Weather Research and Forecasting (WRF)-Chem and Community Atmosphere Model (CAM)-Chem and directly utilized GEMS Level-2 (L2) surface reflectance (BSR) data products (Choi et al., 2024). V2.0 successfully demonstrated the feasibility of geostationary NO2 monitoring, but comparisons with TROPOMI revealed systematic biases: overestimation of tropospheric NO2 and underestimation of stratospheric NO2, resulting in stratospheric distributions contaminated by tropospheric features. To mitigate these issues, Version 3.0 (V3.0) introduced recalculated air mass factors (AMFs) based on a restructured separation of tropospheric and stratospheric shape factors, together with updated auxiliary inputs including aerosol (AERAOD), cloud (CLD), total ozone (O3T), and BSR data products. These improvements were designed to enhance the accuracy of the retrievals and to improve consistency with independent observations.

In this study, we present a comprehensive comparison and evaluation of GEMS NO2 V2.0 and V3.0 data products. Independent datasets, including satellite observations from TROPOMI, ground-based measurements from Pandora, and model simulations, are employed to assess the improvements achieved in V3.0 relative to V2.0. The analysis focuses on spatial distributions, temporal variability, and statistical performance across East Asia. The results highlight the strengths and remaining limitations of current GEMS retrievals and provide insights into the future development of geostationary air quality monitoring algorithms.

2. Data

2.1. GEMS

The GEMS, onboard the GEO-KOMPSAT-2B satellite, is a hyperspectral imaging spectrometer designed to continuously monitor atmospheric trace gases over East Asia. GEMS observes solar backscattered radiance within the 300–500 nm wavelength range with a spectral resolution of 0.6 nm (Kim et al., 2020). The instrument provides hourly NO₂ observations during daytime (09:00 to 18:00 in local solar time) across the domain of 5°S–45°N and 75°E–145°E, with a spatial resolution of 3.5 × 8 km2. The operational GEMS NO2 retrieval algorithm consists of four major steps (Lee et al., 2024): (1) retrieval of slant column densities (SCDs) using Differential Optical Absorption Spectroscopy (DOAS), which fits the spectrum in the 432–450 nm range; (2) calculation of the AMF through radiative transfer simulations; (3) conversion of SCDs to vertical column densities (VCDs); and (4) separation of tropospheric NO2 VCDs from total columns using stratosphere-troposphere separation (STS) method.

The AMF is calculated as the product of a geometric AMF and the vertical integral of scattering weight multiplied by the shape factor of a priori NO2 profiles. The scattering weight and shape factor denote the sensitivity of the backscattered spectrum to the abundance of the absorber at each layer and a normalized vertical profile of the number density, respectively (Palmer et al., 2001). The derived SCD is divided by the AMF to retrieve the VCD. Then, the tropospheric and stratospheric NO2 VCDs are separated using a new stratospheric and tropospheric NO2 retrieval algorithm (Bucsela et al., 2013).

In this study, we compare two algorithm versions, V2.0 and V3.0. The main updates in the GEMS NO2 retrieval algorithm V3.0 include the correction of errors in the tropospheric and stratospheric shape factors that previously caused biases in the AMF calculations, leading to overestimation of tropospheric NO2 and underestimation of stratospheric NO2 in V2.0. In V3.0, these shape factors were refined, resulting in improved AMF accuracy and reduced such biases. In addition, the tropopause pressure was adjusted to 230 hPa to eliminate the artificial inclusion of tropospheric signals within the stratosphere. Furthermore, a priori NO2 profiles derived from the GEOS-Chem model were revised, and the latest auxiliary input datasets were incorporated, including GEMS L2 aerosol, clouds, surface reflectance, and total ozone data products, ensuring a more accurate representation of radiative transfer conditions in the retrieval process.

2.2. TROPOMI

The TROPOMI, onboard European Space Agency (ESA)’s Sentinel-5 Precursor (S5P) satellite, is a nadir-viewing grating spectrometer covering the ultraviolet (UV), visible (VIS), near-infrared (NIR), and shortwave infrared (SWIR) spectral bands (270–2,385 nm; Veefkind et al., 2012). The TROPOMI provides daily global coverage with a spatial resolution of 3.5 × 5.5 km at nadir-view and local overpass times over Korea between 12:30 and 14:30 local solar time (Veefkind et al., 2012). The operational total NO2 column is retrieved using the DOAS technique within a 405–465 nm fitting window, developed by the Royal Netherlands Meteorological Institute (KNMI) (van Geffen et al., 2022). In this study, we used reprocessed (RPRO) and offline (OFFL) products from processor version 02.04. Pixels with a quality assurance (qa_value) below 0.75 or cloud fraction above 0.3 were excluded (Eskes et al., 2022). The TROPOMI product provides an essential reference for evaluating GEMS retrievals due to its high accuracy and global consistency.

