Articles | Volume 25, issue 10
https://doi.org/10.5194/acp-25-5159-2025
© 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/acp-25-5159-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
21 May 2025
Research article |
| 21 May 2025
| 21 May 2025
High-resolution greenhouse gas flux inversions using a machine learning surrogate model for atmospheric transport
Nikhil Dadheech
Department of Atmospheric and Climate Science, University of Washington, Seattle, WA, USA
Tai-Long He
Department of Atmospheric and Climate Science, University of Washington, Seattle, WA, USA
now at: School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, USA
Department of Atmospheric and Climate Science, University of Washington, Seattle, WA, USA
Related authors
Tai-Long He, Nikhil Dadheech, Tammy M. Thompson, and Alexander J. Turner
Geosci. Model Dev., 18, 1661–1671, https://doi.org/10.5194/gmd-18-1661-2025, https://doi.org/10.5194/gmd-18-1661-2025, 2025
Short summary
Short summary
It is computationally expensive to infer greenhouse gas (GHG) emissions using atmospheric observations. This is partly due to the detailed model used to represent atmospheric transport. We demonstrate how a machine learning (ML) model can be used to simulate high-resolution atmospheric transport. This type of ML model will help estimate GHG emissions using dense observations, which are becoming increasingly common with the proliferation of urban monitoring networks and geostationary satellites.
Eric John Mei, Gregory J. Hakim, Max Taniguchi-King, Dominik Stiller, and Alexander J. Turner
EGUsphere, https://doi.org/10.5194/egusphere-2025-3258, https://doi.org/10.5194/egusphere-2025-3258, 2025
This preprint is open for discussion and under review for Atmospheric Chemistry and Physics (ACP).
Short summary
Short summary
Chemistry-climate models are used to investigate how physical climate influences the composition of the atmosphere but are slow and expensive to run. We train a linear inverse model that can replicate the behavior of chemistry-climate models at low computational cost. It captures how large-scale climate features like El Niño affect atmospheric composition and can make accurate forecasts up to a year ahead. This model enables fast hypothesis testing and estimates of past atmospheric composition.
Tai-Long He, Nikhil Dadheech, Tammy M. Thompson, and Alexander J. Turner
Geosci. Model Dev., 18, 1661–1671, https://doi.org/10.5194/gmd-18-1661-2025, https://doi.org/10.5194/gmd-18-1661-2025, 2025
Short summary
Short summary
It is computationally expensive to infer greenhouse gas (GHG) emissions using atmospheric observations. This is partly due to the detailed model used to represent atmospheric transport. We demonstrate how a machine learning (ML) model can be used to simulate high-resolution atmospheric transport. This type of ML model will help estimate GHG emissions using dense observations, which are becoming increasingly common with the proliferation of urban monitoring networks and geostationary satellites.
Matthew S. Johnson, Sofia D. Hamilton, Seongeun Jeong, Yuyan Cui, Dien Wu, Alex Turner, and Marc Fischer
EGUsphere, https://doi.org/10.5194/egusphere-2024-2152, https://doi.org/10.5194/egusphere-2024-2152, 2024
Short summary
Short summary
Satellites, such as NASA’s Orbiting Carbon Observatory-2 and -3 (OCO-2/3), retrieve carbon dioxide (CO2) concentrations which provide vital information for estimating surface CO2 emissions. Here we investigate the ability of OCO-2/3 retrievals to constrain CO2 emissions for the state of California for the major emission sectors (i.e., fossil fuels, net ecosystem exchange, wildfire).
Xueying Yu, Dylan B. Millet, Daven K. Henze, Alexander J. Turner, Alba Lorente Delgado, A. Anthony Bloom, and Jianxiong Sheng
Atmos. Chem. Phys., 23, 3325–3346, https://doi.org/10.5194/acp-23-3325-2023, https://doi.org/10.5194/acp-23-3325-2023, 2023
Short summary
Short summary
We combine satellite measurements with a novel downscaling method to map global methane emissions at 0.1°×0.1° resolution. These fine-scale emission estimates reveal unreported emission hotspots and shed light on the roles of agriculture, wetlands, and fossil fuels for regional methane budgets. The satellite-derived emissions point in particular to missing fossil fuel emissions in the Middle East and to a large emission underestimate in South Asia that appears to be tied to monsoon rainfall.
Tai-Long He, Dylan B. A. Jones, Kazuyuki Miyazaki, Kevin W. Bowman, Zhe Jiang, Xiaokang Chen, Rui Li, Yuxiang Zhang, and Kunna Li
Atmos. Chem. Phys., 22, 14059–14074, https://doi.org/10.5194/acp-22-14059-2022, https://doi.org/10.5194/acp-22-14059-2022, 2022
Short summary
Short summary
We use a deep-learning (DL) model to estimate Chinese NOx emissions by combining satellite analysis and in situ measurements. Our results are consistent with conventional analyses of Chinese NOx emissions. Comparison with mobility data shows that the DL model has a better capability to capture changes in NOx. We analyse Chinese NOx emissions during the COVID-19 pandemic lockdown period. Our results illustrate the potential use of DL as a complementary tool for conventional air quality studies.
Weichao Han, Tai-Long He, Zhaojun Tang, Min Wang, Dylan Jones, and Zhe Jiang
Geosci. Model Dev., 15, 4225–4237, https://doi.org/10.5194/gmd-15-4225-2022, https://doi.org/10.5194/gmd-15-4225-2022, 2022
Short summary
Short summary
We present an application of a hybrid deep learning (DL) model on prediction of surface CO in China from 2015 to 2020, which utilizes both convolutional neural networks and long short-term memory neural networks. The DL model performance is better than a Kalman filter (KF) system in the training period (2005–2018). Furthermore, the DL model demonstrates good temporal extensibility: the mean bias and correlation coefficients are 95.7 ppb and 0.93 in the test period (2019–2020) over eastern China.
