Articles | Volume 18, issue 18
https://doi.org/10.5194/amt-18-4771-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/amt-18-4771-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
25 Sep 2025
Research article |
| 25 Sep 2025
| 25 Sep 2025
Machine learning data fusion for high spatio-temporal resolution PM2.5
Andrea Porcheddu
CORRESPONDING AUTHOR
Department of Technical Physics, University of Eastern Finland, Kuopio, Finland
Ville Kolehmainen
Department of Technical Physics, University of Eastern Finland, Kuopio, Finland
Timo Lähivaara
Department of Technical Physics, University of Eastern Finland, Kuopio, Finland
Antti Lipponen
Atmospheric Research Centre of Eastern Finland, Finnish Meteorological Institute, Kuopio, Finland
Related authors
Andrea Porcheddu, Ville Kolehmainen, Timo Lähivaara, and Antti Lipponen
Atmos. Meas. Tech., 17, 5747–5764, https://doi.org/10.5194/amt-17-5747-2024, https://doi.org/10.5194/amt-17-5747-2024, 2024
Short summary
Short summary
This study focuses on improving the accuracy of satellite-based PM2.5 retrieval, crucial for monitoring air quality and its impact on health. It employs machine learning to correct the AOD-to-PM2.5 conversion ratio using various data sources. The approach produces high-resolution PM2.5 estimates with improved accuracy. The method is flexible and can incorporate additional training data from different sources, making it a valuable tool for air quality monitoring and epidemiological studies.
Vishnu Nair, Edward Gryspeerdt, Antti Arola, Antti Lipponen, and Timo Virtanen
EGUsphere, https://doi.org/10.5194/egusphere-2025-4272, https://doi.org/10.5194/egusphere-2025-4272, 2025
This preprint is open for discussion and under review for Atmospheric Chemistry and Physics (ACP).
Short summary
Short summary
This work investigates how surface winds affect cloud properties via driving sea salt aerosols, and evaporating water from the ocean surface. Current studies consider snapshots from satellites; here we use observations of evolving clouds which captures feedbacks due to time-dependent adjustments of clouds to aerosol increases. We show that even though sea salt changes droplet sizes, the evaporation from the ocean surface has a stronger impact on cloud properties, hiding the real aerosol effect.
William Wandji Nyamsi, Anders V. Lindfors, Angela Meyer, Antti Lipponen, and Antti Arola
EGUsphere, https://doi.org/10.5194/egusphere-2025-3743, https://doi.org/10.5194/egusphere-2025-3743, 2025
This preprint is open for discussion and under review for Atmospheric Measurement Techniques (AMT).
Short summary
Short summary
This paper proposes a new, fast and accurate method for estimating the cloud optical depth (τc) from photovoltaic (PV) power measurements under overcast sky conditions. The method performs very well with the European Centre for Medium-Range Weather Forecasts (ECMWF) products as inputs describing the state of the atmosphere. The method exhibits a similar or better performance than published state-of-the-art methods when compared to ground and satellite-based τc retrievals serving as reference.
Antti Kukkurainen, Antti Mikkonen, Antti Arola, Antti Lipponen, Ville Kolehmainen, and Neus Sabater
EGUsphere, https://doi.org/10.5194/egusphere-2025-220, https://doi.org/10.5194/egusphere-2025-220, 2025
Short summary
Short summary
HAPI2LIBIS is a new software tool that enhances the capabilities of the radiative transfer model libRadtran. It simplifies high wavelength resolution simulations by using up-to-date molecular data from the HITRAN database and streamlining computations. This tool helps researchers analyze how gases interact with radiation in the Earth's atmosphere and surface, improving atmospheric studies and satellite observations, and making detailed modeling more accurate and accessible.
Timo H. Virtanen, Anu-Maija Sundström, Elli Suhonen, Antti Lipponen, Antti Arola, Christopher O'Dell, Robert R. Nelson, and Hannakaisa Lindqvist
Atmos. Meas. Tech., 18, 929–952, https://doi.org/10.5194/amt-18-929-2025, https://doi.org/10.5194/amt-18-929-2025, 2025
Short summary
Short summary
We find that small particles suspended in the air (aerosols) affect the satellite observations of carbon dioxide (CO2) made by the Orbiting Carbon Observatory-2 satellite instrument. Satellite estimates of CO2 appear to be too high for clean areas and too low for polluted areas. Our results show that CO2 and aerosols are often co-emitted, and this is partly masked out in the current retrievals. Correctly accounting for the aerosol effect is important for CO2 emission estimates by satellites.
