Tracking regional ch4 emissions through collocated air pollution measurement: a pilot application and robustness analysis in china

ABSTRACT Accurate regional methane (CH4) emission monitoring is essential for effective climate mitigation but remains constrained by limited observational networks and challenged by diffuse

emission sources. Here, we present an innovative regional CH4 inversion system integrating satellite-based carbon monoxide (CO) observations with ground-based CH4-to-CO flux ratios. Our

study estimates China’s CH4 fluxes between 2000 and 2021, revealing an average of 48.4 ± 13.8 Tg yr−1 and a significant increasing trend of 1.1 ± 0.2 Tg yr−2. Over the 22-year period,

socio-economic development drove a 92.1 Tg increase in China’s CH4 fluxes, partially offset by a 78.1 Tg reduction due to declining emission intensity; however, this mitigating effect

weakened after 2015. Our results demonstrate comparability with independent estimates, and comprehensive sensitivity and uncertainty analyses confirm the robustness of our approach. This

study highlights the potential of integrating air pollution monitoring into tracking regional greenhouse gas emissions. SIMILAR CONTENT BEING VIEWED BY OTHERS ATMOSPHERIC OBSERVATIONS

SUGGEST METHANE EMISSIONS IN NORTH-EASTERN CHINA GROWING WITH NATURAL GAS USE Article Open access 17 November 2022 URBAN METHANE EMISSION MONITORING ACROSS NORTH AMERICA USING TROPOMI DATA:

AN ANALYTICAL INVERSION APPROACH Article Open access 19 April 2024 IMPROVED ESTIMATION OF CARBON DIOXIDE AND METHANE USING MACHINE LEARNING WITH SATELLITE OBSERVATIONS OVER THE ARABIAN

PENINSULA Article Open access 04 January 2025 INTRODUCTION Methane (CH4), the second most abundant anthropogenic greenhouse gas, has contributed to over 30% of global warming since the

Industrial Revolution1,2. Due to its relatively short atmospheric lifetime and high warming potential, mitigating CH4 emissions is crucial for curbing global near-term temperature rise.

Cutting CH4 emissions is now a central focus in climate change mitigation strategies. This priority has been recognized in the Paris Climate Agreement, adopted at COP21 under the United

Nations Framework Convention on Climate Change (UNFCCC), and reaffirmed in subsequent conferences3,4. The Global Methane Pledge, launched at COP26, aims to collectively reduce anthropogenic

CH4 emissions by 30% from the 2020 levels by 2030 (https://www.globalmethanepledge.org/). COP28 further facilitated CH4 abatement agreements across climate, economic, and industrial

sectors5. Efficient reduction of CH4 emissions is contingent upon accurate and timely accounting. The scientific community has developed atmospheric observation-based inversion systems,

known as the top-down (TD) approach, to track CH4 fluxes across global, regional, and infrastructure scales6,7,8,9. Principles of global mass conservation and observation station data can

constrain the global CH4 budget. However, monitoring CH4 emissions at regional scales poses challenges due to sparse observational networks, complicating the establishment of correlation

between measured CH4 spatial gradients with regional CH4 sources and sinks. The increasing availability of satellite observations with dense spatial coverage is advancing the development of

region- and infrastructure-level inversion systems. Nonetheless, these systems face several obstacles, including satellite retrieval biases, the effects of clouds and aerosols, high

background CH4 levels, and errors in transport models10,11,12,13,14. As a result, regional CH4 budget monitoring remains in its early stages, typically relying on global-scale atmospheric

inversion systems to estimate regional budgets, despite the associated uncertainties. We even observe a considerable spread in model estimates for the same region across different

systems10,13,15,16. We face substantial challenges in monitoring CH4 fluxes in large countries with sparse observation systems and widely distributed sources. For instance, TD estimates of

CH4 emissions for China, the world’s largest anthropogenic emitter, have shown disparities of up to 60% after 2000, with disagreements on emission magnitudes, trends, and

variations12,17,18,19,20,21,22. These TD inversions also differ substantially from bottom-up (BU) estimates, highlighting the considerable uncertainty in understanding China’s CH4 emissions.

Additionally, both TD and BU approaches are hindered by issues of timeliness and latency in emissions reporting. Given the limitations in current methods, developing new approaches for

accurate, transparent, and timely CH4 emissions monitoring is crucial. Such advancements would improve current estimate ensembles and constrain the uncertainty range, which is urgently

needed to support effective emissions reduction efforts for major emitters. Here, we developed a regional CH4 flux inversion approach based on satellite-based carbon monoxide (CO) inversion

and observation-based CH4-to-CO flux ratios (Fig. 1), fundamentally differing from previous TD CH4 inversion methods. CO, due to its short atmospheric lifetime, exhibits sharp spatial

gradients of concentrations around emission sources, making it detectable from space and suitable for flux inversion. Our previous research inferred global gridded, weekly CO fluxes from

satellite CO column retrievals over the past two decades23. Building on this, we derived daily CH4 to CO flux ratios from ground-based measurements downwind of China. The hybrid

single-particle Lagrangian integrated trajectory (HYSPLIT) backward model was used to ensure that the CH4 to CO concentration ratios reflect China’s emissions, with conversion factors (CFs)

accounting for atmospheric transport losses of both gases. By combining satellite-constrained CO fluxes with observation-based CH4 to CO flux ratios, we estimate China’s CH4 fluxes. We

further performed a perturbation sensitivity analysis to ensure the robustness of our inversion estimation results. Our approach offers a new perspective on TD regional CH4 flux estimates

based on air pollution inversion, aiming to leverage the widespread air quality monitoring networks in climate actions. RESULTS INCREASING CH4 TO CO RATIOS IN AIR MASSES ORIGINATING FROM

CHINA A crucial step in inferring CH4 fluxes from CO fluxes is estimating the CH4 to CO flux ratios that represent China’s emission budget. We utilized observational data from the

Yonagunijima (YON) station, located downwind of China, to measure the air masses originating from the country. To determine regional background levels, we used observational data from the

Minamitorishima (MNM) station, situated in the remote clean Pacific at the same latitude band (Fig. 2a). We calculated the enhancement concentrations (EnhCs) caused by China’s emissions by

subtracting the daily averages measured at the MNM station from the hourly CH4 and CO concentrations recorded at the YON station on the corresponding day (Fig. 2b). Air parcel transport

pathway analysis indicates that the YON station effectively captures polluted plumes from China (Fig. 2a). The average travel time for air masses from China to the YON station is 96 h, which

is much shorter than the atmospheric lifetimes of CH4 (~10 years) and CO (~2 months) over the same latitude band1, 23. For a year, the trajectories cover nearly the entire China region,

with the 0.5° × 0.5° grid cells intersected by at least 10 backward trajectories per year accounting for more than 80% of China’s annual anthropogenic CH4 emissions (Fig. 2a). We relied on

the observations at the YON station to a reflect the joint variation of China’s CH4 and CO fluxes, and the single station observation-induced uncertainties are discussed and quantified

further in details. The ratios of CH4 to CO EnhCs have increased from 0.3 ppb-CH4 ppb-CO−1 in 2000 to 0.7 ppb-CH4 ppb-CO−1 in 2021, with a significant annual growth rate of 2.1% ± 0.5% yr−1