Johannes Gensheimer, Alexander J. Turner, Philipp Köhler, Christian Frankenberg, and Jia Chen
Biogeosciences, 19, 1777–1793, https://doi.org/10.5194/bg-19-1777-2022, https://doi.org/10.5194/bg-19-1777-2022, 2022
Short summary
Short summary
We develop a convolutional neural network, named SIFnet, that increases the spatial resolution of SIF from TROPOMI by a factor of 10 to a spatial resolution of 0.005°. SIFnet utilizes coarse SIF observations, together with a broad range of high-resolution auxiliary data. The insights gained from interpretable machine learning techniques allow us to make quantitative claims about the relationships between SIF and other common parameters related to photosynthesis.
Helen L. Fitzmaurice, Alexander J. Turner, Jinsol Kim, Katherine Chan, Erin R. Delaria, Catherine Newman, Paul Wooldridge, and Ronald C. Cohen
Atmos. Chem. Phys., 22, 3891–3900, https://doi.org/10.5194/acp-22-3891-2022, https://doi.org/10.5194/acp-22-3891-2022, 2022
Short summary
Short summary
On-road emissions are thought to vary widely from existing predictions, as the effects of the age of the vehicle fleet, the performance of emission control systems, and variations in speed are difficult to assess under ambient driving conditions. We present an observational approach to characterize on-road emissions and show that the method is consistent with other approaches to within ~ 3 %.
Alexander J. Turner, Philipp Köhler, Troy S. Magney, Christian Frankenberg, Inez Fung, and Ronald C. Cohen
Biogeosciences, 18, 6579–6588, https://doi.org/10.5194/bg-18-6579-2021, https://doi.org/10.5194/bg-18-6579-2021, 2021
Short summary
Short summary
This work builds a high-resolution estimate (500 m) of gross primary productivity (GPP) over the US using satellite measurements of solar-induced chlorophyll fluorescence (SIF) from the TROPOspheric Monitoring Instrument (TROPOMI) between 2018 and 2020. We identify ecosystem-specific scaling factors for estimating gross primary productivity (GPP) from TROPOMI SIF. Extreme precipitation events drive four regional GPP anomalies that account for 28 % of year-to-year GPP differences across the US.
Vikram Khade, Saroja M. Polavarapu, Michael Neish, Pieter L. Houtekamer, Dylan B. A. Jones, Seung-Jong Baek, Tai-Long He, and Sylvie Gravel
Geosci. Model Dev., 14, 2525–2544, https://doi.org/10.5194/gmd-14-2525-2021, https://doi.org/10.5194/gmd-14-2525-2021, 2021
Short summary
Short summary
A new modeling system has been developed at Environment and Climate Change Canada to ingest observations of carbon monoxide (CO) into a coupled weather and constituent transport model. We show that accounting for the uncertainty in surface flux leads to a better estimate of CO distributions. The benefit of assimilating observations from different simulated networks varies with region. This is the first step towards developing a state and flux estimation system for greenhouse gases.
Alexander J. Turner, Philipp Köhler, Troy S. Magney, Christian Frankenberg, Inez Fung, and Ronald C. Cohen
Biogeosciences, 17, 405–422, https://doi.org/10.5194/bg-17-405-2020, https://doi.org/10.5194/bg-17-405-2020, 2020
Short summary
Short summary
We present the highest resolution solar-induced chlorophyll fluorescence (SIF) dataset from satellite measurements, providing previously unobservable phenomena related to plant photosynthesis. We find a strong correspondence between TROPOMI SIF and AmeriFlux GPP. We then observe a double peak in the seasonality of California's photosynthesis, not seen by traditional vegetation indices (e.g., MODIS). This is further corroborated by EOF/PC analysis.
Jacob K. Hedelius, Tai-Long He, Dylan B. A. Jones, Bianca C. Baier, Rebecca R. Buchholz, Martine De Mazière, Nicholas M. Deutscher, Manvendra K. Dubey, Dietrich G. Feist, David W. T. Griffith, Frank Hase, Laura T. Iraci, Pascal Jeseck, Matthäus Kiel, Rigel Kivi, Cheng Liu, Isamu Morino, Justus Notholt, Young-Suk Oh, Hirofumi Ohyama, David F. Pollard, Markus Rettinger, Sébastien Roche, Coleen M. Roehl, Matthias Schneider, Kei Shiomi, Kimberly Strong, Ralf Sussmann, Colm Sweeney, Yao Té, Osamu Uchino, Voltaire A. Velazco, Wei Wang, Thorsten Warneke, Paul O. Wennberg, Helen M. Worden, and Debra Wunch
Atmos. Meas. Tech., 12, 5547–5572, https://doi.org/10.5194/amt-12-5547-2019, https://doi.org/10.5194/amt-12-5547-2019, 2019
Short summary
Short summary
We seek ways to improve the accuracy of column measurements of carbon monoxide (CO) – an important tracer of pollution – made from the MOPITT satellite instrument. We devise a filtering scheme which reduces the scatter and also eliminates bias among the MOPITT detectors. Compared to ground-based observations, MOPITT measurements are about 6 %–8 % higher. When MOPITT data are implemented in a global assimilation model, they tend to reduce the model mismatch with aircraft measurements.
Daniel H. Cusworth, Daniel J. Jacob, Jian-Xiong Sheng, Joshua Benmergui, Alexander J. Turner, Jeremy Brandman, Laurent White, and Cynthia A. Randles
Atmos. Chem. Phys., 18, 16885–16896, https://doi.org/10.5194/acp-18-16885-2018, https://doi.org/10.5194/acp-18-16885-2018, 2018
Short summary
Short summary
Methane emissions from oil/gas fields originate from a large number of small and densely clustered point sources. We examine the potential of recently launched or planned satellites to locate these high-mode emitters through measurements of atmospheric methane. We find that the recently launched TROPOMI and the planned GeoCARB instruments are successful at locating high-emitting sources for fields of 20-50 emitters within the 50 × 50 km2 geographic domain but are unsuccessful for denser fields.