Harri Kokkola, Juha Tonttila, Silvia M. Calderón, Sami Romakkaniemi, Antti Lipponen, Aapo Peräkorpi, Tero Mielonen, Edward Gryspeerdt, Timo Henrik Virtanen, Pekka Kolmonen, and Antti Arola
Atmos. Chem. Phys., 25, 1533–1543, https://doi.org/10.5194/acp-25-1533-2025, https://doi.org/10.5194/acp-25-1533-2025, 2025
Short summary
Short summary
Understanding how atmospheric aerosols affect clouds is a scientific challenge. One question is how aerosols affects the amount of cloud water. We used a cloud-scale model to study these effects on marine clouds. The study showed that variations in cloud properties and instrument noise can cause bias in satellite-derived cloud water content. However, our results suggest that for similar weather conditions with well-defined aerosol concentrations, satellite data can reliably track these effects.
Andrea Porcheddu, Ville Kolehmainen, Timo Lähivaara, and Antti Lipponen
Atmos. Meas. Tech., 17, 5747–5764, https://doi.org/10.5194/amt-17-5747-2024, https://doi.org/10.5194/amt-17-5747-2024, 2024
Short summary
Short summary
This study focuses on improving the accuracy of satellite-based PM2.5 retrieval, crucial for monitoring air quality and its impact on health. It employs machine learning to correct the AOD-to-PM2.5 conversion ratio using various data sources. The approach produces high-resolution PM2.5 estimates with improved accuracy. The method is flexible and can incorporate additional training data from different sources, making it a valuable tool for air quality monitoring and epidemiological studies.
Xiaoxia Shang, Antti Lipponen, Maria Filioglou, Anu-Maija Sundström, Mark Parrington, Virginie Buchard, Anton S. Darmenov, Ellsworth J. Welton, Eleni Marinou, Vassilis Amiridis, Michael Sicard, Alejandro Rodríguez-Gómez, Mika Komppula, and Tero Mielonen
Atmos. Chem. Phys., 24, 1329–1344, https://doi.org/10.5194/acp-24-1329-2024, https://doi.org/10.5194/acp-24-1329-2024, 2024
Short summary
Short summary
In June 2019, smoke particles from a Canadian wildfire event were transported to Europe. The long-range-transported smoke plumes were monitored with a spaceborne lidar and reanalysis models. Based on the aerosol mass concentrations estimated from the observations, the reanalysis models had difficulties in reproducing the amount and location of the smoke aerosols during the transport event. Consequently, more spaceborne lidar missions are needed for reliable monitoring of aerosol plumes.
Kalle Nordling, Jukka-Pekka Keskinen, Sami Romakkaniemi, Harri Kokkola, Petri Räisänen, Antti Lipponen, Antti-Ilari Partanen, Jaakko Ahola, Juha Tonttila, Muzaffer Ege Alper, Hannele Korhonen, and Tomi Raatikainen
Atmos. Chem. Phys., 24, 869–890, https://doi.org/10.5194/acp-24-869-2024, https://doi.org/10.5194/acp-24-869-2024, 2024
Short summary
Short summary
Our results show that the global model is stable and it provides meaningful results. This way we can include a physics-based presentation of sub-grid physics (physics which happens on a 100 m scale) in the global model, whose resolution is on a 100 km scale.
Tuuli Miinalainen, Harri Kokkola, Antti Lipponen, Antti-Pekka Hyvärinen, Vijay Kumar Soni, Kari E. J. Lehtinen, and Thomas Kühn
Atmos. Chem. Phys., 23, 3471–3491, https://doi.org/10.5194/acp-23-3471-2023, https://doi.org/10.5194/acp-23-3471-2023, 2023
Short summary
Short summary
We simulated the effects of aerosol emission mitigation on both global and regional radiative forcing and city-level air quality with a global-scale climate model. We used a machine learning downscaling approach to bias-correct the PM2.5 values obtained from the global model for the Indian megacity New Delhi. Our results indicate that aerosol mitigation could result in both improved air quality and less radiative heating for India.