2.3. Pandora

The Pandora spectrometer is a ground-based hyperspectral UV–visible instrument that measures direct-sun and sky radiance between 280 and 530 nm with a spectral resolution of 0.6 nm (Herman et al., 2009). It was developed through collaboration between the NASA and the ESA to complement satellite observations by providing high-precision, long-term measurements of atmospheric trace gases. Pandora applies geometric AMFs to minimize uncertainties in converting slant to vertical columns and has been widely used as a reference for validating satellite NO2 retrievals (Verhoelst et al., 2021; Bae et al., 2025). The measurement data are distributed through the Pandonia Global Network (PGN) (https://www.pandonia-global-network.org/), which has been utilized in several validation studies (Herman et al., 2009; 2018; Choi et al., 2020; Judd et al., 2020; Kim et al., 2023). The retrieval algorithms for total and tropospheric NO₂ columns are well described in Herman et al. (2009). In this study, L2 Pandora data over East Asia during 2022–2023 were used, selecting only measurements with quality flags of 0 and 10 to ensure high data reliability.

3. Results

3.1. Intercomparison between GEMS NO2 V2.0 and V3.0 Retrievals

Fig. 1 shows a pixel-by-pixel comparison of the total, tropospheric, and stratospheric NO2 VCDs retrieved from the GEMS NO2 retrieval algorithms V2.0 and V3.0 for June and December 2023. The two versions show strong internal consistency (R > 0.9 for total columns), confirming the algorithm stability of the GEMS retrieval. However, regression slopes lower than unity indicate that V3.0 generally yields smaller NO2 columns than V2.0. The mean slope for total VCDs decreased from 0.77 to 0.76, corresponding to an overall reduction of about 20–25%. This decrease mainly results from the use of updated a priori NO2 profiles and revised AMF calculations in V3.0.

OGCSBN_2025_v41n5_803_4_f0001.png 이미지

Fig. 1. Comparison of total (a, d), tropospheric (b, e), and stratospheric (c, f) NO2 VCD from GEMS V2.0 and V3.0 for June (upper) and December (lower) 2023, respectively.

For the tropospheric NO2 VCDs, correlations remain high (R ≈ 0.98), and slopes below unity indicate a systematic reduction in retrieved tropospheric NO2 in V3.0 relative to V2.0, with the largest decrease occurring over highly polluted source regions in V3.0. In contrast, the stratospheric NO2 VCDs in V3.0 show higher columns and a tighter scatter (lower root-mean-square errors [RMSE]) than in V2.0. This improvement is attributed to the revision of the vertical distribution (shape factor) and the corresponding AMF recalculations. In V2.0, errors in the tropospheric and stratospheric shape factors caused unrealistic partitioning of NO2 columns-stratospheric VCDs were significantly underestimated, and residual tropospheric signals were often visible in the stratospheric field, indicating imperfect STS. In V3.0, the shape factors were reconstructed using updated model profiles, and AMFs were recalculated for both layers.

Consequently, the stratospheric NO2 columns increased notably, resulting in more physically realistic magnitudes and spatial patterns, particularly at mid- and high latitudes, while the tropospheric NO2 columns decreased, reflecting the removal of erroneously attributed stratospheric sensitivity in V2.0. Overall, the updates in V3.0 substantially improved the vertical separation of NO2, reducing cross-contamination between tropospheric and stratospheric components and providing a more physically consistent representation of atmospheric NO2 distributions over East Asia.