Jian-Xiong Sheng, Daniel J. Jacob, Alexander J. Turner, Joannes D. Maasakkers, Joshua Benmergui, A. Anthony Bloom, Claudia Arndt, Ritesh Gautam, Daniel Zavala-Araiza, Hartmut Boesch, and Robert J. Parker
Atmos. Chem. Phys., 18, 12257–12267, https://doi.org/10.5194/acp-18-12257-2018, https://doi.org/10.5194/acp-18-12257-2018, 2018
Short summary
Short summary
Analysis of 7 years (2010–2016) of GOSAT methane trends over Canada, the contiguous US, and Mexico suggests that US methane emissions increased by 2.5 ± 1.4 % a−1 over the 7-year period, with contributions from both oil–gas systems and livestock in the Midwest. Mexican emissions show a decrease that can be attributed to a decreasing cattle population. Canadian emissions show year-to-year variability driven by wetland emissions and correlated with wetland areal extent.
Alexander J. Turner, Daniel J. Jacob, Joshua Benmergui, Jeremy Brandman, Laurent White, and Cynthia A. Randles
Atmos. Chem. Phys., 18, 8265–8278, https://doi.org/10.5194/acp-18-8265-2018, https://doi.org/10.5194/acp-18-8265-2018, 2018
Short summary
Short summary
We conduct a 1-week WRF-STILT simulation to generate methane column footprints at 1.3 km spatial resolution and hourly temporal resolution over the Barnett Shale. We find that a week of TROPOMI observations should provide regional (~30 km) information on temporally invariant sources and GeoCARB should provide information on temporally invariant sources at 2–7 km spatial resolution. An instrument precision better than 6 ppb is an important threshold for achieving fine resolution of emissions.
Jian-Xiong Sheng, Daniel J. Jacob, Alexander J. Turner, Joannes D. Maasakkers, Melissa P. Sulprizio, A. Anthony Bloom, Arlyn E. Andrews, and Debra Wunch
Atmos. Chem. Phys., 18, 6483–6491, https://doi.org/10.5194/acp-18-6483-2018, https://doi.org/10.5194/acp-18-6483-2018, 2018
Short summary
Short summary
We use observations of boundary layer methane from the SEAC4RS aircraft campaign over the Southeast US to estimate methane emissions in that region. Our results suggest that the EPA inventory is regionally unbiased but there are large local biases, suggesting variable emission factors. Our results also suggest that the choice of landcover map is the dominant source of error for wetland emission estimates.
A. Anthony Bloom, Kevin W. Bowman, Meemong Lee, Alexander J. Turner, Ronny Schroeder, John R. Worden, Richard Weidner, Kyle C. McDonald, and Daniel J. Jacob
Geosci. Model Dev., 10, 2141–2156, https://doi.org/10.5194/gmd-10-2141-2017, https://doi.org/10.5194/gmd-10-2141-2017, 2017
Short summary
Short summary
Wetland emissions are a principal source of uncertainty in the global atmospheric methane budget due to poor knowledge of wetland processes. We construct a wetland methane emission and uncertainty dataset for use in global atmospheric methane models. Our wetland model ensemble is based on static wetland maps, satellite-derived inundation and carbon cycle models. The ensemble performs favourably against regional flux estimates and atmospheric methane measurements relative to previous studies.
Michael Buchwitz, Oliver Schneising, Maximilian Reuter, Jens Heymann, Sven Krautwurst, Heinrich Bovensmann, John P. Burrows, Hartmut Boesch, Robert J. Parker, Peter Somkuti, Rob G. Detmers, Otto P. Hasekamp, Ilse Aben, André Butz, Christian Frankenberg, and Alexander J. Turner
Atmos. Chem. Phys., 17, 5751–5774, https://doi.org/10.5194/acp-17-5751-2017, https://doi.org/10.5194/acp-17-5751-2017, 2017
Short summary
Short summary
Methane is an important greenhouse gas and increasing atmospheric concentrations result in global warming. We present a simple method to derive annual methane emission estimates of methane hotspot areas from satellite data. We present results for four source areas. We found that our estimates are in good agreement with other studies/data sets for the Four Corners region in the USA and for Azerbaijan but we also found higher emissions for parts of California and Turkmenistan.
Whitney Bader, Benoît Bovy, Stephanie Conway, Kimberly Strong, Dan Smale, Alexander J. Turner, Thomas Blumenstock, Chris Boone, Martine Collaud Coen, Ancelin Coulon, Omaira Garcia, David W. T. Griffith, Frank Hase, Petra Hausmann, Nicholas Jones, Paul Krummel, Isao Murata, Isamu Morino, Hideaki Nakajima, Simon O'Doherty, Clare Paton-Walsh, John Robinson, Rodrigue Sandrin, Matthias Schneider, Christian Servais, Ralf Sussmann, and Emmanuel Mahieu
Atmos. Chem. Phys., 17, 2255–2277, https://doi.org/10.5194/acp-17-2255-2017, https://doi.org/10.5194/acp-17-2255-2017, 2017
Short summary
Short summary
An increase of 0.31 ± 0.03 % year−1 of atmospheric methane is reported using 10 years of solar observations performed at 10 ground-based stations since 2005. These trend agree with a GEOS-Chem-tagged simulation that accounts for the contribution of each emission source and one sink in the total methane. The GEOS-Chem simulation shows that anthropogenic emissions from coal mining and gas and oil transport and exploration have played a major role in the increase methane since 2005.
Daniel J. Jacob, Alexander J. Turner, Joannes D. Maasakkers, Jianxiong Sheng, Kang Sun, Xiong Liu, Kelly Chance, Ilse Aben, Jason McKeever, and Christian Frankenberg
Atmos. Chem. Phys., 16, 14371–14396, https://doi.org/10.5194/acp-16-14371-2016, https://doi.org/10.5194/acp-16-14371-2016, 2016
Short summary
Short summary
Methane is a greenhouse gas emitted by a range of natural and anthropogenic sources. Atmospheric methane has been measured continuously from space since 2003, and new instruments are planned to launch in the near future that will greatly expand the capabilities of space-based observations. We review the value of current, future, and proposed satellite observations to better quantify methane emissions from the global scale down to the scale of point sources.
Alexander J. Turner, Alexis A. Shusterman, Brian C. McDonald, Virginia Teige, Robert A. Harley, and Ronald C. Cohen
Atmos. Chem. Phys., 16, 13465–13475, https://doi.org/10.5194/acp-16-13465-2016, https://doi.org/10.5194/acp-16-13465-2016, 2016
Short summary
Short summary
Our paper investigates the ability of different types of observational networks to estimate urban CO2 emissions. We have quantified the trade-off between precision and network density for estimating urban greenhouse gas emissions. Our results show that different observing systems may fall into noise- or site-limited regimes where reducing the uncertainty in the estimated emissions is governed by a single factor.