Chao Yan, Yicheng Shen, Dominik Stolzenburg, Lubna Dada, Ximeng Qi, Simo Hakala, Anu-Maija Sundström, Yishuo Guo, Antti Lipponen, Tom V. Kokkonen, Jenni Kontkanen, Runlong Cai, Jing Cai, Tommy Chan, Liangduo Chen, Biwu Chu, Chenjuan Deng, Wei Du, Xiaolong Fan, Xu-Cheng He, Juha Kangasluoma, Joni Kujansuu, Mona Kurppa, Chang Li, Yiran Li, Zhuohui Lin, Yiliang Liu, Yuliang Liu, Yiqun Lu, Wei Nie, Jouni Pulliainen, Xiaohui Qiao, Yonghong Wang, Yifan Wen, Ye Wu, Gan Yang, Lei Yao, Rujing Yin, Gen Zhang, Shaojun Zhang, Feixue Zheng, Ying Zhou, Antti Arola, Johanna Tamminen, Pauli Paasonen, Yele Sun, Lin Wang, Neil M. Donahue, Yongchun Liu, Federico Bianchi, Kaspar R. Daellenbach, Douglas R. Worsnop, Veli-Matti Kerminen, Tuukka Petäjä, Aijun Ding, Jingkun Jiang, and Markku Kulmala
Atmos. Chem. Phys., 22, 12207–12220, https://doi.org/10.5194/acp-22-12207-2022, https://doi.org/10.5194/acp-22-12207-2022, 2022
Short summary
Short summary
Atmospheric new particle formation (NPF) is a dominant source of atmospheric ultrafine particles. In urban environments, traffic emissions are a major source of primary pollutants, but their contribution to NPF remains under debate. During the COVID-19 lockdown, traffic emissions were significantly reduced, providing a unique chance to examine their relevance to NPF. Based on our comprehensive measurements, we demonstrate that traffic emissions alone are not able to explain the NPF in Beijing.
Jaakko Ahola, Tomi Raatikainen, Muzaffer Ege Alper, Jukka-Pekka Keskinen, Harri Kokkola, Antti Kukkurainen, Antti Lipponen, Jia Liu, Kalle Nordling, Antti-Ilari Partanen, Sami Romakkaniemi, Petri Räisänen, Juha Tonttila, and Hannele Korhonen
Atmos. Chem. Phys., 22, 4523–4537, https://doi.org/10.5194/acp-22-4523-2022, https://doi.org/10.5194/acp-22-4523-2022, 2022
Short summary
Short summary
Clouds are important for the climate, and cloud droplets have a significant role in cloud properties. Cloud droplets form when air rises and cools and water vapour condenses on small particles that can be natural or of anthropogenic origin. Currently, the updraft velocity, meaning how fast the air rises, is poorly represented in global climate models. In our study, we show three methods that will improve the depiction of updraft velocity and which properties are vital to updrafts.
Antti Lipponen, Jaakko Reinvall, Arttu Väisänen, Henri Taskinen, Timo Lähivaara, Larisa Sogacheva, Pekka Kolmonen, Kari Lehtinen, Antti Arola, and Ville Kolehmainen
Atmos. Meas. Tech., 15, 895–914, https://doi.org/10.5194/amt-15-895-2022, https://doi.org/10.5194/amt-15-895-2022, 2022
Short summary
Short summary
We have developed a machine-learning-based model that can be used to correct the Sentinel-3 satellite-based aerosol parameter data of the Synergy data product. The strength of the model is that the original satellite data processing does not have to be carried out again but the correction can be carried out with the data already available. We show that the correction significantly improves the accuracy of the satellite aerosol parameters.