(non-parametric Mann-Kendall test, 95% confidence interval, Fig. 2c). This rising trend is primarily driven by rapidly increasing CH4 concentrations and declining CO concentrations, after

correcting for regional background levels (SI Appendix, Fig. S1). The steady decline in anthropogenic CO emissions from China is attributed to the improvements in fuel combustion efficiency

in industrial boilers, gasoline vehicles, and residential stoves23,24, with the decline accelerating after 2013, following the implementation of China’s Clean Air Actions25. In contrast,

China’s anthropogenic CH4 emissions have tended to increase since 2000 mainly driven by fossil fuel production and waste treatment, according to various TD and BU estimates despite different

magnitudes and trends26,27. The combination of declining CO with increasing CH4 emissions led to the rising enhancement ratios (EnhRs) of CH4 to CO from 2000 to 2021 (Fig. 2c). The CH4 to

CO flux ratios derived from the measured EnhRs, accounting for chemical loss during transport (see Methods), are estimated to average 0.22 g-CH4 g-CO−1 for the period 2000–2019, with a

significant increasing trend of 1.0% ± 0.3% yr−1. These measurement-based estimates closely align with the emission ratio of 0.22 g-CH4 g-CO−1 and the growing trend of 0.9% ± 0.1% yr−1 for

China as a whole, based on our CO inversion emissions and anthropogenic CH4 emissions derived from the community emissions data system (CEDS) dataset28. The CH4 to CO emission ratios

estimated for the regions traversed by different numbers of back-trajectories over China span from 0.19 to 0.24 g-CH4 g-CO−1 (SI Appendix, Fig. S2) for the annual average between 2000 and

2019, covering our measurement-based estimate of 0.22 g-CH4 g-CO−1. The broad consistency between these independent estimates indicates that the YON measurement station represents China’s

CH4 to CO flux ratios on a national, annual scale. RAPID INCREASES IN CHINA’S CH4 FLUXES INFERRED FROM CO-BASED INVERSION China’s CH4 fluxes have increased from 41.6 ± 11.9 Tg yr−1 in 2000

(with an uncertainty range at a 95% confidence interval approximated by sensitivity analysis) to 60.8 ± 18.9 Tg yr−1 in 2021 (Fig. 3 and SI Appendix, Table S1), based on our CO inversion

fluxes and the observed CH4 to CO flux ratios. This represents a significant upward trend of 1.1 ± 0.2 Tg yr−2 from 2000 to 2021. Our CH4 inversion results fall within the broad range of the

latest Global Carbon Project (GCP) surface-based (purple curve in Fig. 3a) and satellite-based (green curve in Fig. 3a) inversions for China29. Despite broad consistency in the decadal

averages, discrepancies in the trend estimates remain even considering their respective uncertainty ranges (Fig. 3b). For the period 2000–2020, our inversion reveals an increasing trend of

1.1 ± 0.2 Tg yr−2 (red solid diamond in Fig. 3b), which is faster than the GCP surface-based inversion estimate of 0.5 ± 0.1 Tg yr−2 (purple solid diamond in Fig. 3b). Both our study and the

GCP surface-based inversion models rely on ground measurement stations with incomplete spatial coverage over China, which may affect the accuracy of inversion results for this region. The

substantial differences between our study and the GCP surface-based inversion highlight the challenges of constraining regional CH4 fluxes based on limited surface networks, with the derived

inversion results probably dependent on the approach used. For the period between 2000 and 2010, our inversion estimates a trend comparable to two previous inversion studies constrained by

satellite retrievals and ground measurements, respectively18,20. For 2010–2020, the increasing trends in China’s CH4 fluxes in our study (1.1 ± 0.8 Tg yr−2, red hollow square in Fig. 3b) is

close to the GCP satellite-based inversion (0.7 ± 0.4 Tg yr−2, green hollow square in Fig. 3b), both of which are faster than the GCP surface-based inversion (0.2 ± 0.3 Tg yr−2, purple

hollow square in Fig. 3b). Overall, both our CO-based inversion and the GCP inversions agree on the steady increase in CH4 fluxes from China since 2000, although the growth rates diverge. We

compared our inversion results with anthropogenic emission inventories (Fig. 3c, d) to investigate the drivers behind China’s increasing CH4 fluxes. The decadal average fluxes from our

inversion are close to the emissions data reported by CEDS v2021_04_2128 and Peking University (PKU-CH4 v2) datasets (https://figshare.com/s/b38a368111749f1412be)26,30. However, our CH4 flux

estimates are nearly 15% higher than the latest CEDS v2024_07_0831 and 20% lower than those from the Emissions Database for Global Atmospheric Research version 8 (EDGARv8.0)32 and the Food

and Agriculture Organization of the United Nations (FAO) (https://www.fao.org/faostat/en/#data/GT)33. It should be noted that previous atmospheric inversion studies using EDGAR emissions

data as prior also reported an overestimation of CH4 emissions by EDGAR in China29. The increasing trends from 2000 to 2020 observed in our inversion estimates (red solid diamond in Fig. 3d)

agree with those from BU inventories (other solid diamonds in Fig. 3d), probably indicating that the rise in CH4 fluxes from China is primarily driven by anthropogenic sources. However, the

BU inventories suggest slightly lower rates of increase compared to our inversions during 2010–2020. The CEDS (including both of the two versions), EDGARv8.0, and FAO data show modestly

increasing trends since 2010, while the PKU-CH4 v2 even indicates a slight decline after 2010 (Fig. 3d). The comparison between our inversions with emission inventories here is relatively

rough, because inversion estimates give total net fluxes including anthropogenic and natural sources as well as sinks, whereas emission inventories involved here only account for

anthropogenic sources. Resolving the divergent trends across estimates needs to reconcile the different system boundaries for emission fluxes accounting, which is rather challenging here

with the limited data available we have in this study. The insufficient available observations in the YON station, particularly between June and October, represent a key limitation and

uncertainty source in our study, leading to lower reliability for the summer and autumn compared to the other seasons34,35. During these months, air masses predominantly originated from the

coastal areas, suggesting our estimates are more reflective of emissions from these regions rather than nationwide sources. Since coastal emissions are significantly higher than the national

average32, the growth of EnhRs in summer and autumn after 2015 may be overestimated, potentially explaining why the post-2010 increase rates in our study are higher than those reported

elsewhere (Fig. 3b, d, and SI Appendix, Figs. S3 and S4). The focus of our study is to propose a methodological framework that, with improved data availability, would provide more accurate

and representative results. DRIVERS BEHIND THE INCREASE IN CHINA’S ANTHROPOGENIC CH4 EMISSIONS The logarithmic mean divisia index (LMDI) decomposition analysis, an improvement of index

decomposition analysis, is widely used to quantify the driving forces behind emission changes36,37. We used LMDI38 to assess the contributions of factors, including CH4 fluxes per unit of

gross domestic product (GDP), GDP per capita, and population level, to changes in China’s CH4 fluxes over time (See Method). Our results reveal that socioeconomic development in China

dominates anthropogenic CH4 emission trends, as demonstrated by the increased GDP per capita that spurs CH4 fluxes growth (Fig. 4a). Conversely, the reduction in CH4 fluxes per unit of GDP

has shown a mitigating effect on these emissions. Our decomposition analysis indicates that China’s rising GDP per capita consistently drove up CH4 emissions by nearly 92.1 Tg from 2000 to