Alexis A. Shusterman, Virginia E. Teige, Alexander J. Turner, Catherine Newman, Jinsol Kim, and Ronald C. Cohen
Atmos. Chem. Phys., 16, 13449–13463, https://doi.org/10.5194/acp-16-13449-2016, https://doi.org/10.5194/acp-16-13449-2016, 2016
Short summary
Short summary
We describe the design of and first results from the BErkeley Atmospheric CO2 Observation Network, a distributed instrument of 28 CO2 sensors stationed across and around the city of Oakland, California at ~ 2 km intervals. We evaluate the network via 4 performance parameters (cost, reliability, precision, systematic uncertainty) and find this high density technique to be sufficiently cost-effective and rigorous to inform understanding of small-scale urban emissions relevant to climate regulation.
Zeli Tan, Qianlai Zhuang, Daven K. Henze, Christian Frankenberg, Ed Dlugokencky, Colm Sweeney, Alexander J. Turner, Motoki Sasakawa, and Toshinobu Machida
Atmos. Chem. Phys., 16, 12649–12666, https://doi.org/10.5194/acp-16-12649-2016, https://doi.org/10.5194/acp-16-12649-2016, 2016
Short summary
Short summary
Methane emissions from the pan-Arctic could be important in understanding the global carbon cycle but are still poorly constrained to date. This study demonstrated that satellite retrievals can be used to reduce the uncertainty of the estimates of these emissions. We also provided additional evidence for the existence of large methane emissions from pan-Arctic lakes in the Siberian yedoma permafrost region. We found that biogeochemical models should be improved for better estimates.
Nicolas Bousserez, Daven K. Henze, Brigitte Rooney, Andre Perkins, Kevin J. Wecht, Alexander J. Turner, Vijay Natraj, and John R. Worden
Atmos. Chem. Phys., 16, 6175–6190, https://doi.org/10.5194/acp-16-6175-2016, https://doi.org/10.5194/acp-16-6175-2016, 2016
Short summary
Short summary
This work provides new insight into the observational constraints provided by current low-Earth orbit (LEO) and future potential geostationary (GEO) satellite missions on methane emissions in North America. Using efficient numerical tools, the information content (error reductions, spatial resolution of the constraints) of methane inversions using different instrument configurations (TIR, SWIR and multi-spectral) was estimated at model grid-scale resolution (0.5° × 0.7°).
J. R. Worden, A. J. Turner, A. Bloom, S. S. Kulawik, J. Liu, M. Lee, R. Weidner, K. Bowman, C. Frankenberg, R. Parker, and V. H. Payne
Atmos. Meas. Tech., 8, 3433–3445, https://doi.org/10.5194/amt-8-3433-2015, https://doi.org/10.5194/amt-8-3433-2015, 2015
Short summary
Short summary
Here we demonstrate the potential for estimating lower tropospheric CH4 concentrations through the combination of free-tropospheric methane measurements from the Aura Tropospheric Emission Spectrometer (TES) and XCH4 (dry-mole air fraction of methane) from the Greenhouse Gases Observing Satellite - Thermal And Near-infrared for carbon Observation (GOSAT TANSO).
A. J. Turner and D. J. Jacob
Atmos. Chem. Phys., 15, 7039–7048, https://doi.org/10.5194/acp-15-7039-2015, https://doi.org/10.5194/acp-15-7039-2015, 2015
A. J. Turner, D. J. Jacob, K. J. Wecht, J. D. Maasakkers, E. Lundgren, A. E. Andrews, S. C. Biraud, H. Boesch, K. W. Bowman, N. M. Deutscher, M. K. Dubey, D. W. T. Griffith, F. Hase, A. Kuze, J. Notholt, H. Ohyama, R. Parker, V. H. Payne, R. Sussmann, C. Sweeney, V. A. Velazco, T. Warneke, P. O. Wennberg, and D. Wunch
Atmos. Chem. Phys., 15, 7049–7069, https://doi.org/10.5194/acp-15-7049-2015, https://doi.org/10.5194/acp-15-7049-2015, 2015
A. J. Turner, A. M. Fiore, L. W. Horowitz, and M. Bauer
Atmos. Chem. Phys., 13, 565–578, https://doi.org/10.5194/acp-13-565-2013, https://doi.org/10.5194/acp-13-565-2013, 2013
Related subject area
Subject: Gases | Research Activity: Machine Learning | Altitude Range: Troposphere | Science Focus: Chemistry (chemical composition and reactions)
Multi-Machine Learning Approaches to Modeling Small-Scale Source Attribution of Ozone Formation
Technical note: Towards atmospheric compound identification in chemical ionization mass spectrometry with pesticide standards and machine learning
Identifying Drivers of Surface Ozone Bias in Global Chemical Reanalysis with Explainable Machine Learning
Diagnosing ozone–NOx–VOC–aerosol sensitivity and uncovering causes of urban–nonurban discrepancies in Shandong, China, using transformer-based estimations
A machine learning approach to downscale EMEP4UK: analysis of UK ozone variability and trends
Automated detection and monitoring of methane super-emitters using satellite data
Spatiotemporal modeling of air pollutant concentrations in Germany using machine learning
Estimating nitrogen and sulfur deposition across China during 2005 to 2020 based on multiple statistical models
Technical note: Improving the European air quality forecast of the Copernicus Atmosphere Monitoring Service using machine learning techniques
Zheng Xiao, Yifeng Lu, and Guangli Xiu
EGUsphere, https://doi.org/10.5194/egusphere-2025-160, https://doi.org/10.5194/egusphere-2025-160, 2025
Short summary
Short summary
This study innovates air pollution tracking in industrial areas by merging AI with traditional methods. Analyzing three years of data from a Shanghai chemical park, we identified ozone pollution sources and seasonal variations, revealing that chemical solvents and fuel vapors are key contributors. Our method enables faster, accurate source identification, aiding better air quality decisions in rapidly developing regions.