3.2. Assessment of GEMS NO2 V2.0 and V3.0 against TROPOMI Observations

To evaluate the accuracy and consistency of the GEMS NO2 retrievals, GEMS V2.0 and V3.0 were compared with the TROPOMI for total, tropospheric and stratospheric NO2 VCDs in June and December 2023. For the comparison, TROPOMI data were collocated to the GEMS observations by selecting the nearest TROPOMI pixel within a ± 30-minute time window and within 0.1° of each GEMS observation point, using the GEMS spatiotemporal domain as the reference. Fig. 2 shows spatial distributions of total NO2 VCDs from TROPOMI, GEMS V2.0, and GEMS V3.0, together with their differences. Both TROPOMI and GEMS consistently show high NO2 columns over major industrial regions of East Asia, including northeastern China, the Korean Peninsula, and Japan, with substantially increased NO2 columns in December than in June, reflecting the influence of wintertime stagnation. GEMS V2.0 generally overestimates total NO2 relative to TROPOMI, particularly in high NO2 regions. In contrast, GEMS V3.0 shows a marked improvement, significantly reducing these discrepancies and achieving improved spatial consistency with TROPOMI. The reduction in bias primarily results from the use of updated a priori NO2 profiles, revised tropopause pressure, improved AMF calculations, and the adoption of updated auxiliary inputs (aerosol optical depth, cloud, total ozone, and surface reflectance) that more accurately represent the radiative transfer conditions in the lower atmosphere.

OGCSBN_2025_v41n5_803_5_f0001.png 이미지

Fig. 2. Comparison of total NO2 VCD from (a) TROPOMI, (b) GEMS V2.0, (c) the difference between GEMS V2.0 and TROPOMI, (d) GEMS V3.0, and (e) the difference between GEMS V3.0 and TROPOMI for June (upper) and December (lower) 2023, respectively.

The quantitative comparison in Fig. 3 shows that total NO2 VCD in V2.0 systematically overestimates those of TROPOMI, with slopes of 1.31 and 1.98 (R = 0.90 and 0.93 for June and December, respectively). In contrast, V3.0 yields slopes closer to unity (1.05 and 1.48) and reduced RMSE values (0.09–0.30 × 1016 molec. cm–2). Correlation coefficients increase to R > 0.92, confirming that V3.0 achieves substantially better agreement with TROPOMI. These improvements demonstrate that the revised a priori NO2 profiles reduce both systematic and random errors. Fig. 4 shows the comparison of tropospheric and stratospheric NO2 VCDs derived from GEMS V2.0 and V3.0, and TROPOMI for December 2023. The tropospheric NO2 from GEMS V2.0 shows a significant overestimation (slope = 1.42) compared to TROPOMI, which is substantially reduced in V3.0. In particular, the reduced dispersion in the high NO2 values leads to a substantial decrease in the overall bias, resulting in improved consistency with TROPOMI. For the stratospheric NO2, GEMS V2.0 shows a strong underestimation overall, while V3.0 retrievals show higher values with an improved correlation (R = 0.81) and a reduced RMSE from 0.15 to 0.05, indicating a more effective separation of stratospheric and tropospheric signals. This improvement reflects the enhanced AMF calculations resulting from the revised shape factors for both the stratosphere and troposphere and auxiliary data that better capture vertical sensitivity in the lower atmosphere. Overall, the comparisons with TROPOMI clearly demonstrate that V3.0 has significantly improved retrieval performance relative to V2.0. Spatial patterns are more consistent with independent LEO measurements, and the statistical metrics confirm enhanced accuracy and physical consistency in total, tropospheric and stratospheric NO2 columns across East Asia.

OGCSBN_2025_v41n5_803_5_f0002.png 이미지

Fig. 3. Comparison of total NO2 VCD from GEMSV2.0 (a,c) and V3.0 (b, d) with TROPOMI for June (upper) and December (lower) 2023, respectively.

OGCSBN_2025_v41n5_803_5_f0003.png 이미지

Fig. 4. Comparison of tropospheric (a, b) and stratospheric (c, d) NO2 VCD from GEMSV2.0 and V3.0 with TROPOMI for December 2023, respectively.

3.3. Validation with Pandora Ground-Based Measurements

For the comparison between Pandora and GEMS NO2 data products, the Pandora observations were averaged within a ± 15-minute time window around GEMS overpass and within a ± 0.1° spatial radius of each measurement location. To minimize the cloud contamination, the GEMS NO2 data were filtered using the effective cloud fraction (ECF) product, retaining only scenes with ECF < 0.3. The ECF represents the fraction of sunlight reflected by clouds, ranging from 0 (clear) to 1 (fully cloudy) (Choi et al., 2024). Fig. 5 and 6 present comparisons of total NO2 VCDs from GEMS V2.0 and V3.0 with Pandora measurements at six PGN sites representing diverse pollution environments across East Asia: Beijing and Seoul (highly urbanized megacities with heavy anthropogenic emissions), Busan (a coastal industrial city), Bangkok (a dense metropolitan area with strong traffic influence), Tsukuba, and Tokyo (urban background and suburban sites in Japan with relatively lower pollution levels).