2021. The nation’s rapid economic growth and social development, both of which depend on intensive anthropogenic activities, such as fossil fuel production and consumption, livestock

farming, and waste treatment, have been key contributors to this increase (Fig. 4b). Fortunately, CH4 fluxes per unit of GDP have decreased due to a series of mitigation measures implemented

across multiple sectors, such as the advancements in irrigation techniques for rice cultivation within the agriculture sector39,40,41 and the increased recovery and utilization of mine gas

in coal mining42,43. These measures have offset 78.1 Tg of emissions growth over the past two decades. However, the impact of these reductions decreased from 28.6 Tg between 2000 and 2005 to

8.5 Tg between 2015 and 2021, which shifted from fully counteracting the emission increases driven by GDP per capita and population growth to offset less than half of that growth. This

trend underscores the growing challenge and urgency of achieving deeper emission reductions. It is thus critical to identify key emission sources and regions to design targeted mitigation

strategies that decouple emissions from socio-economic development. In line with the consistent trends between our inversion and other estimates, the sectoral and regional details of other

data allow us to identify the drivers of CH4 emissions growth in China (Fig. 4c, d, and SI Appendix, Fig. S5 for sectors; Fig. 4e, f for regions). Based on EDGAR v8.0 and GCP surface-based

inversion, fossil fuel production, particularly coal, has been consistently identified as the dominant force behind CH4 emissions growth over the past two decades. While the fossil fuel

sector accounted for ~30% of total emissions on average, it contributed nearly 70% of the increase, with coal representing up to 85% of this sector’s growth (Fig. 4c, d). Rising energy

demand drove a tripling of coal production in China since 2000, supplying nearly 70% of the country’s total energy consumption until 2013, when the “coal-to-gas’’ energy transition policy

was implemented (https://www.gov.cn/zwgk/2013-09/12/content_2486773.htm). With a series of “low-carbon’’ policies introduced after 2010, coal’s share in energy consumption declined to below

57% after 2020, leading to a marked slowdown in coal-related emissions44. Meanwhile, emissions from natural gas have continued to rise due to increased production, maintaining fossil fuel’s

contribution to emission growth, albeit at a slower rate (Fig. 4b). Moreover, the agriculture and waste sectors are the second-largest contributors to CH4 emission growth, together

accounting for over half of total emissions (Fig. 4c, d). Notably, emissions from waste management alone contributed nearly 30% of the total emission increase, primarily driven by solid

waste disposal and wastewater treatment according to the EDGARv8.0 emission inventory (SI Appendix, Fig. S6). These characteristics align with the near doubling of solid waste disposal

(e.g., municipal solid waste) and wastewater discharge in China over the past two decades (Fig. 4b). The vital role of the agriculture and waste sectors in emission growth is also supported

by other inversions, despite variations in their findings on fossil fuel emissions9,30,45. Spatial emission distributions show greater discrepancies across different datasets compared to the

sectoral distributions discussed above, except for East China, whose role is consistently emphasized. East China accounts for nearly 24% of total emissions in both GCP inversions and the

EDGAR inventory, contributing 27% and 23% of the emission increase from 2000 to 2021, respectively (Fig. 4e, f, with the provinces in each region listed in SI Appendix, Table S2). The waste

sector drove its rapid growth, as indicated by the EDGARv8.0 (SI Appendix, Fig. S6) and other independent estimates26,27,46, with GCP surface-based inversions attributing 36% of China’s

waste and agricultural sectors emission growth to this region (SI Appendix, Fig. S7). Besides, EDGARv8.0 identifies North and Northwest China as major contributors to CH4 emissions growth

(35% and 27% of growth, respectively), driven by coal exploitation in provinces like Shanxi, Inner Mongolia, Shaanxi, and Henan (Fig. 4f, and SI Appendix, Fig. S6)27,43. In contrast, GCP

inversions highlighted substantial contributions from South (21%) and Southwest (18%) China, with the Southwest leading fossil fuel emission growth (SI Appendix, Fig. S7). These apparent

spatial emission discrepancies reveal ongoing challenges in accurately quantifying fine-scale CH4 budget with the current method and data. ROBUSTNESS ANALYSIS OF INVERSION ESTIMATES Through

a series of perturbations on key parameters and model configurations (SI Appendix, Table S3), we demonstrated the robustness of our inversion estimate of CH4 flux magnitude (Fig. 5). The

22-year average fluxes varied between −2% and 10%, corresponding to an uncertainty range of 47.5–52.8 Tg yr−1. Among the factors tested, those related to the EnhRs had a more pronounced

impact on inversion fluxes than the others like CF. In the EnhR sensitivity tests, the height setting of the HYSPLIT backward model emerged as the most influential parameter. The atmospheric

conditions at different heights influence observation data selection and EnhR estimation. We tested different heights of 100, 500, 1000, 1500, and 2000 m and observed variations in the

22-year average CH4 flux estimates of up to 5.4 Tg (equivalent to 10–11% of average estimates). Here, we initiated the HYSPLIT model at a height of 500 m, a common practice in previous

studies to reduce topographical influences47,48 and capture CH4 transport. The sensitivity tests on most other factors showed less than 1% differences from the reference inversion results,

which fall within the lower ranges of both satellite-based and surface-based inversion ensembles during corresponding periods. All sensitivity tests confirmed the increasing trend from 2000

to 2020, with trend estimates ranging from 0.6 to 1.3 Tg yr−2. Our characteristic of faster flux growth compared to the GCP inversion results remained consistent, with the height setting in

the HYSPLIT model exerting the greatest influence on the trend estimate, similar to the impact on the magnitude estimate. Additionally, our flux trend is highly sensitive to the

meteorological field used to drive the HYSPLIT model, with a trend discrepancy of 27% between the two meteorological products: the NCEP/NCAR global reanalysis data and the global data

assimilation system (GDAS) dataset (https://www.ready.noaa.gov/data/archives/). We opted to use the NCEP/NCAR data in the HYSPLIT model due to its longer time coverage. Overall, our study

underscores the critical importance of key parameters in the backward trajectory reconstruction for accurately estimating CH4 flux magnitudes and trends in our inversion system, pointing to

a need for further research to optimize these parameters for enhancing reliability and reducing uncertainties in regional flux estimation. The overall uncertainty of China’s annual CH4

fluxes from our inversion is ~±29% within a 95% confidence interval, as determined through error propagation (see Method). This accounts for the respective uncertainties of the CO inversion

results, EnhRs, and CFs. The primary sources of uncertainty are the CO fluxes and EnhRs, each with an uncertainty of ~±20%, with the former having been discussed in our previous study23. For