Federica Bortolussi, Hilda Sandström, Fariba Partovi, Joona Mikkilä, Patrick Rinke, and Matti Rissanen
Atmos. Chem. Phys., 25, 685–704, https://doi.org/10.5194/acp-25-685-2025, https://doi.org/10.5194/acp-25-685-2025, 2025
Short summary
Short summary
Chemical ionization mass spectrometry (CIMS) is widely used in atmospheric chemistry studies. We still have a limited understanding of the complex functioning of the instrument; therefore, we applied machine learning to provide insights from CIMS analyses. We were able to predict both detection and signal intensity with a fair error, and we found out the most important structural fragments for negative ionization schemes (NH and OH) and positive ones (nitrogen-containing groups).
Kazuyuki Miyazaki, Yuliya Marchetti, James Montgomery, Steven Lu, and Kevin Bowman
EGUsphere, https://doi.org/10.5194/egusphere-2024-3753, https://doi.org/10.5194/egusphere-2024-3753, 2025
Short summary
Short summary
This study employs explainable machine learning to analyze the causes of significant biases in surface ozone estimates from chemical reanalysis. By analyzing global observations and chemical reanalysis outputs, key bias drivers such as meteorological conditions and precursor emissions were identified. This provides actionable insights to improve chemical transport models, observation systems, and emissions inventories, ultimately enhancing ozone reanalysis for better air pollution management.
Chenliang Tao, Yanbo Peng, Qingzhu Zhang, Yuqiang Zhang, Bing Gong, Qiao Wang, and Wenxing Wang
Atmos. Chem. Phys., 24, 4177–4192, https://doi.org/10.5194/acp-24-4177-2024, https://doi.org/10.5194/acp-24-4177-2024, 2024
Short summary
Short summary
We developed a novel transformer framework to bridge the sparse surface monitoring for inferring ozone–NOx–VOC–aerosol sensitivity and their urban–nonurban discrepancies at a finer scale with implications for improving our understanding of ozone variations. The change in urban–rural disparities in ozone was dominated by PM2.5 from 2019 to 2020. An aerosol-inhibited regime on top of the two traditional NOx- and VOC-limited regimes was identified in Jiaodong Peninsula, Shandong, China.
Lily Gouldsbrough, Ryan Hossaini, Emma Eastoe, Paul J. Young, and Massimo Vieno
Atmos. Chem. Phys., 24, 3163–3196, https://doi.org/10.5194/acp-24-3163-2024, https://doi.org/10.5194/acp-24-3163-2024, 2024
Short summary
Short summary
High-resolution spatial fields of surface ozone are used to understand spikes in ozone concentration and predict their impact on public health. Such fields are routinely output from complex mathematical models for atmospheric conditions. These outputs are on a coarse spatial resolution and the highest concentrations tend to be biased. Using a novel data-driven machine learning methodology, we show how such output can be corrected to produce fields with both lower bias and higher resolution.
Berend J. Schuit, Joannes D. Maasakkers, Pieter Bijl, Gourav Mahapatra, Anne-Wil van den Berg, Sudhanshu Pandey, Alba Lorente, Tobias Borsdorff, Sander Houweling, Daniel J. Varon, Jason McKeever, Dylan Jervis, Marianne Girard, Itziar Irakulis-Loitxate, Javier Gorroño, Luis Guanter, Daniel H. Cusworth, and Ilse Aben
Atmos. Chem. Phys., 23, 9071–9098, https://doi.org/10.5194/acp-23-9071-2023, https://doi.org/10.5194/acp-23-9071-2023, 2023
Short summary
Short summary
Using two machine learning models, which were trained on TROPOMI methane satellite data, we detect 2974 methane plumes, so-called super-emitters, in 2021. We detect methane emissions globally related to urban areas or landfills, coal mining, and oil and gas production. Using our monitoring system, we identify 94 regions with frequent emissions. For 12 locations, we target high-resolution satellite instruments to enlarge and identify the exact infrastructure responsible for the emissions.
Vigneshkumar Balamurugan, Jia Chen, Adrian Wenzel, and Frank N. Keutsch
Atmos. Chem. Phys., 23, 10267–10285, https://doi.org/10.5194/acp-23-10267-2023, https://doi.org/10.5194/acp-23-10267-2023, 2023
Short summary
Short summary
In this study, machine learning models are employed to model NO2 and O3 concentrations. We employed a wide range of sources of data, including meteorological and column satellite measurements, to model NO2 and O3 concentrations. The spatial and temporal variability, and their drivers, were investigated. Notably, the machine learning model established the relationship between NOx and O3. Despite the fact that metropolitan regions are NO2 hotspots, rural areas have high O3 concentrations.
Kaiyue Zhou, Wen Xu, Lin Zhang, Mingrui Ma, Xuejun Liu, and Yu Zhao
Atmos. Chem. Phys., 23, 8531–8551, https://doi.org/10.5194/acp-23-8531-2023, https://doi.org/10.5194/acp-23-8531-2023, 2023
Short summary
Short summary
We developed a dataset of the long-term (2005–2020) variabilities of China’s nitrogen and sulfur deposition, with multiple statistical models that combine available observations and chemistry transport modeling. We demonstrated the strong impact of human activities and national pollution control actions on the spatiotemporal changes in deposition and indicated a relatively small benefit of emission abatement on deposition (and thereby ecological risk) for China compared to Europe and the USA.
Jean-Maxime Bertrand, Frédérik Meleux, Anthony Ung, Gaël Descombes, and Augustin Colette
Atmos. Chem. Phys., 23, 5317–5333, https://doi.org/10.5194/acp-23-5317-2023, https://doi.org/10.5194/acp-23-5317-2023, 2023
Short summary
Short summary
Post-processing methods based on machine learning algorithms were applied to refine the forecasts of four key pollutants at monitoring sites across Europe. Performances show significant improvements compared to those of the deterministic model raw outputs. Taking advantage of the large modelling domain extension, an innovative
globalapproach is proposed to drastically reduce the period necessary to train the models and thus facilitate the implementation in an operational context.