OGCSBN_2025_v41n5_803_6_f0001.png 이미지

Fig. 5. Comparison of total NO2 VCD from GEMS V2.0 (upper) and V3.0 (lower) with Pandora at (a) Beijing, (b) Seoul, (c) Busan, (d) Bangkok, (e) Tsukuba, and (f) Tokyo, respectively.

For V2.0, the GEMS retrievals showed both overestimation and underestimation relative to Pandora observations, depending on the pollution characteristics of each site. The regression slopes ranged from 0.35 to 1.05 and correlation coefficients (R) varied between 0.56 and 0.84, indicating substantial site-to-site variability. At highly polluted urban sites such as Beijing and Seoul, the GEMS retrievals tended to significantly overestimate NO2 VCDs, likely due to limitations in the a priori profile and AMF assumption under high aerosol loading and complex surface conditions. In V3.0, these discrepancies were generally reduced across all sites. Although the regression slopes did not show a dramatic improvement compared with V2.0, the overall distribution became closer to the unity, and the RMSE decreased by approximately 20%. The systematic overestimation observed at Beijing and Seoul was substantially mitigated, and the correlation with Pandora measurements improved from 0.84 to 0.90 and from 0.78 to 0.81, respectively. In contrast, at relatively lower pollution sites such as Tsukuba and Tokyo, V3.0 retrievals were also closer to unity and showed smaller biases, confirming that the updated AMF and shape factor configurations perform robustly under various pollution conditions. Overall, the V3.0 retrievals demonstrate improved consistency and accuracy compared with V2.0 retrievals when validated against ground-based Pandora observations across urban, coastal, and background environments, suggesting hat V3.0 retrievals provide more reliable data for air quality monitoring and long-term trend analysis over East Asia.

4. Discussion and Conclusions

In this study, we compared and validated the GEMS NO2 retrieval algorithm V2.0 and V3.0 using independent satellite (TROPOMI) and ground-based Pandora observations. The analysis demonstrates that the algorithmic updates in V3.0—particularly the revision of the a priori NO2 profile shapes, the revised tropopause pressure, and the recalculation of AMFs—have substantially improved the accuracy and physical consistency of the GEMS NO2 retrievals over East Asia.

The intercomparison between V2.0 and V3.0 revealed that the refinement of shape factors effectively resolved the unrealistic vertical partitioning observed in V2.0, where stratospheric NO2 VCDs were severely underestimated and contaminated by tropospheric NO2 signals. As a result, V3.0 showed increased stratospheric NO2 and decreased tropospheric NO2 VCDs, yielding a more realistic representation of the atmospheric NO2 distribution. These improvements led to reduced total NO2 VCDs by approximately 20–25%, consistent with the revised vertical sensitivity and updated auxiliary inputs. The comparison with TROPOMI confirmed that V3.0 mitigated the systematic overestimation over high-emission regions such as northeastern China and the Korean Peninsula. The spatial patterns of V3.0 are more consistent with those of TROPOMI, and the correlation coefficients increased, indicating improved agreement and retrieval stability. Ground-based validation using Pandora observations further confirmed that V3.0 reduced regional biases and RMSE by about 20%, with correlation coefficients increasing across multiple sites. This result highlights the enhanced reliability of the V3.0 product under both polluted and less-polluted atmospheric conditions.

The algorithm improvements in V3.0 not only enhanced the accuracy of the retrievals by revising a priori NO2 profiles and tropopause pressure and updating AMF calculations, but also strengthened the capability of GEMS to capture diurnal and seasonal variability of NO2 over East Asia. These updates are particularly important for interpreting pollution transport, evaluating emission inventories, and improving air-quality modeling. In future algorithm versions, further optimization of the STS method and the incorporation of more dynamic aerosol and cloud treatments are expected to further reduce retrieval uncertainties.

Author Contributions

Conceptualization: Jung Y, Jeong U; Data curation: All authors; Formal analysis: Jung Y, Jeong U; Funding acquisition: Jeong U, Hong H; Investigation, Methodology: Jung Y, Park J, Lee H, Jeong U; Project administration: Jeong U, Hong H, Lee H; Supervision: Jung Y, Lee H, Jeong U; Validation: Jung Y, Jeong U; Writing–original draft: Jung Y, Jeong U; Writing–review & editing: All authors.

Conflicts of Interest

No potential conflict of interest relevant to this article was reported.

Funding

This research was supported by Global - Learning & Academic research institution for Master’s · PhD students, and Postdocs (LAMP) Program of the National Research Foundation of Korea (NRF) grant funded by the Ministry of Education (No. RS-2023-00301702).