EnhRs, the main source of error arises from the spatial representativeness of single-station observations, with an uncertainty of about ±15%. This is influenced by the geographic location

of the YON station and the spatial coverage of the associated backward trajectory. In China, the prevailing onshore winds in winter make the YON station observation have better spatial and

temporal coverage, whereas the prevailing monsoons in summer result in fewer available observations. While these errors may be unavoidable, we have implemented measures to minimize their

impacts on our results. By correcting outliers (values more than 2 standard deviations from the mean) in seasonal values and applying seasonal and three-year moving averages to smooth the

interannual variation of EnhRs, we mitigate apparent seasonal anomalies in individual years and achieve a modest interannual fluctuation of less than 14% (SI Appendix, Figs. S3 and S4). The

constrained annual average CH4 to CO flux ratio (0.22 g-CH4 g-CO−1) is comparable to the independent estimate derived from the CEDS CH4 emissions dataset (0.22 g-CH4 g-CO−1) (SI Appendix,

Fig. S2), confirming that such an estimate is reasonable. DISCUSSION Our study developed an innovative method for estimating regional CH4 fluxes by combining measurement-based CH4 to CO flux

ratios with satellite-derived CO inversion fluxes. Applying this approach, we monitored China’s CH4 fluxes from 2000 to 2021, revealing a continuous increase consistent with most estimates.

By integrating socioeconomic factors and detailed sectoral emission data, our study highlights the positive correlation between GDP per capita growth and CH4 emissions and underscores the

dominant emission sectors during different periods. Our methodology overcomes several challenges faced by traditional CH4 inversion methods, including limited observational coverage, high

background concentrations, and computationally intensive inversion algorithms. However, this preliminary method still has limitations that warrant future improvements. First, our study

simplifies the calculation of EnhRs by relying on a ground-based station to convert CO fluxes to CH4 fluxes. Given the complexity and dynamic variability of atmospheric transport processes,

influenced by multiple meteorological factors, this simplification may introduce uncertainties, particularly in years with large seasonal disparities in the number of available observations.

Second, although the robustness of our inversion results based on single-station observations has been confirmed, the limited spatio-temporal coverage of the backward trajectories from the

single station may not be overlooked. The reliance on single-station observations is primarily due to the scarcity of stations capable of providing long-term CH4 and CO measurements

simultaneously across China. Third, there is inherent uncertainty in the HYSPLIT model and limitations in its parameter settings. Specifically, some factors, such as the height, can only be

roughly estimated, posing challenges in accurately representing the transport conditions for each air mass. Finally, since CH4 emission inventories tend to cover only anthropogenic sources,

there are biases in our driver analyses of CH4 fluxes. Nevertheless, according to the inversion results, the contributions of natural sources have minimal effect on overall emission

variations in China. Therefore, further refinements and enhanced observational networks could substantially improve the accuracy and efficiency of our inversion preliminary framework.

Involving additional monitoring stations and utilizing data from multiple locations can help validate and further refine the current method we developed. Additionally, its universal

applicability allows it to be adapted to diverse regions with varying geographical and climatic conditions. By integrating with existing and expanding air pollution surface observation

networks, this method has the potential for widespread application, enabling more efficient, dynamic, and accessible assessments of the CH4 budget on a regional scale. Furthermore, this

approach offers a promising tool for synergistic monitoring of greenhouse gas and air pollutant emissions. The successful use of CO as a tracer for estimating CH4 fluxes underscores the

potential to monitor other long-lived greenhouse gases, such as nitrous oxide, sulfur hexafluoride, and hydrofluorocarbons. By leveraging easily traceable, short-lived air pollutants, our

method provides a valuable framework for comprehensive regional assessments of greenhouse gas budgets. On one hand, this system can effectively monitor long-term target gases, offer

independent and robust data support, and enhance monitoring and verification support capabilities at regional scales. On the other hand, it can lay the groundwork for establishing optimal

observation stations tailored to specific regional needs. Furthermore, while achieving synergistic atmospheric inversion, this system could offer a data foundation responding to both climate

change and air pollution mitigation targets. It can play a pivotal role in swiftly evaluating the effectiveness of emission reduction policies, thereby empowering policymakers with

actionable insights to mitigate adverse impacts both effectively and efficiently. MATERIALS AND METHODS Our method combines satellite-based CO inversion fluxes with ground measurement-based

CH4 to CO flux ratios to infer the CH4 fluxes from China (Fig. 1). The key concept relies on the well-mixed CH4 and CO during atmospheric transport. Their concentrations, simultaneously

observed by a ground station downwind of China, can reflect the relative variations of CH4 and CO fluxes originating from China along backward trajectories. China’s CH4 fluxes are estimated

based on the following two equations. $${F}_{C{H}_{4,i}}\,={F}_{CO,i}\times F{R}_{i}$$ (1) $$F{R}_{i}=Enh{R}_{i}\times C{F}_{i}\times {{\rm{M}}}_{{{\rm{CH}}}_{4}}{/{\rm{M}}}_{{\rm{CO}}}$$

(2) where _i_ represents the year between 2000 and 2021. _F__CH4,i_ denotes the annual CH4 fluxes from China. \({F}_{CO,i}\,\) is the annual CO fluxes of China, derived from the satellite CO

column-based atmospheric inversion23,49. _FR_ is the flux ratio of CH4 to CO, calculated based on the _EnhR_ of the two gases, the _CF_, and their molar mass ratio

(\({{\rm{M}}}_{{{\rm{CH}}}_{4}}{/{\rm{M}}}_{{\rm{CO}}}\)). _EnhR_ represents the ratio of the enhanced concentrations of CH4 to CO above the regional background levels, quantified based on

the linear correlation between their enhanced concentrations. _CF_ accounts for the mass loss due to the chemical removal of both gases during their long-range atmospheric transport from

source regions to the observation station, and it is combined with \({{\rm{M}}}_{{{\rm{CH}}}_{4}}{/{\rm{M}}}_{{\rm{CO}}}\) to convert _EnhR_ to _FR_50. SATELLITE-BASED CO INVERSION FLUXES

(_F_ _CO_) _F__CO_ was estimated using a multi-species atmospheric inversion system in our previous study49. We briefly describe the inversion system framework here and refer the readers to

our previous papers23,49,51 for more details. The global gridded CO budget from 2000 to 2021 was reconstructed with the PYVAR inversion system, which was developed and maintained by the

Laboratoire des Sciences du Climat et de l’Environnement in France over the past two decades. This inversion system was built based on the global 3-D transport model of the Laboratoire de