Anu Kauppi, Antti Kukkurainen, Antti Lipponen, Marko Laine, Antti Arola, Hannakaisa Lindqvist, and Johanna Tamminen
Atmos. Meas. Tech. Discuss., https://doi.org/10.5194/amt-2021-328, https://doi.org/10.5194/amt-2021-328, 2021
Revised manuscript not accepted
Short summary
Short summary
We present a methodology in Bayesian framework for retrieving atmospheric aerosol optical depth and aerosol type from the pre-computed look-up tables (LUTs). Especially, we consider Bayesian model averaging and uncertainty originating from aerosol model selection and imperfect forward modelling. Our aim is to get more realistic uncertainty estimates. We have applied the methodology to TROPOMI/S5P satellite observations and evaluated the results against ground-based data from the AERONET.
Xiaoxia Shang, Tero Mielonen, Antti Lipponen, Elina Giannakaki, Ari Leskinen, Virginie Buchard, Anton S. Darmenov, Antti Kukkurainen, Antti Arola, Ewan O'Connor, Anne Hirsikko, and Mika Komppula
Atmos. Meas. Tech., 14, 6159–6179, https://doi.org/10.5194/amt-14-6159-2021, https://doi.org/10.5194/amt-14-6159-2021, 2021
Short summary
Short summary
The long-range-transported smoke particles from a Canadian wildfire event were observed with a multi-wavelength Raman polarization lidar and a ceilometer over Kuopio, Finland, in June 2019. The optical properties and the mass concentration estimations were reported for such aged smoke aerosols over northern Europe.
Antti Arola, William Wandji Nyamsi, Antti Lipponen, Stelios Kazadzis, Nickolay A. Krotkov, and Johanna Tamminen
Atmos. Meas. Tech., 14, 4947–4957, https://doi.org/10.5194/amt-14-4947-2021, https://doi.org/10.5194/amt-14-4947-2021, 2021
Short summary
Short summary
Methods to estimate surface UV radiation from satellite measurements offer the only means to obtain global coverage, and the development of satellite-based UV algorithms has been ongoing since the early 1990s. One of the main challenges in this development has been how to account for the overall effect of absorption by atmospheric aerosols. One such method was suggested roughly a decade ago, and in this study we propose further improvements for this kind of approach.
Antti Lipponen, Ville Kolehmainen, Pekka Kolmonen, Antti Kukkurainen, Tero Mielonen, Neus Sabater, Larisa Sogacheva, Timo H. Virtanen, and Antti Arola
Atmos. Meas. Tech., 14, 2981–2992, https://doi.org/10.5194/amt-14-2981-2021, https://doi.org/10.5194/amt-14-2981-2021, 2021
Short summary
Short summary
We have developed a new computational method to post-process-correct the satellite aerosol retrievals. The proposed method combines the conventional satellite aerosol retrievals relying on physics-based models and machine learning. The results show significantly improved accuracy in the aerosol data over the operational satellite data products. The correction can be applied to the existing satellite aerosol datasets with no need to fully reprocess the much larger original radiance data.
Johannes Quaas, Antti Arola, Brian Cairns, Matthew Christensen, Hartwig Deneke, Annica M. L. Ekman, Graham Feingold, Ann Fridlind, Edward Gryspeerdt, Otto Hasekamp, Zhanqing Li, Antti Lipponen, Po-Lun Ma, Johannes Mülmenstädt, Athanasios Nenes, Joyce E. Penner, Daniel Rosenfeld, Roland Schrödner, Kenneth Sinclair, Odran Sourdeval, Philip Stier, Matthias Tesche, Bastiaan van Diedenhoven, and Manfred Wendisch
Atmos. Chem. Phys., 20, 15079–15099, https://doi.org/10.5194/acp-20-15079-2020, https://doi.org/10.5194/acp-20-15079-2020, 2020
Short summary
Short summary
Anthropogenic pollution particles – aerosols – serve as cloud condensation nuclei and thus increase cloud droplet concentration and the clouds' reflection of sunlight (a cooling effect on climate). This Twomey effect is poorly constrained by models and requires satellite data for better quantification. The review summarizes the challenges in properly doing so and outlines avenues for progress towards a better use of aerosol retrievals and better retrievals of droplet concentrations.
Nick Schutgens, Andrew M. Sayer, Andreas Heckel, Christina Hsu, Hiren Jethva, Gerrit de Leeuw, Peter J. T. Leonard, Robert C. Levy, Antti Lipponen, Alexei Lyapustin, Peter North, Thomas Popp, Caroline Poulsen, Virginia Sawyer, Larisa Sogacheva, Gareth Thomas, Omar Torres, Yujie Wang, Stefan Kinne, Michael Schulz, and Philip Stier
Atmos. Chem. Phys., 20, 12431–12457, https://doi.org/10.5194/acp-20-12431-2020, https://doi.org/10.5194/acp-20-12431-2020, 2020
Short summary
Short summary
We intercompare 14 different datasets of satellite observations of aerosol. Such measurements are challenging but also provide the best opportunity to globally observe an atmospheric component strongly related to air pollution and climate change. Our study shows that most datasets perform similarly well on a global scale but that locally errors can be quite different. We develop a technique to estimate satellite errors everywhere, even in the absence of surface reference data.
Cited articles
Belward, A. S., Estes, J. E., and Kline, K. D.: The IGBP-DIS global 1-km land-cover data set DISCover: A project overview, Photogrammetric Engineering and Remote Sensing, 65, 1013–1020, 1999. a
Bessho, K., Date, K., Hayashi, M., Ikeda, A., Imai, T., Inoue, H., Kumagai, Y., Miyakawa, T., Murata, H., Ohno, T., Okuyama, A., Oyama, R., Sasaki, Y., Shimazu, Y., Shimoji, K., Sumida, Y., Suzuki, M., Taniguchi, H., Tsuchiyama, H., Uesawa, D., Yokota, H., and Yoshida, R.: An Introduction to Himawari-8/9 – Japan's New-Generation Geostationary Meteorological Satellites, Journal of the Meteorological Society of Japan. Ser. II, 94, 151–183, https://doi.org/10.2151/jmsj.2016-009, 2016. a
Bishop, C. M. and Bishop, H.: Deep Learning: Foundations and Concepts, Springer International Publishing, Cham, ISBN 978-3-031-45467-7, https://doi.org/10.1007/978-3-031-45468-4, 2024. a, b
Buchard, V., Da Silva, A., Randles, C., Colarco, P., Ferrare, R., Hair, J., Hostetler, C., Tackett, J., and Winker, D.: Evaluation of the surface PM2.5 in Version 1 of the NASA MERRA Aerosol Reanalysis over the United States, Atmospheric Environment, 125, 100–111, 2016. a
Chu, Y., Liu, Y., Li, X., Liu, Z., Lu, H., Lu, Y., Mao, Z., Chen, X., Li, N., Ren, M., Liu, F., Tian, L., Zhu, Z., and Xiang, H.: A Review on Predicting Ground PM2.5 Concentration Using Satellite Aerosol Optical Depth, Atmosphere, 7, https://doi.org/10.3390/atmos7100129, 2016. a
Çiçek, Ö., Abdulkadir, A., Lienkamp, S. S., Brox, T., and Ronneberger, O.: 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation, edited by: Ourselin, S., Joskowicz, L., Sabuncu, M., Unal, G., Wells, W., Medical Image Computing and Computer-Assisted Intervention – MICCAI 2016, MICCAI 2016, Lecture Notes in Computer Science, Springer, Cham., 9901. https://doi.org/10.1007/978-3-319-46723-8_49, 2016. a, b
Cohen, A. J., Brauer, M., Burnett, R., Anderson, H. R., Frostad, J., Estep, K., Balakrishnan, K., Brunekreef, B., Dandona„ L., Dandona, R., Feigin, V., Freedman, G., Hubbell, B., Jobling, A., Kan, H., Knibbs, L., Liu, Y., Martin, R., Morawska, L., Pope III, C. A., Shin, H., Straif, K., Shaddick, G., Thomas, M., van Dingenen, R., van Donkelaar, A., Vos, T., Murray, C. J. L., and Forouzanfar, M. H.: Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015, The Lancet, 389, 1907–1918, 2017. a
Deng, L.: Estimation of PM2.5 Spatial Distribution Based on Kriging Interpolation, in: Proceedings of the First International Conference on Information Sciences, Machinery, Materials and Energy, Atlantis Press, ISBN 978-94-62520-67-7, 1791–1794, https://doi.org/10.2991/icismme-15.2015.370, 2015. a