Data Availability Statement

GEMS NO2 data products are openly available at https://nesc.nier.go.kr/en/html/datasvc/index.do. Pandora NO2 data that support the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgments

We thank the National Institute of Environmental Research(NIER) of South Korea for providing the GEMS data. We thank PGN instrument PIs and support staff and acknowledge PGN funding for establishing and maintaining the Pandora sites used in this investigation. We appreciate all anonymous referees for their constructive and detailed comments, which helped improve the paper.

Supplementary Materials

None.

References

  1. Bae, K. Song, C.-K., Van Roozendael, M., Richter, A., Wagner, T, Merlaud, A., et al., 2025. Validation of GEMS operational v2.0 total column NO2 and HCHO during the GMAP/SIJAQ campaign. Science of The Total Environment, 974, 179190. https://doi.org/10.1016/j.scitotenv.2025.179190
  2. Beirle, S., Boersma, K. E. Platt, U., Lawrence, M. G., and Wagner, T., 2011. Megacity emissions and lifetimes of nitrogen oxides probed from space. Science, 333(6050), 1737-1739. https://doi.org/10.1126/science.1207824
  3. Beirle, S., Platt, U., Wenig, M., and Wagner, T., 2003. Weekly cycle of NO2 by GOME measurements: A signature of anthropogenic sources. Atmospheric Chemistry and physics, 3(6), 2225-2232. https//doi.org/10.5194/acp-3-2225-2003
  4. Bucsela, E.J., Krotkov, N.A., Celarier, E.A, Lamsal, L.N., Swartz, W.H., Bhartia, et al., 2013. A new stratospheric and tropospheric NO2 retrieval algorithm for nadir-viewing satellite instruments: Applications to OMI. Atmospheric Measurement Techniques, 6(10), 2607-2626, https://doi.org/10.5194/amt-6-2607-2013
  5. Choi, S., Lamsal, L.N., Follette-Cook, M., Joiner, J., Krotkov, N.A., Swartz, W.H., et al., 2020. Assessment of NO2 observations during DISCOVER-AQ and KORUS-AQ field campaigns. Atmospheric Measurement Techniques, 13(5), 2523-2546. https://doi.org/105194/amt-13-2523-2020 105194/amt-13-2523-2020
  6. Choi, Y.-S., Kim, G., Cho, M., Kim, B.-R., and Kwon, M.-J. 2024. Geostationary Emvironment Monitoring Spectrometer (GEMS), Algorithm theoretical basis document, cloud refrieval algorithm (Report No. NIER-2024-04-02-028). Environmental Satellite Center, National Institute of Environmental Research, Ministry of Environment. https://nesc.nier.go.kr/ko/html/satellite/doc/doc.do
  7. Crutzen, P.J., 1979. The role of NO and NO2 in the chemistry of the troposphere and stratosphere. Annual Review of Earth and Planetary Sciences, 7, 443-472. https://doi.org/10.1146/annurev.ea.07.050179.002303
  8. Eskes, H., van Geffen,J., Boersma, F., Eichmann,K.-U.,Apituley, A., Pedergnana, M., et al., 2022. Sentinel-5 precursor/TROPOMI level 2 product user manual – Nitrogen dioxide (Document No. S5P-KNMI-L2-0021-MA). Royal Netherlands Meteorological Institute, Ministry of Infrastructure and Water Management. https://sentinels.copernicus.eu/documents/ 247904/2474726/Sentinel-5P-Level-2-Product-User-Manual-Nitrogen-Dioxide.pdf
  9. Herman, J.. Cede, A., Spinei, E., Mount, G., Tzortziou, M., and Abuhassan, N., 2009. NO2 column amounts from ground-based Pandora and MFDOAS spectrometers using the direct-sun DOAS technique: Intercomparisons and application to OMI validation. Journal of Geophysical Research: Atmospheres, 114, D13307. https://doi.org/10.1029/2009JD011848
  10. Herman, J., Spinei, E., Fried, A., Kim, J., Kim, J., Kim, W., et al., 2018. NO2 and HCHO measurements in Korea from 2012 to 2016 from Pandora spectrometer instruments compared with OMI retrievals and with aircraft measurements during the KORUS-AQ campaign. Atmospheric Measurement Techniques, 11(8), 4583-4603. https://doi.org/10.5194/amt-11-4583-2018