Météorologie Dynamique, which was integrated with a simplified chemistry module simplified atmospheric chemistry assimilation System52. We assimilated the latest retrievals of CO columns

(version 9) from the measurements of pollution in the troposphere satellite and incorporated prior information on the initial model state, surface source emissions, CO chemical production,

and OH field to constrain and optimize different components of CO sources and sinks. This inversion system was successfully utilized to investigate global long-term CO budget and wildfire CO

and CO2 emission patterns in our previous studies23,49. The model evaluation results against independent CO measurement and different bottom-up CO emission datasets confirm the system’s

robustness in CO flux estimation. Our CO inversion provides global gridded CO fluxes at a spatial resolution of 3.75° × 1.9°, which includes six CO sources: anthropogenic, oceanic, biomass

burning, biogenic, CH4 oxidation, and non-methane volatile organic compounds oxidation, as well as CO sink by OH reaction. The surface CO emissions of different sources are quantified by

multiplying the optimized 8-daily surface CO total fluxes by the proportion of each CO source within each model grid cell, as provided by prior emissions53,54,55. Here, we aggregated all CO

sources, excluding oceanic contribution, to represent the CO fluxes on land and then extracted the fluxes over China’s mainland for the estimation of CH4 fluxes for China. ENHANCEMENT RATIO

OF CH4 TO CO CONCENTRATIONS (_ENHR_) Selecting appropriate observation stations is crucial for estimating _EnhR_, which in turn influences the estimation of _FR_. We selected the YON ground

station (YON, 24.47°N, 123.01°E), located at the western end of the Ryukyu Archipelago at an altitude of 30 m, to monitor hourly concentrations of CH4 and CO. YON is situated downwind of

China and can capture the air masses originating from China facilitated by the East Asian monsoon. Moreover, the MNM station (MNM, 24.29°N, 153.98°E), located in the remote and clean North

Pacific Ocean at an altitude of 7 m, was used as the regional background reference. To extract hourly flux signals from China and ensure the stability of background data, we utilized the

hourly concentrations observed by YON and the corresponding daily average from MNM to calculate the EnhC (Eq. 3), which represents CH4 and CO concentrations exceeding the regional background

levels. $$EnhC={C}_{h,YON}-{C}_{d,MNM}$$ (3) Where \({C}_{h,YON}\) represents hourly concentrations of CH4 or CO measured at YON and \({C}_{d,MNM}\) represents daily average CH4 or CO

concentrations based on hourly observational data at MNM for the corresponding day. The hourly observation data of CH4 and CO both from YON and MNM stations were derived from the World Data

Center for Greenhouse Gases (WDCGG)4,56,57,58,59. We employed the HYSPLIT model to reconstruct backward trajectories of air masses observed at YON, to select EnhC that represents air masses

from China’s mainland. The HYSPLIT model, developed by NOAA’s air resources laboratory60, is capable of simulating the trajectory of substances in the atmosphere with hourly resolution

across local to global scales. We configured the HYSPLIT model to track backward trajectories at a height of 500 meters above sea level over a 7-day (168-h) period. The meteorological data

used in our reference inversion was the NCEP/NCAR global reanalysis data, with a spatial resolution of 2.5° × 2.5°, which provides data from 1948 to the present. Based on the spatial

analysis of simulated backward trajectories of air masses, we selected those crossing over mainland China before being captured by YON and the corresponding hourly EnhCs for CH4 flux

inversion in our study. Few air masses that transport over a long distance originate from other countries (e.g., Mongolia, Korea) before entering China. Since China dominates CO and CH4

emissions in East Asia, we retained those data in our study to represent China’s fluxes, increasing the amount of available observation data. To enhance the robustness of our analysis, we

grouped the hourly EnhCs by season and performed linear regression analysis to estimate seasonal EnhRs derived from the slope of CH4 EnhCs relative to CO EnhCs. The average Pearson

correlation coefficients of all seasons exceeded 0.75, confirming the well-mixed nature of CH4 and CO emissions before measurement (SI Appendix, Fig. S8). Outliers in the seasonal EnhR,

defined as values exceeding two standard deviations from the seasonal mean between 2000 and 2021—primarily due to inadequate observation amount—were removed and replaced with the average

values for the same season from adjacent years. The annual EnhRs from 2000 to 2021 were estimated based on the average of seasonal EnhRs for each year. We then applied a 3-year moving

average method to reduce unexpected inter-annual fluctuations in EnhRs and to focus on trends. ESTIMATION OF _CF_ We introduced _CF_ to convert _EnhR_ to _FR_, accounting for the annual mean

travel time of selected air masses and the atmospheric lifetimes of CO and CH4 (Eq. 4)50. We simulated the annual average travel time from each grid cell in China to the YON station and

calculated the regional average travel time weighted by the annual average emission proportions of each grid cell based on the EDGARv8.0 dataset. Besides, we incorporated CO and CH4

lifetimes with annual variability under the increasing tendency of atmospheric hydroxyl radical (OH) levels, the primary sink for the two gases (Eq. 5). $$C{F}_{i}=\frac{\exp (t/{\tau

}_{C{H}_{4},i}\,\,)}{\exp (t/{\tau }_{CO,i})}$$ (4) $${\tau }_{gas,i}=\frac{1}{{[OH]}_{i}\cdot {k}_{gas}}$$ (5) where _t_ is the annual mean travel time of all air masses transported from

mainland China to the YON station. \({\tau }_{C{H}_{4},i}\) and \({\tau }_{CO,i}\) are the lifetimes of CH4 and CO, respectively, calculated based on the annual average atmospheric OH

concentrations ([_OH_], molecules cm−3) and the corresponding rate constants (\({k}_{gas}\)). For this study, [_OH_] was assumed to increase linearly by 1% per year since 2000, with a

baseline concentration of 1 × 106 molecules cm−3 in 200561. The estimation of \({k}_{gas}\) is based on the formulation in the Chemical Kinetics and Photochemical Data for Use in Atmospheric

Studies provided by NASA/JPL Data Evaluation (https://jpldataeval.jpl.nasa.gov/contact.html) and Zhao et al.62. The atmospheric temperature over China was obtained from the World Bank Group

(https://climateknowledgeportal.worldbank.org/country/china/climate-data-historical). SOCIO-ECONOMIC DRIVER ANALYSIS OF CHINA’S CH4 FLUXES We employed the LMDI decomposition analysis

method38,63 to evaluate the impact of socio-economic development on China’s CH4 flux variations. The LMDI method can quantify the contributions from each explained factor to the total CH4

flux variations in the combination of all factors, which has been widely applied in analyzing emission trends and drivers38,64. Here, we decomposed China’s CH4 fluxes into three factors

related to anthropogenic activities to elucidate the underlying drivers behind flux variations across time, as shown in Eq. 6. $${F}_{C{H}_{4}}=\frac{{F}_{C{H}_{4}}}{G}\cdot \frac{G}{P}\cdot

P$$ (6) where _G_ and _P_ represent the GDP and total population in China, respectively, derived from the National Bureau of Statistics of China (https://www.stats.gov.cn/sj/).