Fujisada, H., Urai, M., and Iwasaki, A.: Advanced methodology for ASTER DEM generation, IEEE Transactions on Geoscience and Remote Sensing, 49, 5080–5091, 2011. a
Fujisada, H., Urai, M., and Iwasaki, A.: Technical methodology for ASTER global DEM, IEEE Transactions on Geoscience and Remote Sensing, 50, 3725–3736, 2012. a
Goodfellow, I., Bengio, Y., and Courville, A.: Deep Learning, MIT Press, http://www.deeplearningbook.org (last access: 1 December 2024), 2016. a
Health Effects Institute: State of global air 2019, Health Effects Institute, ISSN 2578-6873, 2019. a
Inness, A., Ades, M., Agustí-Panareda, A., Barré, J., Benedictow, A., Blechschmidt, A.-M., Dominguez, J. J., Engelen, R., Eskes, H., Flemming, J., Huijnen, V., Jones, L., Kipling, Z., Massart, S., Parrington, M., Peuch, V.-H., Razinger, M., Remy, S., Schulz, M., and Suttie, M.: The CAMS reanalysis of atmospheric composition, Atmos. Chem. Phys., 19, 3515–3556, https://doi.org/10.5194/acp-19-3515-2019, 2019. a
Keller, C. A., Knowland, K. E., Duncan, B. N., Liu, J., Anderson, D. C., Das, S., Lucchesi, R. A., Lundgren, E. W., Nicely, J. M., Nielsen, E., Ott, L. E., Saunders, E., Strode, S. A., Wales, P. A., Jacob, D. J., and Pawson, S.: Description of the NASA GEOS Composition Forecast Modeling System GEOS-CF v1.0, Journal of Advances in Modeling Earth Systems, 13, e2020MS002413, https://doi.org/10.1029/2020MS002413, 2021. a
Koo, J.-S., Wang, K.-H., Yun, H.-Y., Kwon, H.-Y., and Koo, Y.-S.: Development of PM2.5 Forecast Model Combining ConvLSTM and DNN in Seoul, Atmosphere, 15, 1276, https://doi.org/10.3390/atmos15111276, 2024. a, b
LeCun, Y., Boser, B., Denker, J., Henderson, D., Howard, R., Hubbard, W., and Jackel, L.: Handwritten Digit Recognition with a Back-Propagation Network, in: Advances in Neural Information Processing Systems, edited by: Touretzky, D., vol. 2, Morgan-Kaufmann, https://proceedings.neurips.cc/paper_files/paper/1989/file/53c3bce66e43be4f209556518c2fcb54-Paper.pdf (last access: 1 December 2024), 1989. a
Levy, R. C., Mattoo, S., Munchak, L. A., Remer, L. A., Sayer, A. M., Patadia, F., and Hsu, N. C.: The Collection 6 MODIS aerosol products over land and ocean, Atmos. Meas. Tech., 6, 2989–3034, https://doi.org/10.5194/amt-6-2989-2013, 2013. a
Lipponen, A., Reinvall, J., Väisänen, A., Taskinen, H., Lähivaara, T., Sogacheva, L., Kolmonen, P., Lehtinen, K., Arola, A., and Kolehmainen, V.: Deep-learning-based post-process correction of the aerosol parameters in the high-resolution Sentinel-3 Level-2 Synergy product, Atmos. Meas. Tech., 15, 895–914, https://doi.org/10.5194/amt-15-895-2022, 2022. a, b
Loveland, T. R. and Belward, A.: The international geosphere biosphere programme data and information system global land cover data set (DISCover), Acta Astronautica, 41, 681–689, 1997. a
Lundberg, S. M. and Lee, S.-I.: A Unified Approach to Interpreting Model Predictions, in: Advances in Neural Information Processing Systems 30, edited by: Guyon, I., Luxburg, U. V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., and Garnett, R., Curran Associates, Inc., 4765–4774, http://papers.nips.cc/paper/7062-a-unified-approach-to-interpreting-model-predictions.pdf (last access: 1 December 2024), 2017. a
Michaelides, S., Lane, J., and Kasparis, T.: Effect of Vertical Air Motion on Disdrometer Derived Z-R Coefficients, Atmosphere, 10, 77, https://doi.org/10.3390/atmos10020077, 2019. a
Muthukumar, P., Cocom, E., Nagrecha, K., Comer, D., Burga, I., Taub, J., Calvert, C. F., Holm, J., and Pourhomayoun, M.: Predicting PM2.5 atmospheric air pollution using deep learning with meteorological data and ground-based observations and remote-sensing satellite big data, Air Quality, Atmosphere, & Health, 15, 1221–1234, https://doi.org/10.1007/s11869-021-01126-3, 2022. a
NASA/METI/AIST/Japan Spacesystems and US/Japan ASTER Science Team: ASTER Global Digital Elevation Model V003, distributed by NASA EOSDIS Land Processes DAAC, https://doi.org/10.5067/ASTER/ASTGTM.003, 2019. a
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., Desmaison, A., Köpf, A., Yang, E., DeVito, Z., Raison, M., Tejani, A., Chilamkurthy, S., Steiner, B., Fang, L., Bai, J., and Chintala, S.: PyTorch: An Imperative Style, High-Performance Deep Learning Library, arXiv [preprint], https://doi.org/10.48550/arXiv.1912.01703, 3 December 2019. a
Pope, C. A. I. and Dockery, D. W.: Health Effects of Fine Particulate Air Pollution: Lines that Connect, Journal of the Air & Waste Management Association, 56, 709–742, https://doi.org/10.1080/10473289.2006.10464485, 2006. a
Raissi, M., Perdikaris, P., and Karniadakis, G.: Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations, Journal of Computational Physics, 378, 686–707, https://doi.org/10.1016/j.jcp.2018.10.045, 2019. a
Randles, C. A., da Silva, A., Buchard, V., Colarco, P. R., Darmenov, A. S., Govindaraju, R. C., Smirnov, A., Ferrare, R. A., Hair, J. W., Shinozuka, Y., and Flynn C.: The MERRA-2 aerosol reanalysis, 1980 onward. Part I: System description and data assimilation evaluation, Journal of Climate, 30, 6823–6850, 2017. a, b
Ronneberger, O., Fischer, P., and Brox, T.: U-Net: Convolutional Networks for Biomedical Image Segmentation, edited by: Navab, N., Hornegger, J., Wells, W., Frangi, A., Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, MICCAI 2015, Lecture Notes in Computer Science, Springer, Cham., 9351, https://doi.org/10.1007/978-3-319-24574-4_28, 2015. a
Shi, X., Chen, Z., Wang, H., Yeung, D.-Y., kin Wong, W., and chun Woo, W.: Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting, arXiv [preprint], https://doi.org/10.48550/arXiv.1506.04214, 13 June 2015. a
Sulla-Menashe, D. and Friedl, M. A.: User guide to collection 6 MODIS land cover (MCD12Q1 and MCD12C1) product, USGS, Reston, VA, USA, 1, 18, https://modis.ornl.gov/documentation/guides/MCD12_User_Guide_V6.pdf (last access: 13 April 2023), 2018. a
Tang, D., Zhan, Y., and Yang, F.: A review of machine learning for modeling air quality: Overlooked but important issues, Atmospheric Research, 300, 107261, https://doi.org/10.1016/j.atmosres.2024.107261, 2024. a
Thangavel, P., Park, D., and Lee, Y.-C.: Recent Insights into Particulate Matter (PM2.5)-Mediated Toxicity in Humans: An Overview, International Journal of Environmental Research and Public Health, 19, 7511, https://doi.org/10.3390/ijerph19127511, 2022. a
Wang, Z., Shrestha, R., and Román, M. O.: VIIRS/NPP Lunar BRDF-Adjusted Nighttime Lights Yearly L3 Global 15 arc second Linear Lat Lon Grid [data set], https://doi.org/10.5067/VIIRS/VNP46A4.001, 2020. a
Short summary
This study proposes a novel machine learning method to estimate pollution levels (PM2.5) on urban areas at fine scale. Our model generates hourly PM2.5 maps with high spatial resolution, by combining satellite data, ground measurements, geophysical model data, and different geographical indicators. The model properly accounts for spatial and temporal variability of the urban pollution levels, and can be highly beneficial for air quality monitoring and health protection.
This study proposes a novel machine learning method to estimate pollution levels (PM2.5) on...