Cited articles
Asimow, N. G., Turner, A. J., and Cohen, R. C.: Sustained Reductions of Bay Area CO2 Emissions 2018–2022, Environ. Sci. Technol., 58, 6586–6594, https://doi.org/10.1021/acs.est.3c09642, 2024. a
Bishop, C. M.: Pattern Recognition and Machine Learning, Springer Science+Business Media, LLC, ISBN 978-0-387-31073-2, 2006. a
Bovensmann, H., Buchwitz, M., Burrows, J. P., Reuter, M., Krings, T., Gerilowski, K., Schneising, O., Heymann, J., Tretner, A., and Erzinger, J.: A remote sensing technique for global monitoring of power plant CO2 emissions from space and related applications, Atmos. Meas. Tech., 3, 781–811, https://doi.org/10.5194/amt-3-781-2010, 2010. a
Brandt, A. R., Heath, G. A., Kort, E. A., O'Sullivan, F., Pétron, G., Jordaan, S. M., Tans, P., Wilcox, J., Gopstein, A. M., Arent, D., Wofsy, S., Brown, N. J., Bradley, R., Stucky, G. D., Eardley, D., and Harriss, R.: Methane Leaks from North American Natural Gas Systems, Science, 343, 733–735, https://doi.org/10.1126/science.1247045, 2014. a, b
Brecht, R., Bakels, L., Bihlo, A., and Stohl, A.: Improving trajectory calculations by FLEXPART 10.4+ using single-image super-resolution, Geosci. Model Dev., 16, 2181–2192, https://doi.org/10.5194/gmd-16-2181-2023, 2023. a
Cartwright, L., Zammit-Mangion, A., and Deutscher, N. M.: Emulation of greenhouse-gas sensitivities using variational autoencoders, Environmetrics, 34, e2754, https://doi.org/10.1002/env.2754, 2023. a, b
Chen, Y., Sherwin, E. D., Berman, E. S., Jones, B. B., Gordon, M. P., Wetherley, E. B., Kort, E. A., and Brandt, A. R.: Quantifying Regional Methane Emissions in the New Mexico Permian Basin with a Comprehensive Aerial Survey, Environ. Sci. Technol., 56, 4317–4323, https://doi.org/10.1021/acs.est.1c06458, 2022. a, b
Cohen Research: Berkeley Environmental Air-quality & CO2 Network (BEACO2N), University of California Berkeley [data set], http://beacon.berkeley.edu, last access: 15 May 2025. a
Crisp, D., Fisher, B. M., O'Dell, C., Frankenberg, C., Basilio, R., Bösch, H., Brown, L. R., Castano, R., Connor, B., Deutscher, N. M., Eldering, A., Griffith, D., Gunson, M., Kuze, A., Mandrake, L., McDuffie, J., Messerschmidt, J., Miller, C. E., Morino, I., Natraj, V., Notholt, J., O'Brien, D. M., Oyafuso, F., Polonsky, I., Robinson, J., Salawitch, R., Sherlock, V., Smyth, M., Suto, H., Taylor, T. E., Thompson, D. R., Wennberg, P. O., Wunch, D., and Yung, Y. L.: The ACOS CO2 retrieval algorithm – Part II: Global
Cusworth, D. H., Thorpe, A. K., Ayasse, A. K., Stepp, D., Heckler, J., Asner, G. P., Miller, C. E., Yadav, V., Chapman, J. W., Eastwood, M. L., Green, R. O., Hmiel, B., Lyon, D. R., and Duren, R. M.: Strong methane point sources contribute a disproportionate fraction of total emissions across multiple basins in the United States, P. Natl. Acad. Sci. USA, 119, e2202338 119, https://doi.org/10.1073/pnas.2202338119, 2022. a, b
Dadheech, N.: High-resolution greenhouse gas flux inversions using a machine learning surrogate model, Zenodo [data set and code], https://doi.org/10.5281/zenodo.13750963, 2024. a
Duren, R. M., Thorpe, A. K., Foster, K. T., Rafiq, T., Hopkins, F. M., Yadav, V., Bue, B. D., Thompson, D. R., Conley, S., Colombi, N. K., Frankenberg, C., McCubbin, I. B., Eastwood, M. L., Falk, M., Herner, J. D., Croes, B. E., Green, R. O., and Miller, C. E.: California's methane super-emitters, Nature, 575, 180–184, https://doi.org/10.1038/s41586-019-1720-3, 2019. a, b
Eldering, A., Boland, S., Solish, B., Crisp, D., Kahn, P., and Gunson, M.: High precision atmospheric CO2 measurements from space: The design and implementation of OCO-2, in: 2012 IEEE aerospace conference, Big Sky, MT, USA, 3–10 March 2012, IEEE, https://doi.org/10.1109/AERO.2012.6187176, pp. 1–10, 2012. a
Eldering, A., Taylor, T. E., O'Dell, C. W., and Pavlick, R.: The OCO-3 mission: measurement objectives and expected performance based on 1 year of simulated data, Atmos. Meas. Tech., 12, 2341–2370, https://doi.org/10.5194/amt-12-2341-2019, 2019. a
Fillola, E., Santos-Rodriguez, R., Manning, A., O'Doherty, S., and Rigby, M.: A machine learning emulator for Lagrangian particle dispersion model footprints: a case study using NAME, Geosci. Model Dev., 16, 1997–2009, https://doi.org/10.5194/gmd-16-1997-2023, 2023. a, b, c
Frankenberg, C., Thorpe, A. K., Thompson, D. R., Hulley, G., Kort, E. A., Vance, N., Borchardt, J., Krings, T., Gerilowski, K., Sweeney, C., Conley, S., Bue, B. D., Aubrey, A. D., Hook, S., and Green, R. O.: Airborne methane remote measurements reveal heavy-tail flux distribution in Four Corners region, P. Natl. Acad. Sci. USA, 113, 9734–9739, https://doi.org/10.1073/pnas.1605617113, 2016. a, b
Guo, W., Shi, Y., Liu, Y., and Su, M.: CO2 emissions retrieval from coal-fired power plants based on OCO-2/3 satellite observations and a Gaussian plume model, J. Clean. Prod., 397, 136 525, https://doi.org/10.1016/j.jclepro.2023.136525, 2023. a
Hammerling, D. M., Michalak, A. M., and Kawa, S. R.: Mapping of CO2 at high spatiotemporal resolution using satellite observations: Global distributions from OCO-2, J. Geophys. Res.-Atmos., 117, D06306, https://doi.org/10.1029/2011JD017015, 2012. a
He, T.-L.: STILT footprints data set 2, Zenodo [data set], https://doi.org/10.5281/zenodo.12803736, 2024. a
He, T.-L., Boyd, R. J., Varon, D. J., and Turner, A. J.: Increased methane emissions from oil and gas following the Soviet Union's collapse, P. Natl. Acad. Sci. USA, 121, e2314600121, https://doi.org/10.1073/pnas.2314600121, 2024. a
Hutyra, L. R., Duren, R., Gurney, K. R., Grimm, N., Kort, E. A., Larson, E., and Shrestha, G.: Urbanization and the carbon cycle: Current capabilities and research outlook from the natural sciences perspective, Earths Future, 2, 473–495, https://doi.org/10.1002/2014EF000255, 2014. a
IPCC: Climate change 2023: synthesis report. Contribution of working groups I, II and III to the sixth assessment report of the intergovernmental panel on climate change, The Australian National University, https://doi.org/10.59327/IPCC/AR6-9789291691647.001, 2023. a
Janardanan, R., Maksyutov, S., Oda, T., Saito, M., Kaiser, J. W., Ganshin, A., Stohl, A., Matsunaga, T., Yoshida, Y., and Yokota, T.: Comparing GOSAT observations of localized CO2 enhancements by large emitters with inventory-based estimates, Geophys. Res. Lett., 43, 3486–3493, 2016. a
Janardanan, R., Maksyutov, S., Wang, F., Nayagam, L., Sahu, S. K., Mangaraj, P., Saunois, M., Lan, X., and Matsunaga, T.: Country-level methane emissions and their sectoral trends during 2009–2020 estimated by high-resolution inversion of GOSAT and surface observations, Environ. Res. Lett., 19, 034007, https://doi.org/10.1088/1748-9326/ad2436, 2024. a
Kiel, M., Eldering, A., Roten, D. D., Lin, J. C., Feng, S., Lei, R., Lauvaux, T., Oda, T., Roehl, C. M., Blavier, J.-F., and Iraci, L. T.: Urban-focused satellite CO2 observations from the Orbiting Carbon Observatory-3: A first look at the Los Angeles megacity, Remote Sens. Environ., 258, 112314, https://doi.org/10.1016/j.rse.2021.112314, 2021. a
Lauvaux, T., Giron, C., Mazzolini, M., d'Aspremont, A., Duren, R., Cusworth, D., Shindell, D., and Ciais, P.: Global assessment of oil and gas methane ultra-emitters, Science, 375, 557–561, https://doi.org/10.1126/science.abj4351, 2022. a
Lin, J. C., Gerbig, C., Wofsy, S. C., Andrews, A. E., Daube, B. C., Davis, K. J., and Grainger, C. A.: A near-field tool for simulating the upstream influence of atmospheric observations: The Stochastic Time-Inverted Lagrangian Transport (STILT) model, J. Geophys. Res.-Atmos., 108, 4493, https://doi.org/10.1029/2002JD003161, 2003. a, b, c
Nassar, R., Hill, T. G., McLinden, C. A., Wunch, D., Jones, D. B. A., and Crisp, D.: Quantifying CO2 Emissions From Individual Power Plants From Space, Geophys. Res. Lett., 44, 10045–10053, https://doi.org/10.1002/2017GL074702, 2017. a, b
Nassar, R., Mastrogiacomo, J.-P., Bateman-Hemphill, W., McCracken, C., MacDonald, C. G., Hill, T., O'Dell, C. W., Kiel, M., and Crisp, D.: Advances in quantifying power plant CO2 emissions with OCO-2, Remote Sens. Environ., 264, 112579, https://doi.org/10.1016/j.rse.2021.112579, 2021. a
Nayagam, L., Maksyutov, S., Oda, T., Janardanan, R., Trisolino, P., Zeng, J., Kaiser, J. W., and Matsunaga, T.: A top-down estimation of subnational CO2 budget using a global high-resolution inverse model with data from regional surface networks, Environ. Res. Lett., 19, 014031, https://doi.org/10.1088/1748-9326/ad0f74, 2023. a
O'Dell, C. W., Connor, B., Bösch, H., O'Brien, D., Frankenberg, C., Castano, R., Christi, M., Eldering, D., Fisher, B., Gunson, M., McDuffie, J., Miller, C. E., Natraj, V., Oyafuso, F., Polonsky, I., Smyth, M., Taylor, T., Toon, G. C., Wennberg, P. O., and Wunch, D.: The ACOS CO2 retrieval algorithm – Part 1: Description and validation against synthetic observations, Atmos. Meas. Tech., 5, 99–121, https://doi.org/10.5194/amt-5-99-2012, 2012. a
Rodgers, C. D.: Inverse methods for atmospheric sounding: theory and practice, vol. 2, World Scientific, ISBN 981-02-2740-X, 2000. a
Roten, D., Wu, D., Fasoli, B., Oda, T., and Lin, J. C.: An Interpolation Method to Reduce the Computational Time in the Stochastic Lagrangian Particle Dispersion Modeling of Spatially Dense XCO2 Retrievals, Earth Space Sci., 8, e2020EA001343, https://doi.org/10.1029/2020EA001343, 2021. a, b
Sherwin, E. D., Rutherford, J. S., Chen, Y., Aminfard, S., Kort, E. A., Jackson, R. B., and Brandt, A. R.: Single-blind validation of space-based point-source detection and quantification of onshore methane emissions, Sci. Rep.-UK, 13, 3836, https://doi.org/10.1038/s41598-023-30761-2, 2023. a
Shusterman, A. A., Teige, V. E., Turner, A. J., Newman, C., Kim, J., and Cohen, R. C.: The BErkeley Atmospheric CO2 Observation Network: initial evaluation, Atmos. Chem. Phys., 16, 13449–13463, https://doi.org/10.5194/acp-16-13449-2016, 2016. a
Shusterman, A. A., Kim, J., Lieschke, K. J., Newman, C., Wooldridge, P. J., and Cohen, R. C.: Observing local CO2 sources using low-cost, near-surface urban monitors, Atmos. Chem. Phys., 18, 13773–13785, https://doi.org/10.5194/acp-18-13773-2018, 2018. a
Steiner, M., Peters, W., Luijkx, I., Henne, S., Chen, H., Hammer, S., and Brunner, D.: European CH4 inversions with ICON-ART coupled to the CarbonTracker Data Assimilation Shell, Atmos. Chem. Phys., 24, 2759–2782, https://doi.org/10.5194/acp-24-2759-2024, 2024. a, b
Taylor, T. E., Eldering, A., Merrelli, A., Kiel, M., Somkuti, P., Cheng, C., Rosenberg, R., Fisher, B., Crisp, D., Basilio, R., Bennett, M., Cervantes, D., Chang, A., Dang, L., Frankenberg, C., Haemmerle, V. R., Keller, G. R., Kurosu, T., Laughner, J. L., Lee, R., Marchetti, Y., Nelson, R. R., O'Dell, C. W., Osterman, G., Pavlick, R., Roehl, C., Schneider, R., Spiers, G., To, C., Wells, C., Wennberg, P. O., Yelamanchili, A., and Yu, S.: OCO-3 early mission operations and initial (vEarly) XCO2 and SIF retrievals, Remote Sens. Environ., 251, 112032, https://doi.org/10.1016/j.rse.2020.112032, 2020. a
Turner, A. J., Shusterman, A. A., McDonald, B. C., Teige, V., Harley, R. A., and Cohen, R. C.: Network design for quantifying urban CO2 emissions: assessing trade-offs between precision and network density, Atmos. Chem. Phys., 16, 13465–13475, https://doi.org/10.5194/acp-16-13465-2016, 2016. a
Turner, A. J., Jacob, D. J., Benmergui, J., Brandman, J., White, L., and Randles, C. A.: Assessing the capability of different satellite observing configurations to resolve the distribution of methane emissions at kilometer scales, Atmos. Chem. Phys., 18, 8265–8278, https://doi.org/10.5194/acp-18-8265-2018, 2018. a
Turner, A. J., Kim, J., Fitzmaurice, H., Newman, C., Worthington, K., Chan, K., Wooldridge, P. J., Köehler, P., Frankenberg, C., and Cohen, R. C.: Observed Impacts of COVID-19 on Urban CO Emissions, Geophys. Res. Lett., 47, e2020GL090037, https://doi.org/10.1029/2020GL090037, 2020. a, b, c, d, e, f, g, h, i, j, k, l, m, n, o, p, q
Varon, D. J., Jacob, D. J., Sulprizio, M., Estrada, L. A., Downs, W. B., Shen, L., Hancock, S. E., Nesser, H., Qu, Z., Penn, E., Chen, Z., Lu, X., Lorente, A., Tewari, A., and Randles, C. A.: Integrated Methane Inversion (IMI 1.0): a user-friendly, cloud-based facility for inferring high-resolution methane emissions from TROPOMI satellite observations, Geosci. Model Dev., 15, 5787–5805, https://doi.org/10.5194/gmd-15-5787-2022, 2022. a
Varon, D. J., Jacob, D. J., Hmiel, B., Gautam, R., Lyon, D. R., Omara, M., Sulprizio, M., Shen, L., Pendergrass, D., Nesser, H., Qu, Z., Barkley, Z. R., Miles, N. L., Richardson, S. J., Davis, K. J., Pandey, S., Lu, X., Lorente, A., Borsdorff, T., Maasakkers, J. D., and Aben, I.: Continuous weekly monitoring of methane emissions from the Permian Basin by inversion of TROPOMI satellite observations, Atmos. Chem. Phys., 23, 7503–7520, https://doi.org/10.5194/acp-23-7503-2023, 2023. a
Veefkind, J., Aben, I., McMullan, K., Förster, H., de Vries, J., Otter, G., Claas, J., Eskes, H., de Haan, J., Kleipool, Q., van Weele, M., Hasekamp, O., Hoogeveen, R., Landgraf, J., Snel, R., Tol, P., Ingmann, P., Voors, R., Kruizinga, B., Vink, R., Visser, H., and Levelt, P.: 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 Sens. Environ., 120, 70–83, https://doi.org/10.1016/j.rse.2011.09.027, 2012. a
Wang, Y., Broquet, G., Bréon, F.-M., Lespinas, F., Buchwitz, M., Reuter, M., Meijer, Y., Loescher, A., Janssens-Maenhout, G., Zheng, B., and Ciais, P.: PMIF v1.0: assessing the potential of satellite observations to constrain CO2 emissions from large cities and point sources over the globe using synthetic data, Geosci. Model Dev., 13, 5813–5831, https://doi.org/10.5194/gmd-13-5813-2020, 2020. a
Wu, D., Lin, J. C., Oda, T., and Kort, E. A.: Space-based quantification of per capita CO2 emissions from cities, Environ. Res. Lett., 15, 035004, https://doi.org/10.1088/1748-9326/ab68eb, 2020. a
Zavala-Araiza, D., Lyon, D. R., Alvarez, R. A., Davis, K. J., Harriss, R., Herndon, S. C., Karion, A., Kort, E. A., Lamb, B. K., Lan, X., Marchese, A. J., Pacala, S. W., Robinson, A. L., Shepson, P. B., Sweeney, C., Talbot, R., Townsend-Small, A., Yacovitch, T. I., Zimmerle, D. J., and Hamburg, S. P.: Reconciling divergent estimates of oil and gas methane emissions, P. Natl. Acad. Sci. USA, 112, 15597–15602, https://doi.org/10.1073/pnas.1522126112, 2015. a, b
Short summary
We developed an efficient GHG (greenhouse gas) flux inversion framework using a machine-learning emulator (FootNet) as a surrogate for an atmospheric transport model, resulting in a 650 × speedup. Paradoxically, the flux inversion using the ML (machine-learning) model outperforms the full-physics model in our case study. We attribute this to the ML model mitigating transport errors in the GHG flux inversion.
We developed an efficient GHG (greenhouse gas) flux inversion framework using a machine-learning...
Altmetrics
Final-revised paper
Preprint