\(\frac{{F}_{C{H}_{4}}}{G}\) is CH4 fluxes per unit of GDP, approximating the CH4 emission intensity. \(\frac{G}{P}\) is GDP per capita, an indicator of economic development level. We

quantified the contributions of these three key factors—emission intensity, economic development level, and population scale—to the variations of total CH4 fluxes across four distinct

periods: 2000–2005, 2005–2010, 2010–2015, and 2015–2021. UNCERTAINTY EVALUATION We evaluated the uncertainty of the estimated \({F}_{C{H}_{4}}\) using an error propagation method that

incorporates uncertainties in \({F}_{CO}\), _EnhR_, and _CF_ (Eq. 7). The uncertainty of \({F}_{CO}\) was assigned a uniform error distribution of ±20%, based on our previous

assessments23,49. The uncertainties of _EnhR_ originated from the configuration of the HYSPLIT model and the linear fitting method, the representativeness of observation at a single station,

and the slopes of the linear fit (Eq. 8). The uncertainty in _CF_ was considered to be mainly caused by the parameter configurations in our study (Eq. 9). $${\sigma

}_{{F}_{C{H}_{{4}}},i}=\sqrt{{({\sigma }_{EnhR,i}/Enh{R}_{i})}^{2}+{({\sigma }_{{F}_{CO},i}/{F}_{CO,i})}^{2}+{({\sigma }_{CF,i}/C{F}_{i})}^{2}}\times {F}_{{C{H}_{{4}}},i}$$ (7) $${\sigma

}_{EnhR,i}=\sqrt{{({\sigma }_{conf,EnhR})}^{2}+{({\sigma }_{slope,i})}^{2}+{({\sigma }_{repr})}^{2}}$$ (8) $${\sigma }_{CF,i}={\sigma }_{conf,CF}$$ (9) where \({\sigma }_{{F}_{C{H}_{4}}}\)

denotes the uncertainty of estimated CH4 fluxes. \({\sigma }_{{F}_{CO}}\), \({\sigma }_{EnhR}\), and \({\sigma }_{CF,i}\) are the uncertainties of \({F}_{CO}\), _EnhR_, and _CF_,

respectively. \({\sigma }_{conf}\) is the uncertainty that arises from the configuration and settings of the HYSPLIT model, linear fitting method, and relevant parameters, which introduces

errors in both _EnhR_ (\({\sigma }_{conf,EnhR}\)) and _CF_ (\({\sigma }_{conf,CF}\)) estimations. \({\sigma }_{slope}\) and \({\sigma }_{repr}\) are uncertainties for the slope of the linear

fit and the representativeness of a single station, respectively. The assessment of \({\sigma }_{conf}\) was based on a series of sensitivity tests on key parameters and steps involved in

estimating _EnhR_ and _CF_ (SI Appendix, Table S3). The deviation of sensitivity test results from the reference inversion results was used to approximate the corresponding error ranges

caused by _EnhR_ and _CF_. For \({\sigma }_{slope}\), we calculated the standard errors of the seasonal slopes for each year using the Monte Carlo method, performing 1000 iterative

calculation tests. Then, we applied the error propagation to evaluate the annual errors caused by corresponding seasonal errors. Considering the difficulty in accurately quantifying

\({\sigma }_{repr}\), we set its error range by referencing the discrepancies of the mean ratios of CEDS CH4 emissions (v2021_04_21) to our CO fluxes between 2000 and 2019 across different

spatial and seasonal scales, as well as with our results for the same period (SI Appendix, Fig. S2). The spatial division was based on areas traversed by different numbers of back

trajectories within a grid cell (0.5° × 0.5°) per year, using the thresholds of larger than 0, 2, 10, 30, 60, and 100 trajectories on an annual and seasonal basis. DATA AVAILABILITY Data are

provided within the manuscript or supplementary information files. CODE AVAILABILITY The code for HYSPLIT Model is available at https://www.ready.noaa.gov/HYSPLIT.php. REFERENCES *

Intergovernmental Panel on Climate Change (IPCC). Short-Lived Climate Forcers. _Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment

Report of the Intergovernmental Panel on Climate Change_ (Cambridge University Press, 2023). * Howarth, R. W., Santoro, R. & Ingraffea, A. Methane and the greenhouse-gas footprint of

natural gas from shale formations. _Clim. Change_ 106, 679–690 (2011). Article  CAS  Google Scholar  * Shindell, D. et al. Simultaneously mitigating near-term climate change and improving

human health and food security. _Science_ 335, 183–189 (2012). Article  CAS  Google Scholar  * UNFCCC. The Paris Agreement.

https://unfccc.int/process-and-meetings/the-paris-agreement/the-paris-agreement (2015). * Nisbet, E. G. New hope for methane reduction. _Science_ 382, 1093–1093 (2023). Article  Google

Scholar  * Houweling, S. et al. Global inverse modeling of CH4 sources and sinks: an overview of methods. _Atmos. Chem. Phys._ 17, 235–256 (2017). Article  CAS  Google Scholar  * Maasakkers,

J. D. et al. 2010–2015 North American methane emissions, sectoral contributions, and trends: a high-resolution inversion of GOSAT observations of atmospheric methane. _Atmos. Chem. Phys._

21, 4339–4356 (2021). Article  CAS  Google Scholar  * Yang, X. et al. City-scale methane emissions from the midstream oil and gas industry: a satellite survey of the Zhoushan archipelago.

_J. Clean. Prod._ 449, 141673 (2024). Article  CAS  Google Scholar  * Niwa, Y. et al. Multi-observational estimation of regional and sectoral emission contributions to the persistent high

growth rate of atmospheric CH4 for 2020–2022. _EGUsphere_ 2024, 1–43 (2024). Google Scholar  * Hemati, M., Mahdianpari, M., Nassar, R., Shiri, H. & Mohammadimanesh, F. Urban methane

emission monitoring across North America using TROPOMI data: an analytical inversion approach. _Sci. Rep._ 14, 9041 (2024). Article  CAS  Google Scholar  * Qu, Z. et al. Global distribution

of methane emissions: a comparative inverse analysis of observations from the TROPOMI and GOSAT satellite instruments. _Atmos. Chem. Phys._ 21, 14159–14175 (2021). Article  CAS  Google

Scholar  * Hasekamp, O., Litvinov, P., Fu, G., Chen, C. & Dubovik, O. Algorithm evaluation for polarimetric remote sensing of atmospheric aerosols. _Atmos. Meas. Tech._ 17, 1497–1525