  11. Hong, H., Lee, H., Kim, J., Jeong, U., Ryu, J., and Lee, D. S., 2017. Investigation of simultaneous effects of aerosol properties and aerosol peak height on the air mass factors for space-borne NO2 retrievals. Remote Sensing, 9(3), 208. https://doi.org/10.3390/rs9030208
  12. Judd, L.M., Al-Saadi, J.A., Szykman, ]. J. Valin, L.C., Janz, S.J., Kowalewski, M.G., et al, 2020. Evaluating Sentinel-5 P TROPOMI tropospheric NO2 column densities with airborne and Pandora spectrometers near New York City and Long Island Sound. Atmospheric Measurement Techniques, 13(11), 6113-6140. https://doi.org/10.5194/amt-13-6113-2020
  13. Kim, J., Jeong, U., Ahn, M.H., Kim, J.H., Park, R.J., Lee, H., et al., 2020. New era of air quality monitoring from space: Geostationary Environment Monitoring Spectrometer (GEMS). Bulletin of the American Meteorological Society, 101(1), E1-E22. https://doi.org/10.1175/BAMS-D-18-0013.1
  14. Kim, S., Kim, D., Hong, H., Chang, L.-S., Lee, H., Kim, D.-R., et al., 2023. First-time comparison between NO2 vertical columns from Geostationary Environment Monitoring Spectrometer (GEMS) and Pandora measurements, Atmospheric Measurement Techniques, 16(16), 3959-3972. https://doi.org/10.5194/amt-16-3959-2023
  15. Lee, H., Park, J., Jung, Y., and Hong, H., 2024. Geostationary Environment Monitoring Spectrometer(GEMS),Algorithm Theoretical Basis Document, NO2 Retrieval Algorithm (Technical Report). Environmental Satellite Center, National Institute of Environmental Research, Ministry of Environment. https://nesc.nier.go.kr/ko/html/satellite/doc/doc.do
  16. Levelt, P. F., Hilsenrath, E., Leppelmeier, G. W., van den Oord, G. H. J., Bhartia, P. K., Tamminen, J., et al., 2006. Science objectives of the Ozone Monitoring Instrument. IEEE Transactions on Geoscience and Remote Sensing, 44(5), 1199-1208. https://doi.org/10.1109/TGRS.2006.872336
  17. Palmer, P.1, Jacob, D.], Chance, K., Martin, R.V, Spurr, R.J.D., Kurosu, T.P. et al. 2001. Air mass factor formulation for spectroscopic measurements from satellites: Application to formaldehyde retrievals from the Global Ozone Monitoring Experiment. Journal of Geophysics Research, 106(D13), 14539-14550, https://doi.org/10.1029/2000JD900772
  18. Richter, A., Begoin, M., Hilboll, A., and Burrows, J. P., 2011. An improved NO2 retrieval for the GOME-2 satellite instrument. Atmospheric Measurement Techniques, 4(6), 1147–1159. https://doi.org/10.5194/amt-4-1147-2011
  19. Van Geffen, J., Eskes, H., Compernolle, S., Pinardi, G., Verhoelst, T., Lambert, J.-C., et al., 2022. Sentinel-5P TROPOMI NO2 retrieval: Impact of version v2.2 improvements and comparisons with OMI and ground-based data. Atmospheric Measurement Techniques, 15(7), 2037-2060. https://doi.org/10.5194/amt-15-2037-2022
  20. Veefkind, P., Aben, I., McMullan, K., Forster, H., de Vries, J., Otter, G., et al., 2012. TROPOMI on the ESA Sentinel-5 Precursor: A GMES mission for global observations of the atmospheric composition for climate, air quality and ozone layer applications. Remote Sensing of Environment, 120, 70-83. https://doi.org/10.1016/j.rse.2011.09.027
  21. Verhoelst, T.,Compernolle, S., Pinardi, G., Lambert, J.-C., Eskes, H. J., Eichmann, K.-U., et al., 2021. Ground-based validation of the Copernicus Sentinel-5P TROPOMI NO2 measurements with the NDACC ZSL-DOAS, MAX-DOAS and Pandonia global networks. Atmospheric Measurement Techniques, 14(1), 481-510. https://doi.org10.5194/amt-14-481-2021
  22. Yang, K., Carn, S.A., Ge, C.,Wang, J., and Dickerson, R. R., 2014. Advancing measurements of tropospheric NO2 from space: New algorithm and first global results from OMPS. Geophysical Research Letters, 41(13), 4777-4786. https://doi.org/10,1002/2014GL060136 10,1002/2014GL060136