(2024). Article  CAS  Google Scholar  * Nesser, H. et al. High-resolution US methane emissions inferred from an inversion of 2019 TROPOMI satellite data: contributions from individual

states, urban areas, and landfills. _Atmos. Chem. Phys._ 24, 5069–5091 (2024). Article  CAS  Google Scholar  * Jacob, D. J. et al. Quantifying methane emissions from the global scale down to

point sources using satellite observations of atmospheric methane. _Atmos. Chem. Phys._ 22, 9617–9646 (2022). Article  CAS  Google Scholar  * Shen, L. et al. Satellite quantification of oil

and natural gas methane emissions in the US and Canada including contributions from individual basins. _Atmos. Chem. Phys._ 22, 11203–11215 (2022). Article  CAS  Google Scholar  * Lu, X. et

al. Observation-derived 2010-2019 trends in methane emissions and intensities from US oil and fields tied to metrics. _Proc. Natl. Acad. Sci. USA_ 120, e2217900120 (2023). Article  CAS 

Google Scholar  * Miller, S. M. et al. China’s coal mine methane regulations have not curbed growing emissions. _Nat. Commun._ 10, 303 (2019). Article  CAS  Google Scholar  * Thompson, R. L.

et al. Methane emissions in East Asia for 2000–2011 were estimated using an atmospheric Bayesian inversion. _J. Geophys. Res. Atmos._ 120, 4352–4369 (2015). Article  CAS  Google Scholar  *

Maasakkers, J. D. et al. Global distribution of methane emissions, emission trends, and OH concentrations and trends inferred from an inversion of GOSAT satellite data for 2010–2015. _Atmos.

Chem. Phys._ 19, 7859–7881 (2019). Article  CAS  Google Scholar  * Bergamaschi, P. et al. Inverse modeling of global and regional CH4 emissions using SCIAMACHY satellite retrievals. _J.

Geophys. Res. Atmos._ 114, JRC51693 (2009). * Zhang, Y. et al. Observed changes in China’s methane emissions linked to policy drivers. _Proc. Natl Acad. Sci._ 119, e2202742119 (2022).

Article  CAS  Google Scholar  * Janardanan, R. et al. 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 (2024). Article  Google Scholar  * Zheng, B. et al. Global atmospheric carbon monoxide budget 2000–2017 inferred from multi-species atmospheric

inversions. _Earth Syst. Sci. Data_ 11, 1411–1436 (2019). Article  Google Scholar  * Zheng, B. et al. Rapid decline in carbon monoxide emissions and export from East Asia between years 2005

and 2016. _Environ. Res. Lett._ 13, 044007 (2018). Article  Google Scholar  * Zheng, B. et al. Trends in China’s anthropogenic emissions since 2010 as the consequence of clean air actions.

_Atmos. Chem. Phys._ 18, 14095–14111 (2018). Article  CAS  Google Scholar  * Peng, S. et al. Inventory of anthropogenic methane emissions in mainland China from 1980 to 2010. _Atmos. Chem.

Phys._ 16, 14545–14562 (2016). Article  CAS  Google Scholar  * Chen, Z. et al. Methane emissions from China: a high-resolution inversion of TROPOMI satellite observations. _Atmos. Chem.

Phys._ 22, 10809–10826 (2022). Article  CAS  Google Scholar  * O'Rourke, P. R. et al. CEDS v_2021_04_21 Release Emission Data, version v_2021_02_05. _Zenodo_,

https://doi.org/10.5281/zenodo.4741285 (2021). * Saunois, M. et al. Global Methane Budget 2000-2020. _Earth Syst. Sci. Data Discuss._ 2024, 1–147 (2024). Google Scholar  * Liu, G. et al.

Recent slowdown of anthropogenic methane emissions in china driven by stabilized coal production. _Environ. Sci. Technol. Lett._ 8, 739–746 (2021). Article  CAS  Google Scholar  * Hoesly, R.

et al. CEDS v_2024_07_08 Release Emission Data, v_2024_07_08. _Zenodo_, https://doi.org/10.5281/zenodo.12803197 (2024). * Crippa, M. et al. GHG emissions of all world countries.

_Publications Office of the European Union_ (Luxembourg, 2023). * Lin, X. et al. A comparative study of anthropogenic CH4 emissions over China based on the ensembles of bottom-up

inventories. _Earth Syst. Sci. Data_ 13, 1073–1088 (2021). Article  Google Scholar  * Patra, P. K. et al. Regional methane emission estimation based on observed atmospheric concentrations

(2002-2012). _J. Meteorol. Soc. Jpn._ 94, 91–113 (2016). Article  Google Scholar  * Wada, A., Matsueda, H., Sawa, Y., Tsuboi, K. & Okubo, S. Seasonal variation of enhancement ratios of

trace gases observed over 10 years in the western North Pacific. _Atmos. Environ._ 45, 2129–2137 (2011). Article  CAS  Google Scholar  * Tu, M. et al. Logarithmic mean divisia index

decomposition of CO2 emissions from urban passenger transport: an empirical study of global cities from 1960–2001. _Sustainability_ 11, 4310 (2019). Article  CAS  Google Scholar  * Jeong, K.

& Kim, S. LMDI decomposition analysis of greenhouse gas emissions in the Korean manufacturing sector. _Energy Policy_ 62, 1245–1253 (2013). Article  CAS  Google Scholar  * Xiang, X.,

Ma, X., Ma, Z., Ma, M. & Cai, W. Python-LMDI: a tool for index decomposition analysis of building carbon emissions. _Buildings_ 12, 83 (2022). Article  Google Scholar  * Li, D. X. Effect

of water-saving irrigation on CH4 emissions from rice fields. _Adv. Mater. Res._ 396-398, 1950–1958 (2012). Article  CAS  Google Scholar  * Yan, X., Akiyama, H., Yagi, K. & Akimoto, H.

Global estimations of the inventory and mitigation potential of methane emissions from rice cultivation conducted using the 2006 intergovernmental panel on climate change guidelines. _Glob.

Biogeochem. Cycles_ 23, GB2002 (2009). * Wang, J. et al. The role of rice cultivation in changes in atmospheric methane concentration and the global methane pledge. _Glob. Change Biol._ 29,

2776–2789 (2023). Article  CAS  Google Scholar  * Liu, Q., Teng, F., Nielsen, C. P., Zhang, Y. & Wu, L. Large methane mitigation potential through prioritized closure of gas-rich coal

mines. _Nat. Clim. Change_ 14, 652–658 (2024). Article  Google Scholar  * Wang, H., Zheng, B., Li, H., Zhang, Q. & He, K. Targeting ultra-emitters for effective methane mitigation in

China’s coal sector. _ACS ES&T Air_ 1, 734–742 (2024). Article  Google Scholar  * Guilhot, L. An analysis of China’s energy policy from 1981 to 2020: transitioning towards to a

diversified and low-carbon energy system. _Energy Policy_ 162, 112806 (2022). Article  Google Scholar  * Chandra, N. et al. Methane emissions decreased in fossil fuel exploitation and

sustainably increased in microbial source sectors during 1990–2020. _Commun. Earth Environ._ 5, 147 (2024). Article  Google Scholar  * Yan, X. et al. Spatial and temporal distribution of

greenhouse gas emissions from municipal wastewater treatment plants in China from 2005 to 2014. _Earth’s Future_ 7, 340–350 (2019). Article  CAS  Google Scholar  * Li, S. et al. Source

identification and apportionment of halogenated compounds observed at a remote site in East Asia. _Environ. Sci. Technol._ 48, 491–498 (2014). Article  CAS  Google Scholar  * Choi, H. et al.

Top-down and bottom-up estimates of anthropogenic methyl bromide emissions from eastern China. _Atmos. Chem. Phys._ 22, 5157–5173 (2022). Article  CAS  Google Scholar  * Zheng, B. et al.

Record-high CO2 emissions from boreal fires in 2021. _Science_ 379, 912–917 (2023). Article  CAS  Google Scholar  * Lutsch, E. et al. Detection and attribution of wildfire pollution in the

Arctic and northern midlatitudes using a network of Fourier-transform infrared spectrometers and GEOS-Chem. _Atmos. Chem. Phys._ 20, 12813–12851 (2020). Article  CAS  Google Scholar  *

Zheng, B. et al. Increasing forest fire emissions despite the decline in global burned area. _Sci. Adv._ 7, eabh2646 (2021). Article  Google Scholar  * Pison, I., Bousquet, P., Chevallier,

F., Szopa, S. & Hauglustaine, D. Multi-species inversion of CH4, CO and H2 emissions from surface measurements. _Atmos. Chem. Phys._ 9, 5281–5297 (2009). Article  CAS  Google Scholar  *

Zheng, B., Chevallier, F., Ciais, P., Yin, Y. & Wang, Y. On the role of the flaming to smoldering transition in the seasonal cycle of African fire emissions. _Geophys. Res. Lett._ 45,

11,998–912,007 (2018). Article  CAS  Google Scholar  * Yin, Y. et al. Variability of fire carbon emissions in equatorial Asia and its nonlinear sensitivity to El Niño. _Geophys. Res. Lett._

43, 10,472–410,479 (2016). Article  Google Scholar  * Jiang, Z. et al. A 15-year record of CO emissions constrained by MOPITT CO observations. _Atmos. Chem. Phys._ 17, 4565–4583 (2017).

Article  CAS  Google Scholar  * Shinya, T. Atmospheric CH4 at Yonagunijima by Japan Meteorological Agency, _The World Data Centre for Greenhouse Gases_, 2024-07-29-0428,

https://doi.org/10.50849/WDCGG_0001-2028-1002-01-01-9999 (2024). * Shinya, T. Atmospheric CH4 at Minamitorishima by Japan Meteorological Agency, _The World Data Centre for Greenhouse Gases_,

2024-07-29-0428, https://doi.org/10.50849/WDCGG_0001-2029-1002-01-01-9999 (2024). * Shinya, T. Atmospheric CO at Minamitorishima by Japan Meteorological Agency, _The World Data Centre for

Greenhouse Gases_, 2024-07-29-0428, https://doi.org/10.50849/WDCGG_0001-2029-3001-01-01-9999 (2024). * Shinya, T. Atmospheric CO at Yonagunijima by Japan Meteorological Agency, _The World

Data Centre for Greenhouse Gases_, 2024-07-29-0428, https://doi.org/10.50849/WDCGG_0001-2028-3001-01-01-9999 (2024). * Stein, A. F. et al. NOAA’s HYSPLIT atmospheric transport and dispersion

modeling system. _Bull. Am. Meteorol. Soc._ 96, 2059–2077 (2015). Article  Google Scholar  * Zhao, Y. et al. Inter-model comparison of global hydroxyl radical (OH) distributions and their

impact on atmospheric methane over the 2000–2016 period. _Atmos. Chem. Phys._ 19, 13701–13723 (2019). Article  CAS  Google Scholar  * Zhao, Y. et al. Reconciling the bottom-up and top-down

estimates of the methane chemical sink using multiple observations. _Atmos. Chem. Phys._ 23, 789–807 (2023). Article  CAS  Google Scholar  * Rivera-Niquepa, J. D., Rojas-Lozano, D., De

Oliveira-De Jesus, P. M. & Yusta, J. M. Methodology for selecting assessment periods of Logarithmic Mean Divisia Index decomposition techniques. _Energy Strateg. Rev._ 50, 101241 (2023).

Article  Google Scholar  * Ang, B. W. & Liu, N. Negative-value problems of the logarithmic mean divisia index decomposition approach. _Energy Policy_ 35, 739–742 (2007). Article  Google

Scholar  * Kurokawa, J. et al. Emissions of air pollutants and greenhouse gases over Asian regions during 2000–2008: Regional Emission inventory in ASia (REAS) version 2. _Atmos. Chem.

Phys._ 13, 11019–11058 (2013). Article  CAS  Google Scholar  * Höglund-Isaksson, L. Global anthropogenic methane emissions 2005–2030: technical mitigation potentials and costs. _Atmos. Chem.

Phys._ 12, 9079–9096 (2012). Article  Google Scholar  Download references ACKNOWLEDGEMENTS This work was supported by the National Natural Science Foundation of China (Grant no. 42375096).

We acknowledge the Japan Meteorological Agency (JMA) and the World Meteorological Organization (WMO) for providing both CO and CH4 observations at YON and MNM surface stations in the World

Data Center for Greenhouse Gases (WDCGG). We are grateful to all those involved in maintaining the network and archiving the observation data. We also thank the Global Methane Budget (GMB)

inverse modeling group for the production of the global gridded CH4 budget inversion results. AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * Institute of Environment and Ecology, Tsinghua

Shenzhen International Graduate School, Tsinghua University, Shenzhen, China Yifan Li & Bo Zheng * State Environmental Protection Key Laboratory of Sources and Control of Air Pollution

Complex, Beijing, China Bo Zheng Authors * Yifan Li View author publications You can also search for this author inPubMed Google Scholar * Bo Zheng View author publications You can also

search for this author inPubMed Google Scholar CONTRIBUTIONS Conceptualization: B.Z., Methodology: B.Z. and Y.L., Investigation: B.Z. and Y.L., Visualization: Y.L., Funding acquisition:

B.Z., Project administration: B.Z., Supervision: B.Z., Writing–original draft: Y.L., Writing–review and editing: B.Z. and Y.L. CORRESPONDING AUTHOR Correspondence to Bo Zheng. ETHICS

DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. ADDITIONAL INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional claims in

published maps and institutional affiliations. SUPPLEMENTARY INFORMATION SUPPLEMENTARY INFORMATION RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons

Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original

author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the

article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use

is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit

http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Li, Y., Zheng, B. Tracking regional CH4 emissions through collocated air pollution

measurement: a pilot application and robustness analysis in China. _npj Clim Atmos Sci_ 8, 127 (2025). https://doi.org/10.1038/s41612-025-01011-1 Download citation * Received: 22 November

2024 * Accepted: 17 March 2025 * Published: 28 March 2025 * DOI: https://doi.org/10.1038/s41612-025-01011-1 SHARE THIS ARTICLE Anyone you share the following link with will be able to read

this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative