Space-based inversion tracks and attributes Shanxi's under-estimated carbon monoxide emissions

Space-based inversion tracks and attributes Shanxi's under-estimated carbon monoxide emissions | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Space-based inversion tracks and attributes Shanxi's under-estimated carbon monoxide emissions Jason Cohen, Xiaolu Li, Pravash Tiwari, Liling Wu, Shuo Wang, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4604393/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 08 May, 2025 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Abstract A space-based mass-conserving framework using observed carbon monoxide (CO) and formaldehyde (HCHO) columns quantifies day-to-day and grid-to-grid CO emissions over energy-consuming Shanxi. Annualized total emissions are 8 times higher than a priori datasets, especially over low emission areas, resulting in an at least 7% increase in CO 2 emissions. Significant forcings include atmospheric lifetime of CO (0.3–16.5 d) and HCHO (0.1-6.5h), and transport. Annual CO emissions decreased year-by-year, although this is only obvious when considering the two to three highest months. The ratio of top-down CO to NO x emissions show source attribution is possible over rural, urban, and five industrial areas (including power, iron/steel, and coke). Cross-border transport of CO is important in the peak emission months, including evolving sources from central Shaanxi and western Hebei. The major reason for the significant increase CO emissions is the fractional increase in non-high emitting area’s energy consumption, resulting in a spatial mis-alignment. Earth and environmental sciences/Environmental sciences/Environmental chemistry/Atmospheric chemistry Earth and environmental sciences/Climate sciences/Atmospheric science/Atmospheric chemistry Earth and environmental sciences/Climate sciences/Climate change/Attribution Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction A variety of atmospheric pollutants are emitted from energy use and industrial processes, including particulate matter, nitrogen oxides (NO x ), carbon monoxide (CO), and other trace gases. Emitted species may also undergo atmospheric chemical processing and be transported by the wind, which in tandem affect the balance of atmospheric oxidation, chemical properties and the ultimate concentration of these species. CO is one such trace gas that impacts air quality, global climate forcing, and the budget and distribution of hydroxyl (OH) radical, thereby affecting the gas-phase chemistry of both methane (CH 4 ) and ozone (O 3 ) 1–3 . Although the direct radiative forcing of CO is relatively small, because of its impacts on the concentrations of the greenhouse gasses CH 4 and O 3 4 , as well as its conversion into carbon dioxide (CO 2 ), emissions of CO contribute significantly to the global radiative balance 5 . CO is produced by incomplete combustion of carbon-based fuels, biomass burning, and chemical oxidation of volatile organic compounds (VOCs) 6 . Its co-emission with NO x depends on thermodynamic conditions, and oxygen and nitrogen availability at the time of combustion 7,8 . From the perspective of bottom-up emission inventories, CO and NO x emissions are calculated by applying emission factors to a set of activity data. Previous studies over China have indicated that biofuels and biomass burning contribute nearly half of CO emissions but only a tenth of NO x emissions 9–11 . High-temperature combustion processes associated with transportation, power generation, and iron and steel production result in significant differences in CO and NO x emissions 8,9 . Therefore, analyzing the CO/NO x ratio variation can provide valuable insights for further investigation and attribution 2 . Discrepancies between CO and NO x representation in emission inventories may lead to under/overestimation from specific sources, posing challenges for current atmospheric models 12 . This is one possible reason why these models still struggle to accurately reproduce observed long-term changes even after more than a decade development 13,14 . Environmental management of CO monitoring and supervision are limited compared with other criteria pollutants. While there are national control sites for CO concentration monitoring and ambient air quality standards that must be followed across many countries 15,16 , the management of CO emissions is almost blank, with most management based on those criteria pollutants that contribute more to the air quality index (AQI), due to CO’s very high ambient air quality standard and AQI cutoff 17 . Specifically in China, policy-based controls for CO emission are primarily implemented in waste incineration and a few other industries 18 , with only a few cities such as Tangshan, Handan and Linfen which all have significant numbers of iron and steel factories introducing CO controls in the iron and steel industry. Moreover, the continuous emissions monitoring systems (CEMS) in China and other counties do not provide CO data, hindering policy-based controls on a stack-by-stack basis 8,19 . Therefore, detailed quantification of CO emissions over a region with various sources can yield significant insights and potentially enhance environmental management efforts and synergistic reduction of air pollution and greenhouse gasses 20–24 . This study employs a top-down constrained mass-conserving inversion method to estimate the daily CO emissions over a mesoscale grid (0.05°×0.05°) from May 2018 through April 2022, utilizing daily observations of the TROPOspheric Monitoring Instrument (TROPOMI) CO and HCHO. A spatially and temporally consistent top-down constrained NO x emission inventory, calculated using the same algorithm, serves as the a priori constraint 8 , enabling interrelationships between these co-emitted pollutants to be established based on the best fit properties of in-situ production, transport, and processing in tandem month-to-month, source-by-source, and grid-by-grid. The underlying CO a priori emissions used in this work are based on daily and grid-by-grid TROPOMI NO 2 constrained emissions and bottom-up CO to NO x ratio over Shanxi province. The geospatial area (showed in Fig. 1 ) is selected since it is an energy-rich region in Northern China that produces more than a quarter of China’s coal (including about half of China’s coking coal) and consumes nearly a tenth 25,26 . The geography is also unique with mountains and basins contributing the majority of surface area, and generally low cloud cover, leading to intense atmospheric processing and observed concentrations not encountered in previous studies. However, this unique set of conditions allows an ideal laboratory to study changes that occur during the combustion and utilization of coal, co-emitted pollutants, and their interrelationships. Results and Discussion Spatial and Temporal Distribution of CO Emissions The emeission was calculated over the domain from 34°N to 41°N and 110°E to 115°E inside the Loess Plateau of China. The yearly total CO emissions (hereafter EI CO ), uncertainty range, and day-to-day variability over Shanxi is 30.4, 15.0, and 17.1 Tg yr − 1 , respectively, showed in Fig. 2 a- 2 b. Year-to-year values for total emissions, uncertainty range, and daily variability (from May to April of the following year) are as follows: 38.1, 18.9, and 23.4 Tg yr − 1 of 2018 ~ 2019, 32.7, 16.1, and 16.3 Tg yr − 1 of 2019 ~ 2020, 27.4, 14.1, and 12.2 Tg yr − 1 of 2020 ~ 2021, 24.9, 11.7, and 12.3 Tg yr − 1 of 2021 ~ 2022. CO emission intensity decreases from 2018 to 2022, consistent with the Chinese Government’s long-term air quality management strategy 27–30 . However, significant temporal variation is observed, the highest emissions occurring in December to January each year, coinciding with the end of the industrial cycle and the start of the Chinese New Year holiday (late January through early February). Total emissions during this period decrease notably, with values of 12.9, 8.7, 6.0, and 5.7 Tg yr − 1 for each year. A small peak is also observed around May, occurring around May with variations in its start and end points. When data from December, January, and May are removed, the emissions do not statistically decrease every year, contradicting claims that COVID-19 dominates emissions changes 31 . The only discernible impact from COVID-19 appears to be a slight shortening of the high-emission period in late January and February of 2020. A similar shortening of a high emissions time is observed during the 2022 Winter Olympic Games 32 . The observed changes are smaller than the year-to-year reduction in both cases. These observations differ from top-down NO x emissions 8 , as showed in Supplementary Fig. 1. However, this disparity is not inconsistent with policy objectives, as efforts primarily target reducing NO x emissions, indirectly affecting CO as an accompanying species. Given the diversity of emissions across different parts of Shanxi, a detailed look as been made over three emissions intensity regions ( 9.5 µg \(\) m −2 \(\) s − 1 ) (Fig. 2 d). Areas with relatively high CO emissions are mainly concentrated in the lower Fen River valley (7 ~ 9.5 µg \(\) m −2 \(\) s − 1 ), which accounts for the majority of population and industry, including coking and other industrial facilities. The highest intensity (> 9.5 µg \(\) m −2 \(\) s − 1 ) occurs within parts of Linfen and Yuncheng where located a large number of iron and steel factories. Additional explanation is in the section “Impacts of Variability and Long-range Transport”. The 2019 EI CO is compared with two widely used bottom- up emissions datasets: 2019 MEIC and 2018 EDGAR (Fig. 3 ). Those grids with higher emission in either MEIC or EDGAR (> 9.5 µg \(\) m −2 \(\) s − 1 ), EI CO is notably lower, averaging 12.1 ± 5.3 µg \(\) m −2 \(\) s − 1 , compared to MEIC and EDGAR averages of 17.7 and 64.6 µg \(\) m −2 \(\) s − 1 respectively. This aligns with a gradual reduction in emissions from well-regulated sources like urban centers and large factories. Conversely, in grids with low initial emissions (< 4.5 µg \(\) m −2 \(\) s − 1 ), EI CO tends to be consistently higher, averaging 3.5 ± 2.8 µg \(\) m −2 \(\) s − 1 , while MEIC and EDGAR average much lower at 0.8 and 0.3 µg \(\) m −2 \(\) s − 1 respectively. The reason is three-fold: first, there are many low values close to zero in MEIC and EDGAR characterized by rural residential burning and wildfires; second, there is an increase in small and moderate sources as non-urban income increases; third, the bottom-up datasets underestimate CO emissions from the rapid increase in iron and steel enterprises in Yuncheng and Linfen. Assuming all additional CO will ultimately decay into CO 2 , the increase of EI CO to MEIC for 2019/2020 are 31.8 and 23.5 Tg yr − 1 , leading to a corresponding increase in CO 2 emissions of 49.9 and 37.0 Tg yr − 1 . This implies a CO 2 emissions increase of approximately 9.8%, 7.0% over Shanxi and 0.5%, 0.4% over China based on MEIC CO 2 in 2019 and 2020 33 . Sensitivity and Robustness of Calculated CO Emissions Sensitivity tests are performed to actively account for the ranges of uncertainties of four different input observations used to compute CO emissions, as shown in Fig. 4 a- 4 g for seven cases (± 30% TROPOMI CO, ± 40% TROPOMI HCHO, ± 30% a priori emission of CO, and different pressure level of wind). The remaining cases can be found in Supplementary Fig. 2. First, in all cases the emissions computed in all cases are stable: there are no new peaks or troughs in the spatial results, and no significant changes in the temporal profiles. Second, the emissions computed in all cases are robust: the resulting computed emissions always have a smaller difference than the magnitude of the perturbations (the range of fractional changes in the seven cases in Fig. 4 are 1.03–1.26, 1.06–1.27, 0.98–1.33 ,0.82–1.21, 0.78–0.98, 0.76–0.95, and 0.93–1.13, respectively). Adjusting the wind pressure level from 850hPa to 900 hPa (showed in Fig. 4 d) leads to an increase in emissions within the high-altitude areas of northern Shanxi simultaneously with a reduction of emissions in the basin areas, with all change less than ± 19%. Intriguingly, in the central and southern regions such as Linfen, Yuncheng, and Jincheng, where CO emission levels are relatively high, emissions still increase despite the altitudes being similarly low to the basin areas. These findings are consistent with the methodology, since any abnormally large or small driving factors computed due to the observational uncertainties are filtered out if physically unrealistic. Similarly, the non-linear effects of uncertainty on the spatial gradient terms are weighted by the linear production and loss terms, ensuring that fluctuations in these components counterbalance one another, bolstering confidence in the reliability and quality of the results. Out of all terms analyzed, the term with the greatest difference between the plus and minus uncertainty is the uncertainty in a priori emissions. Specifically, the impact is larger when the + 30% uncertainty is used as compared to when the − 30% uncertainty, while all of the other terms have a somewhat balanced impact on emissions between the positive and negative uncertainty perturbations. This bias indicates that the a priori emissions is very low and does not provide as much information as a higher a priori dataset would. Overall, this finding provides further support for the results herein which indicate that the emissions are much higher after optimization. Decay and Production of CO The results of the CO’s lifetime are a function of the reaction between CO and the OH radical, forming CO 2 , and mixing between the high concentration plume of CO both as it is emitted and as it evolves in the atmosphere with respect to the lesser polluted air immediately surrounding it within an observed grid. Due to the extremely high level of local CH 4 in Shanxi province 34 and the local emissions of VOCs due to coal to chemicals industries, it is important to also consider the production of CO in-situ. However, due to limited validated global observations from remote sensing platforms, this work uses HCHO as a proxy for CO production, given HCHO’s reasonable retrieval properties 35 . The mean, 10th and 90th percentile of the CO and HCHO atmospheric in-situ net processing decay time are [2.9,1.0, 5.7] d, and [1.0, 0.2, 2.3] hours. The computed lifetime of CO is consistent with observed values of OH over other highly polluted and relatively drier areas 36–39 , as well as downwind from major forest fires 40 . Figure 5 illustrates the monthly distribution of CO and HCHO lifetime in Shanxi. There are two reasons why CO's lifetime has increased from 2018 to 2022 while HCHO's lifetime has not. First there is an an overall decrease in OH concentration 41 consistent with the documented rise in CH 4 emissions due to increased coal production from high coal-bed methane sources in Shanxi over the past decades 25 . Second, changes in the concentrations of air in the surrounding basins due to different controls and economic growth play a non-linear role in the heavily mountainous environment in terms of subgrid mixing between the different atmospheric environments. These factors emphasize the importance of accurately constraining CO emissions and associated processes in tandem. The lifetime of CO is observed to be relatively longer in May and July, and relatively shorter in December and January, consistent with local observations. First, this region has less stable atmosphere in the warmer months, coupled with a relatively dry atmosphere year-round, leading to a higher cloud optical depth in May and July 42 . Secondly, there is a relative large number of absorbing aerosols emitted here due to coal consumed 43,44 , combined with enhanced particle aging in the higher UV seasons 45 and under higher temperature conditions 46 leading to a relatively higher column absorption of UV as secondary aerosol coats and mixes with locally emitted BC 47 , which in turn absorbs more UV from the column when the solar zenith angle is higher, maximizing in May 8 . Observations of the monthly average absorbing aerosol optical depth (AAOD) from Multi-Angle Imaging SpectroRadiometer (MISR) over Shanxi (0.0039 March 2019 to February 2021) confirm that the value is slightly higher in May (0.0063 in 2019 and 0.0071 in 2020), while lower than or similar to the background in December and January (0.0041 in 2019–2020 and 0.0032 in 2020–2021) 48 , as detailed in Supplementary Fig. 3. These factors lead to reduced UV within the lower troposphere, which in turn leads to an influence on OH production due to the joint limitations on UV and water vapor 49 . Third, NO x loadings are observed to be lower during the warmer months. Fourth, the rate of CO decrease is far smaller than that of NO x . This combination leads to a lower net atmospheric oxidation potential. The results are consistent with both the underlying theory and the actual observations in this region, calling into question whether modeled oxidant and CO fields over this region should be examined more carefully. The shorter lifetime of HCHO leadings to more CO production. α 4 is relatively stable, with only a significant change in September and October. However, the ratio between the lifetime of CO and HCHO in two specific sub-regions within Shanxi is larger (Supplementary Fig. 4). They are known to be upwind from other major sources adjacent to Shanxi and have topographic gaps. This is consistent with the major factors in these regions being influenced by emissions and subsequent long-range transport from far-away industrial and urban areas 48 . This effect is increased due to diurnal changes in upslope and downslope mountain winds when more polluted air is buttressed against the outside of the mountains surrounding Shanxi downwind from these external sources. The effect is further supported by the high loadings by the high loadings of absorbing aerosol co-emitted from these sources, which reduces UV at the surface as the air is transported over multiple days from these sources to Shanxi. Attributing Sources Using Emissions Ratios The amounts of CO and NO x emitted are a complex function of energy efficiency, combustion temperature, oxygen availability, and others, with variations of up to an order of magnitude. Thus, it is of interest to investigate the relationship between both species and use this to identify and attribute different source type 2,50 . The ratio of CO emissions to NO x emissions, hereafter called [CO/NO x ], is computed grid-by-grid and day-by-day. On average, CO/NO x vary consistently across four land use types (Fig. 6 a). Urban and industrial generally have a lower CO/NO x than rural areas, which lower than natural areas. Due to the fact that industrial and urban areas have more high-temperature combustion sources associated with power, transportation and gas burning sectors 9 , leading to more NO x and lower CO. Rural areas still have mixed small industry, boilers, and biomass burning leading to less NO x and more CO. Natural areas tend to be protected or have little anthropogenic influence. A detailed analysis is performed by categorizing industrial sub-types identified using CEMS 8 . Cement and power have the lowest median CO/NO x , boilers and coke show intermediate median CO/NO x , while iron and steel is highest. Notably, the CO/NO x ranges (between 75th and 25th percentile) for coke and boilers are wider than others, while the range for iron and steel is smallest. Iron and steel has both the highest 25th percentile and median value, with its 25th percentile value similar to the 75th percentile for cement and power and similar to the median for coke and boiler, while its 75th percentile is slightly lower than coke. Despite high combustion temperatures in iron and steel production, resulting in high NO x emissions, the CO/NO x ratio remains high due to significant CO emissions from sintering and blast furnaces, since carbonaceous material (including metallurgical coke, coal and natural gas) are used as to reduce iron oxide to iron, in addition to being directly used as fuel for combustion. Cement's CO/NO x ratio slightly exceeds power across all percentiles due to higher NO x emissions from rotary kilns. Considering the length, consistency, and temperature of combustion, the overall lesser availability of oxygen leads to more CO than power and slightly bigger CO/NO x observed. Coke and boilers have similar median and 25th percentile values, the values above the median (75th and 90th percentile) and the 10th percentile are smaller for boilers. Both coke and boilers have larger 25th and 50th percentile values compared to power and cement, but lower than iron and steel. The 75th and 90th percentile values of coke are the highest. The main fuel during coke production is natural gas, wherein the combustion process is meant to cook the coal, turning it into coke. The produced CO retained in the coke oven gas instead of being emitted directly. Furthermore, some coke plants have low CO emissions due to the lack of oxygen in charring chambers, which in turn convert a significant amount of coal that is lost (not converted into coke) into black carbon instead of CO. Leakage from combustion chamber to charring chamber in coke ovens, influenced by their design, age, and efficiency, can result in extra CO generation due to the influx of additional oxygen. The combination leads to the overall wide range of CO/NO x . Boilers, operating at lower capacity and efficiency than power plants, hence contributing to lower NO x per unit of coal consumed and a greater potential for incomplete combustion and produce normal to high CO emissions, resulting in a relatively wide CO/NO x range. Impacts of Variability and Long-range Transport Due to its relatively longer lifetime, CO serves as an indicator of dynamical transport, while the daily average transport distance of CO is approximately 58 ~ 471 km (10th and 90th percentile), there is a wide range of variation, up to hundreds of km on specific days in specific parts of Shanxi. Since typical zonal and meridional transport within Shanxi occurs on the order of 1–2 weeks, the calculated lifetime of CO is capable of representing long range transport when its chemical lifetime is similarly long 52–54 . CO transport, as computed through its divergence, peaks during December and January and weakens in May 55 . A north-south corridor is observed from Linfen/Yuncheng to Taiyuan/Datong in the central basin area of Shanxi, particularly evident during December to January (Fig. 7 a). This corridor is surrounded by relatively highly polluted regions, continuously receiving inputs from various sources. The starting point of this corridor is found in Shaanxi Province, where there are strong urban and industrial sources in the megacity of Xi'an and the surrounding area. There is an observed weakening of the transport from 2018 to 2022. However, there is also an observed increase in the local source transport from Weinan, a smaller city in Shaanxi found just west of the Shanxi boarder, into Shanxi. This transition is consistent with policies to combat pollution in megacities such as Xi’an, and to encourage more development in smaller urban areas, such as Weinan. There is similarly a strong source within Yuncheng, that transports CO further upwind into Shanxi. The combination collectively contributes to the high emission intensity in Linfen and Yuncheng, as well as transport into the observed north-south corridor. In northeastern Shanxi bordering Hebei Province, cities like Shijiazhuang, Xingtai, and Handan exhibit strong CO sources, partly due to centralized steel enterprises and residential sources. Transport through Taihang Mountain only occurs via small gaps in Yangquan, Changzhi, and Jincheng at relatively low latitudes. During May the overall concentration of CO decreased while at the same time the long-range transport channels significantly reduced. As observed in Fig. 7 , the entirety of Yuncheng city and adjacent regions in Shaanxi have turned into sources, the central transport corridor has an insignificant amount of transport, and the neighboring regions in Hebei turn to sinks. Correspondingly, there is little to no transport either from outside or within Shanxi occurring in May. Discussion Daily CO emissions are computed using MFIEF based on remotely sensed CO and HCHO, considering first-order atmospheric processing and transport. High CO emissions are observed in densely populated and economically active areas, particularly in the lower Fen River valley, with the highest emissions concentrated around iron and steel factories in parts of Linfen and Yuncheng Cities. Under the requirements of multi-year air pollution control, CO emissions have declined significantly in December and January, when emissions were high because of the factories that increase production loads to fulfill annual schedules before the Chinese New Year break, as well as winter residential heating. CO emissions throughout other times do not have significant reduction. Differences between EI CO , EDGAR and MEIC suggest decreases in heavily industrialized areas and increases in low-emission zones. The total difference in EI CO and MEIC may result in CO 2 production at least 7% of the total CO 2 emissions in Shanxi and 0.4% of the total in China. These results are consistent with Chinese policies over the past decade effectively addressing the low-hanging fruit. Future policies will need to focus more deeply on rural areas, rapidly changing areas, and new industries. The observed atmospheric conditions in Shanxi, characterized by dry weather and high air pollutant emissions, lead to active atmospheric photochemistry, influenced by mountainous terrain and increasing CH 4 emissions from coal mining. These conditions uniquely affect atmospheric OH and CO lifetime. While these atmospheric conditions are not unique to Shanxi, they have not frequently been observed or studied in the USA, EU, and Eastern China, and are not widely discussed in the literature. Long-range transport of CO, consistent with the region's topography and upwind economic development, is observed, highlighting the importance of considering energy consumption patterns and high-resolution topography in future planning. The average CO/NO x calculated by MFIEF aligns with different land use types, with urban and industrial areas generally showing lower ratios compared to rural and natural areas. Furthermore, attribution is possible over certain industrial areas, specifically iron and steel (high and narrow ratio), power and cement (less high and less narrow ratio), and coking and boilers (not so high, but wide ratio). This finding is closely related to combustion temperature (NO x ) and efficiency (CO). Future improvements in emissions calculations involve enhancing ground-based and satellite observations, reducing retrieval uncertainties, and expanding the application of MFIEF to other source regions. This includes refining a priori emissions data and expanding the range of species analyzed. Furthermore, expanding into regions with few or no a priori emissions is possible using trained values from Shanxi, possibly adding considerable value in the Global South in areas with similar climatologically and development. The method's adaptability holds promise for use in air pollution and emergency response efforts to identify and predict extreme pollution events rapidly. Methods Satellite and wind data The TROPOMI spectrometer on board Sentinel-5 follows a sun-synchronous, low-Earth orbit with an equator overpass around 13:30 LT, allowing daily measurements globally 56,57 . The work uses all available grids of CO and HCHO columns which have at least one pass over Shanxi during 1 May 2018 to 30 April 2022. Data quality is assured filtering each pixel with a “qa_value” smaller than 0.75, “cloud radiance fraction” larger than 0.5, and that scenes covered by snow/ice, errors and similar problematic retrievals 58 . Then, overlapping column pixels are resampled using weighted polygons to a common 0.05°×0.05° grid ( http://stcorp.github.io/harp/doc/html/index.html ). Climatological maps of 2019 CO columns over East and parts of Southeast and South Asia, and HCHO over Shanxi and parts of surrounding provinces are given in Fig. 1 a and 1 b. The wind data used in this work is from the ERA-5 reanalysis product, available for download at https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5 . This work specifically used the 6:00 UTC u and v wind products (closest in terms of time to the TROPOMI overpass) at 850 hPa and 0.25°×0.25° resolution 8 . The data was subsequently linearly interpolated to the TROPOMI grid. Land use data Daily flux observations of larger industry sources and their locations are provided by CEMS ( http://www.envsc.cn ). The locations were used to categorize those grids which have industrial sources. The data underlying the land use types come from the Department of Civil Affairs of Shanxi Province and the Shanxi Provincial Platform for Common Geospatial Information Services in 2021, which was used to respectively categorize those grids which are urban and rural, based on their respective administrative level. The remaining grids are classified as natural, as shown in Fig. 1 c. Experimental Design Changes in the stock of column CO are simultaneously impacted by emissions (increases the stock), chemical loss (decreases the stock), chemical production (increases the stock), mixing into cleaner air around the edges (decreases the stock), pressure induced and advective transport (increases or decrease the stock) 59 . This work uses model free inversion estimation framework (MFIEF) to estimate CO emissions. This approach utilizes first-order approximations for each term, ensuring mass conservation, and integrates remotely sensed observations, as shown in Eq. ( 1 ). Detailed information is available in previous works 8,60,61 , where the computed daily average CO column emissions (natural and anthropogenic sources) are denoted as E CO [µg m − 2 d − 1 ]. $${E}_{CO}=\frac{{dV}_{CO}}{dt}+{{\alpha }_{2}\bullet V}_{CO}+{\alpha }_{3}\bullet {\nabla (\varvec{u}\bullet V}_{CO}){{-\alpha }_{4}\bullet V}_{HCHO}$$ 1 In order to fit the terms α 2 , α 3 , and α 4 , the equation is driven by a priori emissions, herein generated using the top-down NO x emissions computed using MFIEF 8 , scaled by the CO to NO x emission ratio from MEIC version 1.4 ( http://meicmodel.org.cn ) 22,62 . V CO represents the TROPOMI CO column loading after converted the unit into [µg m − 2 ]. \({\alpha }_{2}{\bullet V}_{CO}\) represents CO’s sink, with α 2 [d − 1 ] as the inverse CO lifetime. Since CO in clean parts of the atmosphere is many times less concentrated than in polluted areas, there is a local transfer around the edges of these plumes, which are roughly similar to the concentration \({V}_{CO}\) . This value α 2 is dependent on both the chemical loss \(\frac{1}{{\alpha }_{chem}}\) and this mixing loss into the background air \(\frac{1}{{\alpha }_{atm}}\) and is computed following Eq. ( 2 ). A balance of both effects is found to occur in the real world, in which significant forest fire plumes of sufficient concentration eventually are not identifiable using satellite after only traveling for one to two weeks 23,24,40,53 . \({{\alpha }_{4}\bullet V}_{HCHO}\) represents CO production, with α 4 [d − 1 ] as the inverse HCHO lifetime. \({\nabla (\varvec{u}\bullet V}_{CO})\) represents the daily zonal and meridional divergence of CO with a unit of [µg m − 2 d − 1 ]. α 3 [m − 1 ] represents the transport distance. $$\frac{1}{{\alpha }_{2}}=\frac{1}{{\alpha }_{chem}}+\frac{1}{{\alpha }_{atm}}$$ 2 Statistical Analysis This work employs multiple linear regression to fit α 2 , α 3 , and α 4 on a month-by-month, grid-by-grid basis using all available measurements and Eq. ( 1 ). Values of α 2 , α 3 , and α 4 are filtered based on statistics (p < 0.1); removal of outliers, defined as being more than three scaled median absolute deviations from the median, following Li et al (2023); and chemical realism ( α 2 0). Bootstrapping is applied to create a new sample representing the parent sample distribution through multiple repetitions, herein generating pseudo α 2 , α 3 , and α 4 across the central 80% of their probability distributions, filling gaps where no a priori data exists. The resulting mean is presented as the daily emission, and the standard deviation is calculated as the uncertainty of this daily emission. Sensitivity and Uncertainty Analysis Emission uncertainty results from combined uncertainties in satellite data, a priori emissions, and model development 63 . The regression uncertainties range is computed to be 32–73% (95% confidence interval) via Eq. ( 1 ), which is still lower than traditional bottom-up inventories 64 . In addition to the previously introduced uncertainties in computed emissions being a function of various uncertainties in model parameters and configuration using true observations, this work now performs a new set of sensitivity tests, in which the uncertainty in the observations is included. This is done to analyze the robustness of the various computed emissions, and ensure that they are no larger than the uncertainties in the observations themselves. This is consistent with how the most robust approaches of inverse modelling are done such as Kalman Filter 43,65 and 3D or 4D variational methods 66,67 . To ensure that the results are explainable and that the process does not take too much computational power, the tests are done using values of uncertainty that are near the upper bounds of what the community currently considers reasonable 68,69 following an approach used by 60,61 . This work specifically tests each of four input observations: TROPOMI CO column loading (uncertainty range of ± 30%), TROPOMI HCHO column loading (uncertainty range of ± 40%), a priori emissions of CO (uncertainty range of ± 30%), and the vertical level of the reanalysis wind (selected from 850 hPa to 900 hPa). Declarations Competing interests The authors declare no competing interest. Author contributions Xiaolu Li, Jason Blake Cohen, Kai Qin developed the research question and set up the whole experimental program. Xiaolu Li wrote the manuscript and performed the data analysis with input from Jason Blake Cohen, Pravash Tiwari, Shuo Wang, Liling Wu and Hailong Yang. Liling Wu and Hailong Yang contributed to downloading and processing of MEIC and other input data, and give some suggestion on the modeling. All authors discussed the results and contributed to the final manuscript. Acknowledgements The authors would like to thank the PIs of the TROPOMI, ERA-5, EDGAR, and MEIC products for making their data available. The study was supported by the National Natural Science Foundation of China (42075147), the Fundamental Research Funds for the Central Universities (2023KYJD1003), and the Shanxi Province Major Science and Technique Program (202101090301013). Data availability The satellite CO and HCHO datasets used in this study are available at https://disc.gsfc.nasa.gov/datasets . The ERA-5 reanalysis product is available at https://doi.org/10.24381/cds.bd0915c6 . The MEIC product can be accessed from https://doi.org/10.6084/m9.figshare.c.5214920.v2 . The data that support the findings of this study are openly available at the following URL/DOI: https://doi.org/10.6084/m9.figshare.24086943 . Code availability The code is available in the Figshare database ( https://doi.org/10.6084/m9.figshare.24086943 ). References Daniel, J. S. & Solomon, S. On the climate forcing of carbon monoxide. J. Geophys. Res. Atmos. 103, 13249–13260 (1998). Li, J., Reiffs, A., Parchatka, U. & Fischer, H. In situ measurements of atmospheric CO and its correlation with NO x and O 3 at a rural mountain site. Metrol. Meas. Syst.22, 25–38 (2015). Thompson, R. L. et al. Estimation of the atmospheric hydroxyl radical oxidative capacity using multiple hydrofluorocarbons (HFCs). EGUsphere 2023, 1–23 (2023). Holloway, T., Levy, H. & Kasibhatla, P. Global distribution of carbon monoxide. J. Geophys. Res. Atmos. 105, 12123–12147 (2000). Saunois, M. et al. The Global Methane Budget 2000–2017. Earth Syst. Sci. Data 12, 1561–1623 (2020). 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). Andreae, M. O. & Merlet, P. Emission of trace gases and aerosols from biomass burning. Global Biogeochem. Cy. 15, 955–966 (2001). Li, X. et al. Remotely sensed and surface measurement-derived mass-conserving inversion of daily NO x emissions and inferred combustion technologies in energy-rich northern China. Atmos. Chem. Phys. 23, 8001–8019 (2023). Wang, Y. X. Asian emissions of CO and NO x : Constraints from aircraft and Chinese station data. J. Geophys. Res. 109, D24304 (2004). Cohen, J. B. & Prinn, R. G. Development of a fast, urban chemistry metamodel for inclusion in global models. Atmos. Chem. Phys. 11, 7629–7656 (2011). Lin, C., Cohen, J. B., Wang, S. & Lan, R. Application of a combined standard deviation and mean based approach to MOPITT CO column data, and resulting improved representation of biomass burning and urban air pollution sources. Remote Sens. Environ. 241, 111720 (2020). Hassler, B. et al. Analysis of long-term observations of NO x and CO in megacities and application to constraining emissions inventories. Geophys. Res. Lett. 43, 9920–9930 (2016). Zheng, B. et al. Resolution dependence of uncertainties in gridded emission inventories: a case study in Hebei, China. Atmos. Chem. Phys. 17, 921–933 (2017). Li, J. et al. Evaluation of the WRF-CMAQ Model Performances on air quality in China with the impacts of the observation nudging on meteorology. Aerosol Air Qual. Res. 22, 220023 (2022). DeWinter, J. L., Brown, S. G., Seagram, A. F., Landsberg, K. & Eisinger, D. S. A national-scale review of air pollutant concentrations measured in the US near-road monitoring network during 2014 and 2015. Atmos. Environ. 183, 94–105 (2018). Wu, H. et al. Probabilistic Automatic Outlier Detection for Surface Air Quality Measurements from the China National Environmental Monitoring Network. Adv. Atmos. Sci. 35, 1522–1532 (2018). Zhan, D. et al. The driving factors of air quality index in China. J. Clean. Prod. 197, 1342–1351 (2018). Chen, X. et al. Emission characteristics and impact factors of air pollutants from municipal solid waste incineration in Shanghai, China. J. Environ. Manage. 310, 114732 (2022). Norhayati, I. & Rashid, M. Adaptive neuro-fuzzy prediction of carbon monoxide emission from a clinical waste incineration plant. Neural. Comput. Appl. 30, 3049–3061 (2018). Hoesly, R. M. et al. Historical (1750–2014) anthropogenic emissions of reactive gases and aerosols from the Community Emissions Data System (CEDS). Geosci. Model. Dev. 11, 369–408 (2018). Crippa, M. et al. Gridded emissions of air pollutants for the period 1970–2012 within EDGAR v4.3.2. Earth Syst. Sci. Data 10, 1987–2013 (2018). Li, M. et al. Anthropogenic emission inventories in China: A review. Natl. Sci. Rev. 4, 834–866, (2017). Lin, C. Y., Cohen, J. B., Wang, S., Lan, R. Y. & Deng, W. Z. A new perspective on the spatial, temporal, and vertical distribution of biomass burning: quantifying a significant increase in CO emissions. Environ. Res. Lett. 15, 104091 (2020). Wang, S., Cohen, J. B., Lin, C. Y. & Deng, W. Z. Constraining the relationships between aerosol height, aerosol optical depth and total column trace gas measurements using remote sensing and models. Atmos. Chem. Phys. 20, 15401–15426 (2020). Qin, K., Hu, W., He, Q., Lu, F. & Cohen, J. B. Individual coal mine methane emissions constrained by eddy-covariance measurements: Low bias and missing sources. Atmos. Chem. Phys. 24, 3009–3028 (2024). Zhu, A., Wang, Q., Liu, D. & Zhao, Y. Analysis of the characteristics of CH 4 emissions in China’s coal mining industry and research on emission reduction measures. Int. J. Environ. Res. Public Health 19, 7408 (2022). Li, C., Hammer, M. S., Zheng, B. & Cohen, R. C. Accelerated reduction of air pollutants in China, 2017–2020. Sci. Total Environ. 803, 150011 (2022). Wei, J. et al. Ground-level gaseous pollutants (NO 2 , SO 2 , and CO) in China: daily seamless mapping and spatiotemporal variations. Atmos. Chem. Phys. 23, 1511–1532 (2023). Jia, M. et al. Rapid decline of carbon monoxide emissions in the Fenwei Plain in China during the three-year action plan on defending the blue sky. J. Environ. Manage. 337, 117735 (2023). State Council of the People’s Republic of China. Action plan for continuous improvement of air quality ( https://www.gov.cn/zhengce/content/202312/content_6919000.htm) (2023). Metya, A., Dagupta, P., Halder, S., Chakraborty, S. & Tiwari, Y. K. COVID-19 lockdowns improve air quality in the South-East Asian regions, as seen by the remote sensing satellites. Aerosol Air Qual. Res. 20, 1772–1782 (2020). Jiang, Y. et al. Large contributions of emission reductions and meteorological conditions to the abatement of PM 2.5 in Beijing during the 24th Winter Olympic Games in 2022. J. Environ. Sci. 136, 172–188 (2024). Ehhalt, D. et al. Atmospheric chemistry and greenhouse gases (2001). Hu, W., Qin, K., Lu, F., Li, D. & Cohen, J. B. Merging TROPOMI and eddy covariance observations to quantify 5-years of daily CH 4 emissions over coal-mine dominated region (ACCEPT). Int. J. Coal Sci. & Technol. (2024). Su, W. et al. An improved TROPOMI tropospheric HCHO retrieval over China. Atmos. Meas. Tech. 13, 6271–6292 (2020). Valin, L. C., Russell, A. R. & Cohen, R. C. Variations of OH radical in an urban plume inferred from NO 2 column measurements. Geophys. Res. Lett. 40, 1856–1860 (2013). Ma, X. et al. OH and HO 2 radical chemistry at a suburban site during the EXPLORE-YRD campaign in 2018. Atmos. Chem. Phys. 22, 7005–7028 (2022). Lama, S. et al. Estimation of OH in urban plumes using TROPOMI-inferred NO 2 /CO. Atmos. Chem. Phys. 22, 16053–16071 (2022). Li, M. et al. Tropospheric OH and stratospheric OH and Cl concentrations determined from CH 4 , CH 3 Cl, and SF 6 measurements. npj Clim. Atmos. Sci. 1, 29 (2018). Forster, C. et al. Transport of boreal forest fire emissions from Canada to Europe. J. Geophys. Res. Atmos. 106, 22887–22906 (2001). 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). Sun, E. et al. Variation in MERRA-2 aerosol optical depth and absorption aerosol optical depth over China from 1980 to 2017. J. Atmos. Solar-Terr. Phys. 186, 8–19 (2019). Cohen, J. B. & Wang, C. Estimating global black carbon emissions using a top-down Kalman Filter approach. J. Geophys. Res. Atmos. 119, 307–323 (2014). Luo, J. et al. Regional impacts of black carbon morphologies on shortwave aerosol–radiation interactions: a comparative study between the US and China. Atmos. Chem. Phys. 22, 7647–7666 (2022). Leresche, F. et al. Photochemical Aging of Atmospheric Particulate Matter in the Aqueous Phase. Environ. Sci. Technol. 55, 13152–13163 (2021). Cheng, C.-H. & Lehmann, J. Ageing of black carbon along a temperature gradient. Chemosphere 75, 1021–1027 (2009). Tiwari, P., Cohen, J., Wang, X., Wang, S. & Qin, K. Radiative forcing bias calculation based on COSMO (core-shell Mie model optimization) and AERONET data. npj Clim. Atmos. Sci. 6, 193 (2023). Liu, Z. et al. Remotely sensed BC columns over rapidly changing Western China show significant decreases in mass and inconsistent changes in number, size, and mixing properties due to policy actions. npj Clim. Atmos. Sci. 7, 124 (2024). Bates, D. R. & Nicolet, M. The photochemistry of atmospheric water vapor. J. Geophys. Res. 55, 301–327 (1950). McDonald, B. C., Gentner, D. R., Goldstein, A. H. & Harley, R. A. Long-term trends in motor vehicle emissions in u.s. urban areas. Environ. Sci. Technol. 47, 10022–10031 (2013). Ma, X., Jia, H., Sha, T., An, J. & Tian, R. Spatial and seasonal characteristics of particulate matter and gaseous pollution in China: Implications for control policy. Environ. Pollut. 248, 421–428 (2019). Duncan, B. N., Strahan, S. E., Yoshida, Y., Steenrod, S. D. & Livesey, N. Model study of the cross-tropopause transport of biomass burning pollution. Atmos. Chem. Phys. 7, 3713–3736 (2007). Wang, S., Cohen, J. B., Deng, W., Qin, K. & Guo, J. Using a new top-down constrained emissions inventory to attribute the previously unknown source of extreme aerosol loadings observed annually in the monsoon Asia free troposphere. Earth's Future 9, e2021EF002167 (2021). Liang, Q. et al. Long-range transport of Asian pollution to the northeast Pacific: Seasonal variations and transport pathways of carbon monoxide. J. Geophys. Res. Atmos. 109 (2004). Volz, A., Ehhalt, D. H. & Derwent, R. G. Seasonal and latitudinal variation of 14 CO and the tropospheric concentration of OH radicals. J. Geophys. Res.: Oceans 86, 5163–5171 (1981). Veefkind, J. P. et al. 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 (2012). Borsdorff, T. et al. Measuring carbon monoxide with TROPOMI: First results and a comparison with ECMWF-IFS analysis data. Geophys. Res. Lett. 45, 2826–2832 (2018). Henk, E. et al. Sentinel-5 precursor/TROPOMI Level 2 Product User Manual Nitrogendioxide (https:// sentinel.esa.int/documents/247904/4682535 /Sentinel-5P-Level-2-Product-User-Manual -Nitrogen-Dioxide/ad25ea4c-3a9a-3067-0d1c-aaa56eb1746b ) (2021). Beirle, S. et al. Pinpointing nitrogen oxide emissions from space. Sci. Adv. 5, eaax9800 (2019). Qin, K. et al. Model-free daily inversion of NO x emissions using TROPOMI (MCMFE-NO x ) and its uncertainty: Declining regulated emissions and growth of new sources. Remote Sens. Environ. 295, 113720 (2023). Liu, J., Cohen, J. B., He, Q., Tiwari, P. & Qin, K. Accounting for NO x emissions from biomass burning and urbanization doubles existing inventories over South, Southeast and East Asia. Commun. Earth Environ. 5, 255 (2024). 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). Ding, J. et al. Intercomparison of NO x emission inventories over East Asia. Atmos. Chem. Phys. 17, 10125–10141 (2017). Bond, T. C. et al. Bounding the role of black carbon in the climate system: A scientific assessment. J. Geophys. Res. Atmos. 118, 5380–5552 (2013). Rigby, M. et al. Renewed growth of atmospheric methane. Geophys. Res. Lett. 35 (2008). Courtier, P. et al. The ECMWF implementation of three-dimensional variational assimilation (3D-Var). I: Formulation. Q. J. Roy. Meteor. Soc. 124, 1783–1807 (1998). Daescu, D. N. On the sensitivity equations of four-dimensional variational (4D-Var) data assimilation. Mon. Weather Rev. 136, 3050–3065 (2008). Pan, L., Gille, J. C., Edwards, D. P., Bailey, P. L. & Rodgers, C. D. Retrieval of tropospheric carbon monoxide for the MOPITT experiment. J. Geophys. Res. Atmos. 103, 32277–32290 (1998). Landgraf, J. Borsdorff, T. Langerock, B. & Keppens, A. S5P Mission Performance Centre Carbon Monoxide [L2-CO] Readme ( https://sentinel.esa.int/documents/247904/3541451/Sentinel-5P-Carbon-Monoxide-Level-2-Product-Readme-File.pdf/f8942626-ffb6- 4951-90fc-a16b6589e39e?t=1610561347131) (2023). Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformationCEE.pdf Cite Share Download PDF Status: Published Journal Publication published 08 May, 2025 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4604393","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":318956314,"identity":"2cdfcd49-08a1-4eff-8a62-5d68f93871c9","order_by":0,"name":"Jason Cohen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1ElEQVRIiWNgGAWjYBACxgYIIcfADuZLgAgDorQYMzBDtEgQ1ALTl9gA0cJAWAtz+/GHjwt32KX3N/OYbmDMsahjYG/eJoHXgp4cY+OZZ5JzZxzmMbvBuA3oMJ5jZfi1NOSwSfO2HchtgGuRyDHDr6X/+TOQlnR5uBb5NwS0zEgwA2lJMEDYwkNIyxtjY94zyYYbD7OV3UjcJiHZxpNWbIFPi2F/+sPHvDvs5OWON2+78XFbHT8/++GNN/BqaUDmJQAxGz7lICBPSMEoGAWjYBSMAgYAflJB9SZzR5wAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-9889-8175","institution":"China University of Mining and Technology","correspondingAuthor":true,"prefix":"","firstName":"Jason","middleName":"","lastName":"Cohen","suffix":""},{"id":318956315,"identity":"615062b7-cf23-4967-a02b-c47caf028ed9","order_by":1,"name":"Xiaolu Li","email":"","orcid":"https://orcid.org/0000-0001-7206-1230","institution":"China University of Mining and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiaolu","middleName":"","lastName":"Li","suffix":""},{"id":318956316,"identity":"87418c14-793c-4279-915e-7693b191b75a","order_by":2,"name":"Pravash Tiwari","email":"","orcid":"https://orcid.org/0000-0003-4770-4526","institution":"China University of Mining and Technology","correspondingAuthor":false,"prefix":"","firstName":"Pravash","middleName":"","lastName":"Tiwari","suffix":""},{"id":318956317,"identity":"84962fb7-7486-4301-8463-1eacb5b781b5","order_by":3,"name":"Liling Wu","email":"","orcid":"","institution":"China University of Mining and Technology","correspondingAuthor":false,"prefix":"","firstName":"Liling","middleName":"","lastName":"Wu","suffix":""},{"id":318956318,"identity":"896c1b6e-9246-46b4-897d-750d59c3dcb2","order_by":4,"name":"Shuo Wang","email":"","orcid":"","institution":"China University of Mining and Technology","correspondingAuthor":false,"prefix":"","firstName":"Shuo","middleName":"","lastName":"Wang","suffix":""},{"id":318956319,"identity":"b73a87ac-1b44-4cda-9095-07a254d0ffa2","order_by":5,"name":"Qin He","email":"","orcid":"https://orcid.org/0000-0002-6087-5293","institution":"China University of Mining and Technology","correspondingAuthor":false,"prefix":"","firstName":"Qin","middleName":"","lastName":"He","suffix":""},{"id":318956320,"identity":"caf6fef0-9138-476d-8f6f-145cbce4cb9b","order_by":6,"name":"Hailong Yang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Hailong","middleName":"","lastName":"Yang","suffix":""},{"id":318956321,"identity":"a08aabbc-43c3-4da1-95d1-c1f6e857d9fa","order_by":7,"name":"Kai Qin","email":"","orcid":"https://orcid.org/0000-0002-1280-6330","institution":"China University of Mining and Technology","correspondingAuthor":false,"prefix":"","firstName":"Kai","middleName":"","lastName":"Qin","suffix":""}],"badges":[],"createdAt":"2024-06-19 08:35:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4604393/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4604393/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s43247-025-02301-5","type":"published","date":"2025-05-08T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":60486368,"identity":"2e325d10-6c94-4cad-a795-a71ce64fd567","added_by":"auto","created_at":"2024-07-17 09:34:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":737782,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTROPOMI column loadings from 2019 and land use types at 0.05°×0.05°. \u003c/strong\u003e(a) CO column loadings over East, South, and Southeast Asia; (b) HCHO column loadings over Shanxi and parts of surrounding provinces; and (c) urban, natural, industrial, and rural areas over Shanxi Province.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/3bd0f7f3f0d6020a9fbe4d5a.png"},{"id":60485627,"identity":"7575b905-b674-4661-80e5-7224bff1d18f","added_by":"auto","created_at":"2024-07-17 09:26:08","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":299597,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCO calculated emissions from May 2018 to April 2022 over Shanxi at 0.05° × 0.05° scale. \u003c/strong\u003e(a) year-to-year spatial distribution of daily average emissions; (b) bootstrapping uncertainty range (10% to 90% of distribution); (c) day-to-day variation of the Shanxi-average emissions; (d) averaged weekly CO emission over three different geographic domains.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/9866b395a52265c0a776a1ae.png"},{"id":60485630,"identity":"f1fbe9f7-52c6-455f-82fc-2b3f3a7ef916","added_by":"auto","created_at":"2024-07-17 09:26:09","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":232382,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifference of emission density between EICO and MEIC or EDGAR. \u003c/strong\u003e(a) EI\u003csub\u003eCO\u003c/sub\u003e minus MEIC in 2019; (b) EI\u003csub\u003eCO\u003c/sub\u003e in 2019 minus EDGAR in 2018. In both (a) and (b) cases blue are grids where MEIC/EDGAR is larger than 9.5 µgm\u003csup\u003e-2\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e, red are grids where MEIC/EDGAR is between 4.0 µgm\u003csup\u003e-2\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e and 9.5 µgm\u003csup\u003e-2\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e, and orange are grids where MEIC/EDGAR is less than 4.0 µgm\u003csup\u003e-2\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e. (c) annual average difference between EI\u003csub\u003eCO\u003c/sub\u003e 2019 and EDGAR 2018. (d) annual average difference between EI\u003csub\u003eCO\u003c/sub\u003e and MEIC in 2019. (e) annual average difference between EI\u003csub\u003eCO\u003c/sub\u003e and MEIC only during the non-high period: February to November 2019.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/5c93f9e95fb02599f60221db.png"},{"id":60485620,"identity":"169bf62f-dddf-4629-b7da-47bb836f2686","added_by":"auto","created_at":"2024-07-17 09:26:07","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":313300,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSensitivity of CO emissions due to adjustment of input observations to account for uncertainty. \u003c/strong\u003eCalculated CO emissions when (a) TROPOMI CO plus 30% divided by the original EI\u003csub\u003eCO\u003c/sub\u003e; (b) a priori emission of CO plus 30% divided by the original EI\u003csub\u003eCO\u003c/sub\u003e; (c) TROPOMI HCHO plus 40% divided by the original EI\u003csub\u003eCO\u003c/sub\u003e; (d) wind data changed to 900hPa divided by the original EI\u003csub\u003eCO \u003c/sub\u003ewith 850hPa wind data; (e) TROPOMI CO minus 30% divided by the original EI\u003csub\u003eCO\u003c/sub\u003e; (f) a priori emission of CO minus 30% divided by the original EI\u003csub\u003eCO\u003c/sub\u003e; (g) TROPOMI HCHO minus 40% divided by the original EI\u003csub\u003eCO\u003c/sub\u003e; and (h) original CO lifetime compare to the adjusted cases (a), (b), (c), (e), (f), and (g) .\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/00ad7c367ca897e9432ca1d1.png"},{"id":60485625,"identity":"c998db7f-53b5-4261-b320-42db3d84df65","added_by":"auto","created_at":"2024-07-17 09:26:08","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":88681,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTime distribution of lifetime calculated by MFIEF.\u003c/strong\u003e (a) monthly α\u003csub\u003e2 \u003c/sub\u003e(CO lifetime), (b) monthly α\u003csub\u003e4\u003c/sub\u003e (HCHO lifetime).\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/994a16165446444dfcd4331a.png"},{"id":60485636,"identity":"fdc01ad7-6149-46ef-8123-88c3a8526031","added_by":"auto","created_at":"2024-07-17 09:26:09","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":89517,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifferent types of CO/NOx ratio. \u003c/strong\u003e(a) different land use types of annual average CO/NO\u003csub\u003ex\u003c/sub\u003e, (b) different industrial types of annual average CO/NO\u003csub\u003ex\u003c/sub\u003e, and (c) different industrial types of monthly average CO/NO\u003csub\u003ex\u003c/sub\u003e ratio.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/7e3b87c964b8c8e8b78bd52c.png"},{"id":60485623,"identity":"9e08f673-6b57-4e1b-93ee-a6ab92a6e720","added_by":"auto","created_at":"2024-07-17 09:26:07","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":801070,"visible":true,"origin":"","legend":"\u003cp\u003eDaily average CO emission and ∇(u∙VCO) during special time periods from 2018 to 2022. (a) With CO emission (top) and \u0026nbsp;∇(u∙VCO) (down) from December to January. (b) With CO emission (top) and \u0026nbsp;∇(u∙VCO) (down) in May.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/6026936880c8827b3d9d64e6.png"},{"id":82329711,"identity":"64f93701-7ea8-460a-b7ae-1acf6e79959f","added_by":"auto","created_at":"2025-05-09 07:08:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3616966,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/cf1b9fd4-4dee-44ce-b58e-4e2faf2603b1.pdf"},{"id":60486369,"identity":"5420998d-b4f1-4b15-a979-e0a69d7285f2","added_by":"auto","created_at":"2024-07-17 09:34:07","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3361051,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"SupplementaryInformationCEE.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4604393/v1/09b60ecd78db38fe3ba7a7f5.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Space-based inversion tracks and attributes Shanxi's under-estimated carbon monoxide emissions","fulltext":[{"header":"Introduction","content":"\u003cp\u003eA variety of atmospheric pollutants are emitted from energy use and industrial processes, including particulate matter, nitrogen oxides (NO\u003csub\u003ex\u003c/sub\u003e), carbon monoxide (CO), and other trace gases. Emitted species may also undergo atmospheric chemical processing and be transported by the wind, which in tandem affect the balance of atmospheric oxidation, chemical properties and the ultimate concentration of these species. CO is one such trace gas that impacts air quality, global climate forcing, and the budget and distribution of hydroxyl (OH) radical, thereby affecting the gas-phase chemistry of both methane (CH\u003csub\u003e4\u003c/sub\u003e) and ozone (O\u003csub\u003e3\u003c/sub\u003e) \u003csup\u003e1\u0026ndash;3\u003c/sup\u003e. Although the direct radiative forcing of CO is relatively small, because of its impacts on the concentrations of the greenhouse gasses CH\u003csub\u003e4\u003c/sub\u003e and O\u003csub\u003e3\u003c/sub\u003e \u003csup\u003e4\u003c/sup\u003e, as well as its conversion into carbon dioxide (CO\u003csub\u003e2\u003c/sub\u003e), emissions of CO contribute significantly to the global radiative balance \u003csup\u003e5\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCO is produced by incomplete combustion of carbon-based fuels, biomass burning, and chemical oxidation of volatile organic compounds (VOCs) \u003csup\u003e6\u003c/sup\u003e. Its co-emission with NO\u003csub\u003ex\u003c/sub\u003e depends on thermodynamic conditions, and oxygen and nitrogen availability at the time of combustion \u003csup\u003e7,8\u003c/sup\u003e. From the perspective of bottom-up emission inventories, CO and NO\u003csub\u003ex\u003c/sub\u003e emissions are calculated by applying emission factors to a set of activity data. Previous studies over China have indicated that biofuels and biomass burning contribute nearly half of CO emissions but only a tenth of NO\u003csub\u003ex\u003c/sub\u003e emissions \u003csup\u003e9\u0026ndash;11\u003c/sup\u003e. High-temperature combustion processes associated with transportation, power generation, and iron and steel production result in significant differences in CO and NO\u003csub\u003ex\u003c/sub\u003e emissions \u003csup\u003e8,9\u003c/sup\u003e. Therefore, analyzing the CO/NO\u003csub\u003ex\u003c/sub\u003e ratio variation can provide valuable insights for further investigation and attribution \u003csup\u003e2\u003c/sup\u003e. Discrepancies between CO and NO\u003csub\u003ex\u003c/sub\u003e representation in emission inventories may lead to under/overestimation from specific sources, posing challenges for current atmospheric models \u003csup\u003e12\u003c/sup\u003e. This is one possible reason why these models still struggle to accurately reproduce observed long-term changes even after more than a decade development \u003csup\u003e13,14\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eEnvironmental management of CO monitoring and supervision are limited compared with other criteria pollutants. While there are national control sites for CO concentration monitoring and ambient air quality standards that must be followed across many countries \u003csup\u003e15,16\u003c/sup\u003e, the management of CO emissions is almost blank, with most management based on those criteria pollutants that contribute more to the air quality index (AQI), due to CO\u0026rsquo;s very high ambient air quality standard and AQI cutoff \u003csup\u003e17\u003c/sup\u003e. Specifically in China, policy-based controls for CO emission are primarily implemented in waste incineration and a few other industries \u003csup\u003e18\u003c/sup\u003e, with only a few cities such as Tangshan, Handan and Linfen which all have significant numbers of iron and steel factories introducing CO controls in the iron and steel industry. Moreover, the continuous emissions monitoring systems (CEMS) in China and other counties do not provide CO data, hindering policy-based controls on a stack-by-stack basis \u003csup\u003e8,19\u003c/sup\u003e. Therefore, detailed quantification of CO emissions over a region with various sources can yield significant insights and potentially enhance environmental management efforts and synergistic reduction of air pollution and greenhouse gasses \u003csup\u003e20\u0026ndash;24\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThis study employs a top-down constrained mass-conserving inversion method to estimate the daily CO emissions over a mesoscale grid (0.05\u0026deg;\u0026times;0.05\u0026deg;) from May 2018 through April 2022, utilizing daily observations of the TROPOspheric Monitoring Instrument (TROPOMI) CO and HCHO. A spatially and temporally consistent top-down constrained NO\u003csub\u003ex\u003c/sub\u003e emission inventory, calculated using the same algorithm, serves as the a priori constraint \u003csup\u003e8\u003c/sup\u003e, enabling interrelationships between these co-emitted pollutants to be established based on the best fit properties of in-situ production, transport, and processing in tandem month-to-month, source-by-source, and grid-by-grid. The underlying CO a priori emissions used in this work are based on daily and grid-by-grid TROPOMI NO\u003csub\u003e2\u003c/sub\u003e constrained emissions and bottom-up CO to NO\u003csub\u003ex\u003c/sub\u003e ratio over Shanxi province. The geospatial area (showed in Fig.\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) is selected since it is an energy-rich region in Northern China that produces more than a quarter of China\u0026rsquo;s coal (including about half of China\u0026rsquo;s coking coal) and consumes nearly a tenth \u003csup\u003e25,26\u003c/sup\u003e. The geography is also unique with mountains and basins contributing the majority of surface area, and generally low cloud cover, leading to intense atmospheric processing and observed concentrations not encountered in previous studies. However, this unique set of conditions allows an ideal laboratory to study changes that occur during the combustion and utilization of coal, co-emitted pollutants, and their interrelationships.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Results and Discussion","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSpatial and Temporal Distribution of CO Emissions\u003c/h2\u003e \u003cp\u003eThe emeission was calculated over the domain from 34\u0026deg;N to 41\u0026deg;N and 110\u0026deg;E to 115\u0026deg;E inside the Loess Plateau of China. The yearly total CO emissions (hereafter EI\u003csub\u003eCO\u003c/sub\u003e), uncertainty range, and day-to-day variability over Shanxi is 30.4, 15.0, and 17.1 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, respectively, showed in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea-\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb. Year-to-year values for total emissions, uncertainty range, and daily variability (from May to April of the following year) are as follows: 38.1, 18.9, and 23.4 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e of 2018\u0026thinsp;~\u0026thinsp;2019, 32.7, 16.1, and 16.3 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e of 2019\u0026thinsp;~\u0026thinsp;2020, 27.4, 14.1, and 12.2 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e of 2020\u0026thinsp;~\u0026thinsp;2021, 24.9, 11.7, and 12.3 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e of 2021\u0026thinsp;~\u0026thinsp;2022. CO emission intensity decreases from 2018 to 2022, consistent with the Chinese Government\u0026rsquo;s long-term air quality management strategy \u003csup\u003e27\u0026ndash;30\u003c/sup\u003e. However, significant temporal variation is observed, the highest emissions occurring in December to January each year, coinciding with the end of the industrial cycle and the start of the Chinese New Year holiday (late January through early February). Total emissions during this period decrease notably, with values of 12.9, 8.7, 6.0, and 5.7 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for each year. A small peak is also observed around May, occurring around May with variations in its start and end points. When data from December, January, and May are removed, the emissions do not statistically decrease every year, contradicting claims that COVID-19 dominates emissions changes \u003csup\u003e31\u003c/sup\u003e. The only discernible impact from COVID-19 appears to be a slight shortening of the high-emission period in late January and February of 2020. A similar shortening of a high emissions time is observed during the 2022 Winter Olympic Games \u003csup\u003e32\u003c/sup\u003e. The observed changes are smaller than the year-to-year reduction in both cases. These observations differ from top-down NO\u003csub\u003ex\u003c/sub\u003e emissions \u003csup\u003e8\u003c/sup\u003e, as showed in Supplementary Fig.\u0026nbsp;1. However, this disparity is not inconsistent with policy objectives, as efforts primarily target reducing NO\u003csub\u003ex\u003c/sub\u003e emissions, indirectly affecting CO as an accompanying species.\u003c/p\u003e \u003cp\u003eGiven the diversity of emissions across different parts of Shanxi, a detailed look as been made over three emissions intensity regions (\u0026lt;\u0026thinsp;7 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, 7\u0026thinsp;~\u0026thinsp;9.5 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and \u0026gt;\u0026thinsp;9.5 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Areas with relatively high CO emissions are mainly concentrated in the lower Fen River valley (7\u0026thinsp;~\u0026thinsp;9.5 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), which accounts for the majority of population and industry, including coking and other industrial facilities. The highest intensity (\u0026gt;\u0026thinsp;9.5 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) occurs within parts of Linfen and Yuncheng where located a large number of iron and steel factories. Additional explanation is in the section \u0026ldquo;Impacts of Variability and Long-range Transport\u0026rdquo;.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe 2019 EI\u003csub\u003eCO\u003c/sub\u003e is compared with two widely used bottom- up emissions datasets: 2019 MEIC and 2018 EDGAR (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Those grids with higher emission in either MEIC or EDGAR (\u0026gt;\u0026thinsp;9.5 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), EI\u003csub\u003eCO\u003c/sub\u003e is notably lower, averaging 12.1\u0026thinsp;\u0026plusmn;\u0026thinsp;5.3 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, compared to MEIC and EDGAR averages of 17.7 and 64.6 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e respectively. This aligns with a gradual reduction in emissions from well-regulated sources like urban centers and large factories. Conversely, in grids with low initial emissions (\u0026lt;\u0026thinsp;4.5 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), EI\u003csub\u003eCO\u003c/sub\u003e tends to be consistently higher, averaging 3.5\u0026thinsp;\u0026plusmn;\u0026thinsp;2.8 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, while MEIC and EDGAR average much lower at 0.8 and 0.3 \u0026micro;g\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003em\u003csup\u003e\u0026minus;2\u003c/sup\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\)\u003c/span\u003e\u003c/span\u003es\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e respectively. The reason is three-fold: first, there are many low values close to zero in MEIC and EDGAR characterized by rural residential burning and wildfires; second, there is an increase in small and moderate sources as non-urban income increases; third, the bottom-up datasets underestimate CO emissions from the rapid increase in iron and steel enterprises in Yuncheng and Linfen. Assuming all additional CO will ultimately decay into CO\u003csub\u003e2\u003c/sub\u003e, the increase of EI\u003csub\u003eCO\u003c/sub\u003e to MEIC for 2019/2020 are 31.8 and 23.5 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, leading to a corresponding increase in CO\u003csub\u003e2\u003c/sub\u003e emissions of 49.9 and 37.0 Tg yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. This implies a CO\u003csub\u003e2\u003c/sub\u003e emissions increase of approximately 9.8%, 7.0% over Shanxi and 0.5%, 0.4% over China based on MEIC CO\u003csub\u003e2\u003c/sub\u003e in 2019 and 2020 \u003csup\u003e33\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eSensitivity and Robustness of Calculated CO Emissions\u003c/h2\u003e \u003cp\u003eSensitivity tests are performed to actively account for the ranges of uncertainties of four different input observations used to compute CO emissions, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea-\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg for seven cases (\u0026plusmn;\u0026thinsp;30% TROPOMI CO, \u0026plusmn;\u0026thinsp;40% TROPOMI HCHO, \u0026plusmn;\u0026thinsp;30% a priori emission of CO, and different pressure level of wind). The remaining cases can be found in Supplementary Fig.\u0026nbsp;2. First, in all cases the emissions computed in all cases are stable: there are no new peaks or troughs in the spatial results, and no significant changes in the temporal profiles. Second, the emissions computed in all cases are robust: the resulting computed emissions always have a smaller difference than the magnitude of the perturbations (the range of fractional changes in the seven cases in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e are 1.03\u0026ndash;1.26, 1.06\u0026ndash;1.27, 0.98\u0026ndash;1.33 ,0.82\u0026ndash;1.21, 0.78\u0026ndash;0.98, 0.76\u0026ndash;0.95, and 0.93\u0026ndash;1.13, respectively). Adjusting the wind pressure level from 850hPa to 900 hPa (showed in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed) leads to an increase in emissions within the high-altitude areas of northern Shanxi simultaneously with a reduction of emissions in the basin areas, with all change less than \u0026plusmn;\u0026thinsp;19%. Intriguingly, in the central and southern regions such as Linfen, Yuncheng, and Jincheng, where CO emission levels are relatively high, emissions still increase despite the altitudes being similarly low to the basin areas.\u003c/p\u003e \u003cp\u003eThese findings are consistent with the methodology, since any abnormally large or small driving factors computed due to the observational uncertainties are filtered out if physically unrealistic. Similarly, the non-linear effects of uncertainty on the spatial gradient terms are weighted by the linear production and loss terms, ensuring that fluctuations in these components counterbalance one another, bolstering confidence in the reliability and quality of the results.\u003c/p\u003e \u003cp\u003eOut of all terms analyzed, the term with the greatest difference between the plus and minus uncertainty is the uncertainty in a priori emissions. Specifically, the impact is larger when the +\u0026thinsp;30% uncertainty is used as compared to when the \u0026minus;\u0026thinsp;30% uncertainty, while all of the other terms have a somewhat balanced impact on emissions between the positive and negative uncertainty perturbations. This bias indicates that the a priori emissions is very low and does not provide as much information as a higher a priori dataset would. Overall, this finding provides further support for the results herein which indicate that the emissions are much higher after optimization.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDecay and Production of CO\u003c/h3\u003e\n\u003cp\u003eThe results of the CO\u0026rsquo;s lifetime are a function of the reaction between CO and the OH radical, forming CO\u003csub\u003e2\u003c/sub\u003e, and mixing between the high concentration plume of CO both as it is emitted and as it evolves in the atmosphere with respect to the lesser polluted air immediately surrounding it within an observed grid. Due to the extremely high level of local CH\u003csub\u003e4\u003c/sub\u003e in Shanxi province \u003csup\u003e34\u003c/sup\u003e and the local emissions of VOCs due to coal to chemicals industries, it is important to also consider the production of CO in-situ. However, due to limited validated global observations from remote sensing platforms, this work uses HCHO as a proxy for CO production, given HCHO\u0026rsquo;s reasonable retrieval properties \u003csup\u003e35\u003c/sup\u003e. The mean, 10th and 90th percentile of the CO and HCHO atmospheric in-situ net processing decay time are [2.9,1.0, 5.7] d, and [1.0, 0.2, 2.3] hours. The computed lifetime of CO is consistent with observed values of OH over other highly polluted and relatively drier areas \u003csup\u003e36\u0026ndash;39\u003c/sup\u003e, as well as downwind from major forest fires\u003csup\u003e40\u003c/sup\u003e. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e illustrates the monthly distribution of CO and HCHO lifetime in Shanxi. There are two reasons why CO's lifetime has increased from 2018 to 2022 while HCHO's lifetime has not. First there is an an overall decrease in OH concentration \u003csup\u003e41\u003c/sup\u003e consistent with the documented rise in CH\u003csub\u003e4\u003c/sub\u003e emissions due to increased coal production from high coal-bed methane sources in Shanxi over the past decades \u003csup\u003e25\u003c/sup\u003e. Second, changes in the concentrations of air in the surrounding basins due to different controls and economic growth play a non-linear role in the heavily mountainous environment in terms of subgrid mixing between the different atmospheric environments. These factors emphasize the importance of accurately constraining CO emissions and associated processes in tandem.\u003c/p\u003e \u003cp\u003eThe lifetime of CO is observed to be relatively longer in May and July, and relatively shorter in December and January, consistent with local observations. First, this region has less stable atmosphere in the warmer months, coupled with a relatively dry atmosphere year-round, leading to a higher cloud optical depth in May and July \u003csup\u003e42\u003c/sup\u003e. Secondly, there is a relative large number of absorbing aerosols emitted here due to coal consumed \u003csup\u003e43,44\u003c/sup\u003e, combined with enhanced particle aging in the higher UV seasons \u003csup\u003e45\u003c/sup\u003e and under higher temperature conditions \u003csup\u003e46\u003c/sup\u003e leading to a relatively higher column absorption of UV as secondary aerosol coats and mixes with locally emitted BC \u003csup\u003e47\u003c/sup\u003e, which in turn absorbs more UV from the column when the solar zenith angle is higher, maximizing in May \u003csup\u003e8\u003c/sup\u003e. Observations of the monthly average absorbing aerosol optical depth (AAOD) from Multi-Angle Imaging SpectroRadiometer (MISR) over Shanxi (0.0039 March 2019 to February 2021) confirm that the value is slightly higher in May (0.0063 in 2019 and 0.0071 in 2020), while lower than or similar to the background in December and January (0.0041 in 2019\u0026ndash;2020 and 0.0032 in 2020\u0026ndash;2021) \u003csup\u003e48\u003c/sup\u003e, as detailed in Supplementary Fig.\u0026nbsp;3. These factors lead to reduced UV within the lower troposphere, which in turn leads to an influence on OH production due to the joint limitations on UV and water vapor \u003csup\u003e49\u003c/sup\u003e. Third, NO\u003csub\u003ex\u003c/sub\u003e loadings are observed to be lower during the warmer months. Fourth, the rate of CO decrease is far smaller than that of NO\u003csub\u003ex\u003c/sub\u003e. This combination leads to a lower net atmospheric oxidation potential. The results are consistent with both the underlying theory and the actual observations in this region, calling into question whether modeled oxidant and CO fields over this region should be examined more carefully.\u003c/p\u003e \u003cp\u003eThe shorter lifetime of HCHO leadings to more CO production. α\u003csub\u003e4\u003c/sub\u003e is relatively stable, with only a significant change in September and October. However, the ratio between the lifetime of CO and HCHO in two specific sub-regions within Shanxi is larger (Supplementary Fig.\u0026nbsp;4). They are known to be upwind from other major sources adjacent to Shanxi and have topographic gaps. This is consistent with the major factors in these regions being influenced by emissions and subsequent long-range transport from far-away industrial and urban areas \u003csup\u003e48\u003c/sup\u003e. This effect is increased due to diurnal changes in upslope and downslope mountain winds when more polluted air is buttressed against the outside of the mountains surrounding Shanxi downwind from these external sources. The effect is further supported by the high loadings by the high loadings of absorbing aerosol co-emitted from these sources, which reduces UV at the surface as the air is transported over multiple days from these sources to Shanxi.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eAttributing Sources Using Emissions Ratios\u003c/h3\u003e\n\u003cp\u003eThe amounts of CO and NO\u003csub\u003ex\u003c/sub\u003e emitted are a complex function of energy efficiency, combustion temperature, oxygen availability, and others, with variations of up to an order of magnitude. Thus, it is of interest to investigate the relationship between both species and use this to identify and attribute different source type \u003csup\u003e2,50\u003c/sup\u003e. The ratio of CO emissions to NO\u003csub\u003ex\u003c/sub\u003e emissions, hereafter called [CO/NO\u003csub\u003ex\u003c/sub\u003e], is computed grid-by-grid and day-by-day. On average, CO/NO\u003csub\u003ex\u003c/sub\u003e vary consistently across four land use types (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). Urban and industrial generally have a lower CO/NO\u003csub\u003ex\u003c/sub\u003e than rural areas, which lower than natural areas. Due to the fact that industrial and urban areas have more high-temperature combustion sources associated with power, transportation and gas burning sectors \u003csup\u003e9\u003c/sup\u003e, leading to more NO\u003csub\u003ex\u003c/sub\u003e and lower CO. Rural areas still have mixed small industry, boilers, and biomass burning leading to less NO\u003csub\u003ex\u003c/sub\u003e and more CO. Natural areas tend to be protected or have little anthropogenic influence.\u003c/p\u003e \u003cp\u003eA detailed analysis is performed by categorizing industrial sub-types identified using CEMS \u003csup\u003e8\u003c/sup\u003e. Cement and power have the lowest median CO/NO\u003csub\u003ex\u003c/sub\u003e, boilers and coke show intermediate median CO/NO\u003csub\u003ex\u003c/sub\u003e, while iron and steel is highest. Notably, the CO/NO\u003csub\u003ex\u003c/sub\u003e ranges (between 75th and 25th percentile) for coke and boilers are wider than others, while the range for iron and steel is smallest. Iron and steel has both the highest 25th percentile and median value, with its 25th percentile value similar to the 75th percentile for cement and power and similar to the median for coke and boiler, while its 75th percentile is slightly lower than coke. Despite high combustion temperatures in iron and steel production, resulting in high NO\u003csub\u003ex\u003c/sub\u003e emissions, the CO/NO\u003csub\u003ex\u003c/sub\u003e ratio remains high due to significant CO emissions from sintering and blast furnaces, since carbonaceous material (including metallurgical coke, coal and natural gas) are used as to reduce iron oxide to iron, in addition to being directly used as fuel for combustion. Cement's CO/NO\u003csub\u003ex\u003c/sub\u003e ratio slightly exceeds power across all percentiles due to higher NO\u003csub\u003ex\u003c/sub\u003e emissions from rotary kilns. Considering the length, consistency, and temperature of combustion, the overall lesser availability of oxygen leads to more CO than power and slightly bigger CO/NO\u003csub\u003ex\u003c/sub\u003e observed.\u003c/p\u003e \u003cp\u003eCoke and boilers have similar median and 25th percentile values, the values above the median (75th and 90th percentile) and the 10th percentile are smaller for boilers. Both coke and boilers have larger 25th and 50th percentile values compared to power and cement, but lower than iron and steel. The 75th and 90th percentile values of coke are the highest. The main fuel during coke production is natural gas, wherein the combustion process is meant to cook the coal, turning it into coke. The produced CO retained in the coke oven gas instead of being emitted directly. Furthermore, some coke plants have low CO emissions due to the lack of oxygen in charring chambers, which in turn convert a significant amount of coal that is lost (not converted into coke) into black carbon instead of CO. Leakage from combustion chamber to charring chamber in coke ovens, influenced by their design, age, and efficiency, can result in extra CO generation due to the influx of additional oxygen. The combination leads to the overall wide range of CO/NO\u003csub\u003ex\u003c/sub\u003e. Boilers, operating at lower capacity and efficiency than power plants, hence contributing to lower NO\u003csub\u003ex\u003c/sub\u003e per unit of coal consumed and a greater potential for incomplete combustion and produce normal to high CO emissions, resulting in a relatively wide CO/NO\u003csub\u003ex\u003c/sub\u003e range.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eImpacts of Variability and Long-range Transport\u003c/h2\u003e \u003cp\u003eDue to its relatively longer lifetime, CO serves as an indicator of dynamical transport, while the daily average transport distance of CO is approximately 58\u0026thinsp;~\u0026thinsp;471 km (10th and 90th percentile), there is a wide range of variation, up to hundreds of km on specific days in specific parts of Shanxi. Since typical zonal and meridional transport within Shanxi occurs on the order of 1\u0026ndash;2 weeks, the calculated lifetime of CO is capable of representing long range transport when its chemical lifetime is similarly long \u003csup\u003e52\u0026ndash;54\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCO transport, as computed through its divergence, peaks during December and January and weakens in May \u003csup\u003e55\u003c/sup\u003e. A north-south corridor is observed from Linfen/Yuncheng to Taiyuan/Datong in the central basin area of Shanxi, particularly evident during December to January (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). This corridor is surrounded by relatively highly polluted regions, continuously receiving inputs from various sources. The starting point of this corridor is found in Shaanxi Province, where there are strong urban and industrial sources in the megacity of Xi'an and the surrounding area. There is an observed weakening of the transport from 2018 to 2022. However, there is also an observed increase in the local source transport from Weinan, a smaller city in Shaanxi found just west of the Shanxi boarder, into Shanxi. This transition is consistent with policies to combat pollution in megacities such as Xi\u0026rsquo;an, and to encourage more development in smaller urban areas, such as Weinan. There is similarly a strong source within Yuncheng, that transports CO further upwind into Shanxi. The combination collectively contributes to the high emission intensity in Linfen and Yuncheng, as well as transport into the observed north-south corridor. In northeastern Shanxi bordering Hebei Province, cities like Shijiazhuang, Xingtai, and Handan exhibit strong CO sources, partly due to centralized steel enterprises and residential sources. Transport through Taihang Mountain only occurs via small gaps in Yangquan, Changzhi, and Jincheng at relatively low latitudes.\u003c/p\u003e \u003cp\u003eDuring May the overall concentration of CO decreased while at the same time the long-range transport channels significantly reduced. As observed in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, the entirety of Yuncheng city and adjacent regions in Shaanxi have turned into sources, the central transport corridor has an insignificant amount of transport, and the neighboring regions in Hebei turn to sinks. Correspondingly, there is little to no transport either from outside or within Shanxi occurring in May.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eDaily CO emissions are computed using MFIEF based on remotely sensed CO and HCHO, considering first-order atmospheric processing and transport. High CO emissions are observed in densely populated and economically active areas, particularly in the lower Fen River valley, with the highest emissions concentrated around iron and steel factories in parts of Linfen and Yuncheng Cities. Under the requirements of multi-year air pollution control, CO emissions have declined significantly in December and January, when emissions were high because of the factories that increase production loads to fulfill annual schedules before the Chinese New Year break, as well as winter residential heating. CO emissions throughout other times do not have significant reduction. Differences between EI\u003csub\u003eCO\u003c/sub\u003e, EDGAR and MEIC suggest decreases in heavily industrialized areas and increases in low-emission zones. The total difference in EI\u003csub\u003eCO\u003c/sub\u003e and MEIC may result in CO\u003csub\u003e2\u003c/sub\u003e production at least 7% of the total CO\u003csub\u003e2\u003c/sub\u003e emissions in Shanxi and 0.4% of the total in China. These results are consistent with Chinese policies over the past decade effectively addressing the low-hanging fruit. Future policies will need to focus more deeply on rural areas, rapidly changing areas, and new industries.\u003c/p\u003e \u003cp\u003eThe observed atmospheric conditions in Shanxi, characterized by dry weather and high air pollutant emissions, lead to active atmospheric photochemistry, influenced by mountainous terrain and increasing CH\u003csub\u003e4\u003c/sub\u003e emissions from coal mining. These conditions uniquely affect atmospheric OH and CO lifetime. While these atmospheric conditions are not unique to Shanxi, they have not frequently been observed or studied in the USA, EU, and Eastern China, and are not widely discussed in the literature. Long-range transport of CO, consistent with the region's topography and upwind economic development, is observed, highlighting the importance of considering energy consumption patterns and high-resolution topography in future planning.\u003c/p\u003e \u003cp\u003eThe average CO/NO\u003csub\u003ex\u003c/sub\u003e calculated by MFIEF aligns with different land use types, with urban and industrial areas generally showing lower ratios compared to rural and natural areas. Furthermore, attribution is possible over certain industrial areas, specifically iron and steel (high and narrow ratio), power and cement (less high and less narrow ratio), and coking and boilers (not so high, but wide ratio). This finding is closely related to combustion temperature (NO\u003csub\u003ex\u003c/sub\u003e) and efficiency (CO).\u003c/p\u003e \u003cp\u003eFuture improvements in emissions calculations involve enhancing ground-based and satellite observations, reducing retrieval uncertainties, and expanding the application of MFIEF to other source regions. This includes refining a priori emissions data and expanding the range of species analyzed. Furthermore, expanding into regions with few or no a priori emissions is possible using trained values from Shanxi, possibly adding considerable value in the Global South in areas with similar climatologically and development. The method's adaptability holds promise for use in air pollution and emergency response efforts to identify and predict extreme pollution events rapidly.\u003c/p\u003e "},{"header":"Methods","content":"\u003ch3\u003eSatellite and wind data\u003c/h3\u003e\u003cp\u003eThe TROPOMI spectrometer on board Sentinel-5 follows a sun-synchronous, low-Earth orbit with an equator overpass around 13:30 LT, allowing daily measurements globally \u003csup\u003e56,57\u003c/sup\u003e. The work uses all available grids of CO and HCHO columns which have at least one pass over Shanxi during 1 May 2018 to 30 April 2022. Data quality is assured filtering each pixel with a “qa_value” smaller than 0.75, “cloud radiance fraction” larger than 0.5, and that scenes covered by snow/ice, errors and similar problematic retrievals \u003csup\u003e58\u003c/sup\u003e. Then, overlapping column pixels are resampled using weighted polygons to a common 0.05°×0.05° grid (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://stcorp.github.io/harp/doc/html/index.html\u003c/span\u003e\u003cspan address=\"http://stcorp.github.io/harp/doc/html/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Climatological maps of 2019 CO columns over East and parts of Southeast and South Asia, and HCHO over Shanxi and parts of surrounding provinces are given in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea and \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb.\u003c/p\u003e\u003cp\u003eThe wind data used in this work is from the ERA-5 reanalysis product, available for download at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5\u003c/span\u003e\u003cspan address=\"https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. This work specifically used the 6:00 UTC u and v wind products (closest in terms of time to the TROPOMI overpass) at 850 hPa and 0.25°×0.25° resolution \u003csup\u003e8\u003c/sup\u003e. The data was subsequently linearly interpolated to the TROPOMI grid.\u003c/p\u003e\u003ch2\u003eLand use data\u003c/h2\u003e\u003cp\u003eDaily flux observations of larger industry sources and their locations are provided by CEMS (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.envsc.cn\u003c/span\u003e\u003cspan address=\"http://www.envsc.cn\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The locations were used to categorize those grids which have industrial sources. The data underlying the land use types come from the Department of Civil Affairs of Shanxi Province and the Shanxi Provincial Platform for Common Geospatial Information Services in 2021, which was used to respectively categorize those grids which are urban and rural, based on their respective administrative level. The remaining grids are classified as natural, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec.\u003c/p\u003e\u003ch2\u003eExperimental Design\u003c/h2\u003e\u003cp\u003eChanges in the stock of column CO are simultaneously impacted by emissions (increases the stock), chemical loss (decreases the stock), chemical production (increases the stock), mixing into cleaner air around the edges (decreases the stock), pressure induced and advective transport (increases or decrease the stock) \u003csup\u003e59\u003c/sup\u003e. This work uses model free inversion estimation framework (MFIEF) to estimate CO emissions. This approach utilizes first-order approximations for each term, ensuring mass conservation, and integrates remotely sensed observations, as shown in Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Detailed information is available in previous works \u003csup\u003e8,60,61\u003c/sup\u003e, where the computed daily average CO column emissions (natural and anthropogenic sources) are denoted as \u003cem\u003eE\u003c/em\u003e\u003csub\u003e\u003cem\u003eCO\u003c/em\u003e\u003c/sub\u003e [µg m\u003csup\u003e− 2\u003c/sup\u003e d\u003csup\u003e− 1\u003c/sup\u003e].\u003c/p\u003e\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$${E}_{CO}=\\frac{{dV}_{CO}}{dt}+{{\\alpha }_{2}\\bullet V}_{CO}+{\\alpha }_{3}\\bullet {\\nabla (\\varvec{u}\\bullet V}_{CO}){{-\\alpha }_{4}\\bullet V}_{HCHO}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003cp\u003eIn order to fit the terms \u003cem\u003eα\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e, \u003cem\u003eα\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e, and \u003cem\u003eα\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e, the equation is driven by a priori emissions, herein generated using the top-down NO\u003csub\u003ex\u003c/sub\u003e emissions computed using MFIEF \u003csup\u003e8\u003c/sup\u003e, scaled by the CO to NO\u003csub\u003ex\u003c/sub\u003e emission ratio from MEIC version 1.4 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://meicmodel.org.cn\u003c/span\u003e\u003cspan address=\"http://meicmodel.org.cn\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) \u003csup\u003e22,62\u003c/sup\u003e. \u003cem\u003eV\u003c/em\u003e\u003csub\u003e\u003cem\u003eCO\u003c/em\u003e\u003c/sub\u003e represents the TROPOMI CO column loading after converted the unit into [µg m\u003csup\u003e− 2\u003c/sup\u003e]. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\alpha }_{2}{\\bullet V}_{CO}\\)\u003c/span\u003e\u003c/span\u003e represents CO’s sink, with \u003cem\u003eα\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e [d\u003csup\u003e− 1\u003c/sup\u003e] as the inverse CO lifetime. Since CO in clean parts of the atmosphere is many times less concentrated than in polluted areas, there is a local transfer around the edges of these plumes, which are roughly similar to the concentration \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({V}_{CO}\\)\u003c/span\u003e\u003c/span\u003e. This value \u003cem\u003eα\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e is dependent on both the chemical loss \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\frac{1}{{\\alpha }_{chem}}\\)\u003c/span\u003e\u003c/span\u003e and this mixing loss into the background air \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\frac{1}{{\\alpha }_{atm}}\\)\u003c/span\u003e\u003c/span\u003e and is computed following Eq.\u0026nbsp;(\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). A balance of both effects is found to occur in the real world, in which significant forest fire plumes of sufficient concentration eventually are not identifiable using satellite after only traveling for one to two weeks \u003csup\u003e23,24,40,53\u003c/sup\u003e. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({{\\alpha }_{4}\\bullet V}_{HCHO}\\)\u003c/span\u003e\u003c/span\u003e represents CO production, with \u003cem\u003eα\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e [d\u003csup\u003e− 1\u003c/sup\u003e] as the inverse HCHO lifetime. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\nabla (\\varvec{u}\\bullet V}_{CO})\\)\u003c/span\u003e\u003c/span\u003e represents the daily zonal and meridional divergence of CO with a unit of [µg m\u003csup\u003e− 2\u003c/sup\u003e d\u003csup\u003e− 1\u003c/sup\u003e]. \u003cem\u003eα\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e [m\u003csup\u003e− 1\u003c/sup\u003e] represents the transport distance.\u003c/p\u003e\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\frac{1}{{\\alpha }_{2}}=\\frac{1}{{\\alpha }_{chem}}+\\frac{1}{{\\alpha }_{atm}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eThis work employs multiple linear regression to fit \u003cem\u003eα\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e, \u003cem\u003eα\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e, and \u003cem\u003eα\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e on a month-by-month, grid-by-grid basis using all available measurements and Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Values of \u003cem\u003eα\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e, \u003cem\u003eα\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e, and \u003cem\u003eα\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e are filtered based on statistics (p \u0026lt; 0.1); removal of outliers, defined as being more than three scaled median absolute deviations from the median, following Li \u003cem\u003eet al\u003c/em\u003e (2023); and chemical realism (\u003cem\u003eα\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e \u0026lt; 0, \u003cem\u003eα\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e \u0026gt; 0). Bootstrapping is applied to create a new sample representing the parent sample distribution through multiple repetitions, herein generating pseudo \u003cem\u003eα\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e, \u003cem\u003eα\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e, and \u003cem\u003eα\u003c/em\u003e\u003csub\u003e4\u003c/sub\u003e across the central 80% of their probability distributions, filling gaps where no a priori data exists. The resulting mean is presented as the daily emission, and the standard deviation is calculated as the uncertainty of this daily emission.\u003c/p\u003e\u003ch2\u003eSensitivity and Uncertainty Analysis\u003c/h2\u003e\u003cp\u003eEmission uncertainty results from combined uncertainties in satellite data, a priori emissions, and model development \u003csup\u003e63\u003c/sup\u003e. The regression uncertainties range is computed to be 32–73% (95% confidence interval) via Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), which is still lower than traditional bottom-up inventories \u003csup\u003e64\u003c/sup\u003e. In addition to the previously introduced uncertainties in computed emissions being a function of various uncertainties in model parameters and configuration using true observations, this work now performs a new set of sensitivity tests, in which the uncertainty in the observations is included. This is done to analyze the robustness of the various computed emissions, and ensure that they are no larger than the uncertainties in the observations themselves. This is consistent with how the most robust approaches of inverse modelling are done such as Kalman Filter \u003csup\u003e43,65\u003c/sup\u003e and 3D or 4D variational methods\u003csup\u003e66,67\u003c/sup\u003e. To ensure that the results are explainable and that the process does not take too much computational power, the tests are done using values of uncertainty that are near the upper bounds of what the community currently considers reasonable\u003csup\u003e68,69\u003c/sup\u003e following an approach used by \u003csup\u003e60,61\u003c/sup\u003e. This work specifically tests each of four input observations: TROPOMI CO column loading (uncertainty range of ± 30%), TROPOMI HCHO column loading (uncertainty range of ± 40%), a priori emissions of CO (uncertainty range of ± 30%), and the vertical level of the reanalysis wind (selected from 850 hPa to 900 hPa).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interest.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eXiaolu Li, Jason Blake Cohen, Kai Qin developed the research question and set up the whole experimental program. Xiaolu Li wrote the manuscript and performed the data analysis with input from Jason Blake Cohen, Pravash Tiwari, Shuo Wang, Liling Wu and Hailong Yang. Liling Wu and Hailong Yang contributed to downloading and processing of MEIC and other input data, and give some suggestion on the modeling. All authors discussed the results and contributed to the final manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThe authors would like to thank the PIs of the TROPOMI, ERA-5, EDGAR, and MEIC products for making their data available. The study was supported by the National Natural Science Foundation of China (42075147), the Fundamental Research Funds for the Central Universities (2023KYJD1003), and the Shanxi Province Major Science and Technique Program (202101090301013).\u003c/p\u003e\u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eThe satellite CO and HCHO datasets used in this study are available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://disc.gsfc.nasa.gov/datasets\u003c/span\u003e\u003cspan address=\"https://disc.gsfc.nasa.gov/datasets\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. The ERA-5 reanalysis product is available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.24381/cds.bd0915c6\u003c/span\u003e\u003cspan address=\"10.24381/cds.bd0915c6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. The MEIC product can be accessed from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.6084/m9.figshare.c.5214920.v2\u003c/span\u003e\u003cspan address=\"10.6084/m9.figshare.c.5214920.v2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. The data that support the findings of this study are openly available at the following URL/DOI: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.6084/m9.figshare.24086943\u003c/span\u003e\u003cspan address=\"10.6084/m9.figshare.24086943\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\u003ch2\u003eCode availability\u003c/h2\u003e \u003cp\u003eThe code is available in the Figshare database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.6084/m9.figshare.24086943\u003c/span\u003e\u003cspan address=\"10.6084/m9.figshare.24086943\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eDaniel, J. S. \u0026amp; Solomon, S. On the climate forcing of carbon monoxide. J. Geophys. Res. Atmos. 103, 13249\u0026ndash;13260 (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, J., Reiffs, A., Parchatka, U. \u0026amp; Fischer, H. In situ measurements of atmospheric CO and its correlation with NO\u003csub\u003ex\u003c/sub\u003e and O\u003csub\u003e3\u003c/sub\u003e at a rural mountain site. Metrol. Meas. Syst.22, 25\u0026ndash;38 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThompson, R. L. \u003cem\u003eet al.\u003c/em\u003e Estimation of the atmospheric hydroxyl radical oxidative capacity using multiple hydrofluorocarbons (HFCs). \u003cem\u003eEGUsphere\u003c/em\u003e 2023, 1\u0026ndash;23 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolloway, T., Levy, H. \u0026amp; Kasibhatla, P. Global distribution of carbon monoxide. J. Geophys. Res. Atmos. 105, 12123\u0026ndash;12147 (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaunois, M. \u003cem\u003eet al.\u003c/em\u003e The Global Methane Budget 2000\u0026ndash;2017. Earth Syst. Sci. Data 12, 1561\u0026ndash;1623 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng, B. \u003cem\u003eet al.\u003c/em\u003e Rapid decline in carbon monoxide emissions and export from East Asia between years 2005 and 2016. Environ. Res. Lett. 13, 044007 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndreae, M. O. \u0026amp; Merlet, P. Emission of trace gases and aerosols from biomass burning. Global Biogeochem. Cy. 15, 955\u0026ndash;966 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, X. \u003cem\u003eet al.\u003c/em\u003e Remotely sensed and surface measurement-derived mass-conserving inversion of daily NO\u003csub\u003ex\u003c/sub\u003e emissions and inferred combustion technologies in energy-rich northern China. Atmos. Chem. Phys. 23, 8001\u0026ndash;8019 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, Y. X. Asian emissions of CO and NO\u003csub\u003ex\u003c/sub\u003e: Constraints from aircraft and Chinese station data. J. Geophys. Res. 109, D24304 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCohen, J. B. \u0026amp; Prinn, R. G. Development of a fast, urban chemistry metamodel for inclusion in global models. Atmos. Chem. Phys. 11, 7629\u0026ndash;7656 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin, C., Cohen, J. B., Wang, S. \u0026amp; Lan, R. Application of a combined standard deviation and mean based approach to MOPITT CO column data, and resulting improved representation of biomass burning and urban air pollution sources. Remote Sens. Environ. 241, 111720 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHassler, B. \u003cem\u003eet al.\u003c/em\u003e Analysis of long-term observations of NO\u003csub\u003ex\u003c/sub\u003e and CO in megacities and application to constraining emissions inventories. Geophys. Res. Lett. 43, 9920\u0026ndash;9930 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng, B. \u003cem\u003eet al.\u003c/em\u003e Resolution dependence of uncertainties in gridded emission inventories: a case study in Hebei, China. Atmos. Chem. Phys. 17, 921\u0026ndash;933 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, J. \u003cem\u003eet al.\u003c/em\u003e Evaluation of the WRF-CMAQ Model Performances on air quality in China with the impacts of the observation nudging on meteorology. Aerosol Air Qual. Res. 22, 220023 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeWinter, J. L., Brown, S. G., Seagram, A. F., Landsberg, K. \u0026amp; Eisinger, D. S. A national-scale review of air pollutant concentrations measured in the US near-road monitoring network during 2014 and 2015. Atmos. Environ. 183, 94\u0026ndash;105 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu, H. \u003cem\u003eet al.\u003c/em\u003e Probabilistic Automatic Outlier Detection for Surface Air Quality Measurements from the China National Environmental Monitoring Network. Adv. Atmos. Sci. 35, 1522\u0026ndash;1532 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhan, D. \u003cem\u003eet al.\u003c/em\u003e The driving factors of air quality index in China. J. Clean. Prod. 197, 1342\u0026ndash;1351 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, X. \u003cem\u003eet al.\u003c/em\u003e Emission characteristics and impact factors of air pollutants from municipal solid waste incineration in Shanghai, China. J. Environ. Manage. 310, 114732 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNorhayati, I. \u0026amp; Rashid, M. Adaptive neuro-fuzzy prediction of carbon monoxide emission from a clinical waste incineration plant. Neural. Comput. Appl. 30, 3049\u0026ndash;3061 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoesly, R. M. \u003cem\u003eet al.\u003c/em\u003e Historical (1750\u0026ndash;2014) anthropogenic emissions of reactive gases and aerosols from the Community Emissions Data System (CEDS). Geosci. Model. Dev. 11, 369\u0026ndash;408 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrippa, M. \u003cem\u003eet al.\u003c/em\u003e Gridded emissions of air pollutants for the period 1970\u0026ndash;2012 within EDGAR v4.3.2. Earth Syst. Sci. Data 10, 1987\u0026ndash;2013 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, M. \u003cem\u003eet al.\u003c/em\u003e Anthropogenic emission inventories in China: A review. Natl. Sci. Rev. 4, 834\u0026ndash;866, (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin, C. Y., Cohen, J. B., Wang, S., Lan, R. Y. \u0026amp; Deng, W. Z. A new perspective on the spatial, temporal, and vertical distribution of biomass burning: quantifying a significant increase in CO emissions. Environ. Res. Lett. 15, 104091 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, S., Cohen, J. B., Lin, C. Y. \u0026amp; Deng, W. Z. Constraining the relationships between aerosol height, aerosol optical depth and total column trace gas measurements using remote sensing and models. Atmos. Chem. Phys. 20, 15401\u0026ndash;15426 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin, K., Hu, W., He, Q., Lu, F. \u0026amp; Cohen, J. B. Individual coal mine methane emissions constrained by eddy-covariance measurements: Low bias and missing sources. Atmos. Chem. Phys. 24, 3009\u0026ndash;3028 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu, A., Wang, Q., Liu, D. \u0026amp; Zhao, Y. Analysis of the characteristics of CH\u003csub\u003e4\u003c/sub\u003e emissions in China\u0026rsquo;s coal mining industry and research on emission reduction measures. Int. J. Environ. Res. Public Health 19, 7408 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, C., Hammer, M. S., Zheng, B. \u0026amp; Cohen, R. C. Accelerated reduction of air pollutants in China, 2017\u0026ndash;2020. Sci. Total Environ. 803, 150011 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei, J. \u003cem\u003eet al.\u003c/em\u003e Ground-level gaseous pollutants (NO\u003csub\u003e2\u003c/sub\u003e, SO\u003csub\u003e2\u003c/sub\u003e, and CO) in China: daily seamless mapping and spatiotemporal variations. Atmos. Chem. Phys. 23, 1511\u0026ndash;1532 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJia, M. \u003cem\u003eet al.\u003c/em\u003e Rapid decline of carbon monoxide emissions in the Fenwei Plain in China during the three-year action plan on defending the blue sky. J. Environ. Manage. 337, 117735 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eState Council of the People\u0026rsquo;s Republic of China. Action plan for continuous improvement of air quality (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.gov.cn/zhengce/content/202312/content_6919000.htm)\u003c/span\u003e\u003cspan address=\"https://www.gov.cn/zhengce/content/202312/content_6919000.htm)\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMetya, A., Dagupta, P., Halder, S., Chakraborty, S. \u0026amp; Tiwari, Y. K. COVID-19 lockdowns improve air quality in the South-East Asian regions, as seen by the remote sensing satellites. Aerosol Air Qual. Res. 20, 1772\u0026ndash;1782 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang, Y. \u003cem\u003eet al.\u003c/em\u003e Large contributions of emission reductions and meteorological conditions to the abatement of PM\u003csub\u003e2.5\u003c/sub\u003e in Beijing during the 24th Winter Olympic Games in 2022. J. Environ. Sci. 136, 172\u0026ndash;188 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEhhalt, D. \u003cem\u003eet al.\u003c/em\u003e Atmospheric chemistry and greenhouse gases (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu, W., Qin, K., Lu, F., Li, D. \u0026amp; Cohen, J. B. Merging TROPOMI and eddy covariance observations to quantify 5-years of daily CH\u003csub\u003e4\u003c/sub\u003e emissions over coal-mine dominated region (ACCEPT). Int. J. Coal Sci. \u0026amp; Technol. (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu, W. \u003cem\u003eet al.\u003c/em\u003e An improved TROPOMI tropospheric HCHO retrieval over China. Atmos. Meas. Tech. 13, 6271\u0026ndash;6292 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValin, L. C., Russell, A. R. \u0026amp; Cohen, R. C. Variations of OH radical in an urban plume inferred from NO\u003csub\u003e2\u003c/sub\u003e column measurements. Geophys. Res. Lett. 40, 1856\u0026ndash;1860 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa, X. \u003cem\u003eet al.\u003c/em\u003e OH and HO\u003csub\u003e2\u003c/sub\u003e radical chemistry at a suburban site during the EXPLORE-YRD campaign in 2018. Atmos. Chem. Phys. 22, 7005\u0026ndash;7028 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLama, S. \u003cem\u003eet al.\u003c/em\u003e Estimation of OH in urban plumes using TROPOMI-inferred NO\u003csub\u003e2\u003c/sub\u003e/CO. Atmos. Chem. Phys. 22, 16053\u0026ndash;16071 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, M. \u003cem\u003eet al.\u003c/em\u003e Tropospheric OH and stratospheric OH and Cl concentrations determined from CH\u003csub\u003e4\u003c/sub\u003e, CH\u003csub\u003e3\u003c/sub\u003eCl, and SF\u003csub\u003e6\u003c/sub\u003e measurements. npj Clim. Atmos. Sci. 1, 29 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eForster, C. \u003cem\u003eet al.\u003c/em\u003e Transport of boreal forest fire emissions from Canada to Europe. J. Geophys. Res. Atmos. 106, 22887\u0026ndash;22906 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng, B. \u003cem\u003eet al.\u003c/em\u003e Global atmospheric carbon monoxide budget 2000\u0026ndash;2017 inferred from multi-species atmospheric inversions. Earth Syst. Sci. Data 11, 1411\u0026ndash;1436 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun, E. \u003cem\u003eet al.\u003c/em\u003e Variation in MERRA-2 aerosol optical depth and absorption aerosol optical depth over China from 1980 to 2017. J. Atmos. Solar-Terr. Phys. 186, 8\u0026ndash;19 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCohen, J. B. \u0026amp; Wang, C. Estimating global black carbon emissions using a top-down Kalman Filter approach. J. Geophys. Res. Atmos. 119, 307\u0026ndash;323 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo, J. \u003cem\u003eet al.\u003c/em\u003e Regional impacts of black carbon morphologies on shortwave aerosol\u0026ndash;radiation interactions: a comparative study between the US and China. Atmos. Chem. Phys. 22, 7647\u0026ndash;7666 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeresche, F. \u003cem\u003eet al.\u003c/em\u003e Photochemical Aging of Atmospheric Particulate Matter in the Aqueous Phase. Environ. Sci. Technol. 55, 13152\u0026ndash;13163 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng, C.-H. \u0026amp; Lehmann, J. Ageing of black carbon along a temperature gradient. Chemosphere 75, 1021\u0026ndash;1027 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTiwari, P., Cohen, J., Wang, X., Wang, S. \u0026amp; Qin, K. Radiative forcing bias calculation based on COSMO (core-shell Mie model optimization) and AERONET data. npj Clim. Atmos. Sci. 6, 193 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, Z. \u003cem\u003eet al.\u003c/em\u003e Remotely sensed BC columns over rapidly changing Western China show significant decreases in mass and inconsistent changes in number, size, and mixing properties due to policy actions. npj Clim. Atmos. Sci. 7, 124 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBates, D. R. \u0026amp; Nicolet, M. The photochemistry of atmospheric water vapor. J. Geophys. Res. 55, 301\u0026ndash;327 (1950).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcDonald, B. C., Gentner, D. R., Goldstein, A. H. \u0026amp; Harley, R. A. Long-term trends in motor vehicle emissions in u.s. urban areas. Environ. Sci. Technol. 47, 10022\u0026ndash;10031 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa, X., Jia, H., Sha, T., An, J. \u0026amp; Tian, R. Spatial and seasonal characteristics of particulate matter and gaseous pollution in China: Implications for control policy. Environ. Pollut. 248, 421\u0026ndash;428 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuncan, B. N., Strahan, S. E., Yoshida, Y., Steenrod, S. D. \u0026amp; Livesey, N. Model study of the cross-tropopause transport of biomass burning pollution. Atmos. Chem. Phys. 7, 3713\u0026ndash;3736 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, S., Cohen, J. B., Deng, W., Qin, K. \u0026amp; Guo, J. Using a new top-down constrained emissions inventory to attribute the previously unknown source of extreme aerosol loadings observed annually in the monsoon Asia free troposphere. Earth's Future 9, e2021EF002167 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang, Q. \u003cem\u003eet al.\u003c/em\u003e Long-range transport of Asian pollution to the northeast Pacific: Seasonal variations and transport pathways of carbon monoxide. J. Geophys. Res. Atmos. 109 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVolz, A., Ehhalt, D. H. \u0026amp; Derwent, R. G. Seasonal and latitudinal variation of \u003csup\u003e14\u003c/sup\u003eCO and the tropospheric concentration of OH radicals. J. Geophys. Res.: Oceans 86, 5163\u0026ndash;5171 (1981).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVeefkind, J. P. \u003cem\u003eet al.\u003c/em\u003e 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\u0026ndash;83 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorsdorff, T. \u003cem\u003eet al.\u003c/em\u003e Measuring carbon monoxide with TROPOMI: First results and a comparison with ECMWF-IFS analysis data. Geophys. Res. Lett. 45, 2826\u0026ndash;2832 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenk, E. \u003cem\u003eet al.\u003c/em\u003e Sentinel-5 precursor/TROPOMI Level 2 Product User Manual Nitrogendioxide (https://\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003esentinel.esa.int/documents/247904/4682535\u003c/span\u003e\u003cspan address=\"http://sentinel.esa.int/documents/247904/4682535\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e/Sentinel-5P-Level-2-Product-User-Manual\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e-Nitrogen-Dioxide/ad25ea4c-3a9a-3067-0d1c-aaa56eb1746b\u003c/span\u003e\u003cspan address=\"http://-Nitrogen-Dioxide/ad25ea4c-3a9a-3067-0d1c-aaa56eb1746b\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeirle, S. \u003cem\u003eet al.\u003c/em\u003e Pinpointing nitrogen oxide emissions from space. Sci. Adv. 5, eaax9800 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin, K. \u003cem\u003eet al.\u003c/em\u003e Model-free daily inversion of NO\u003csub\u003ex\u003c/sub\u003e emissions using TROPOMI (MCMFE-NO\u003csub\u003ex\u003c/sub\u003e) and its uncertainty: Declining regulated emissions and growth of new sources. Remote Sens. Environ. 295, 113720 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, J., Cohen, J. B., He, Q., Tiwari, P. \u0026amp; Qin, K. Accounting for NO\u003csub\u003ex\u003c/sub\u003e emissions from biomass burning and urbanization doubles existing inventories over South, Southeast and East Asia. Commun. Earth Environ. 5, 255 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng, B. \u003cem\u003eet al.\u003c/em\u003e Trends in China's anthropogenic emissions since 2010 as the consequence of clean air actions. Atmos. Chem. Phys. 18, 14095\u0026ndash;14111 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDing, J. \u003cem\u003eet al.\u003c/em\u003e Intercomparison of NO\u003csub\u003ex\u003c/sub\u003e emission inventories over East Asia. Atmos. Chem. Phys. 17, 10125\u0026ndash;10141 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBond, T. C. \u003cem\u003eet al.\u003c/em\u003e Bounding the role of black carbon in the climate system: A scientific assessment. J. Geophys. Res. Atmos. 118, 5380\u0026ndash;5552 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRigby, M. \u003cem\u003eet al.\u003c/em\u003e Renewed growth of atmospheric methane. Geophys. Res. Lett. 35 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCourtier, P. \u003cem\u003eet al.\u003c/em\u003e The ECMWF implementation of three-dimensional variational assimilation (3D-Var). I: Formulation. Q. J. Roy. Meteor. Soc. 124, 1783\u0026ndash;1807 (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaescu, D. N. On the sensitivity equations of four-dimensional variational (4D-Var) data assimilation. Mon. Weather Rev. 136, 3050\u0026ndash;3065 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePan, L., Gille, J. C., Edwards, D. P., Bailey, P. L. \u0026amp; Rodgers, C. D. Retrieval of tropospheric carbon monoxide for the MOPITT experiment. J. Geophys. Res. Atmos. 103, 32277\u0026ndash;32290 (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLandgraf, J. Borsdorff, T. Langerock, B. \u0026amp; Keppens, A. S5P Mission Performance Centre Carbon Monoxide [L2-CO] Readme (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://sentinel.esa.int/documents/247904/3541451/Sentinel-5P-Carbon-Monoxide-Level-2-Product-Readme-File.pdf/f8942626-ffb6-\u003c/span\u003e\u003cspan address=\"https://sentinel.esa.int/documents/247904/3541451/Sentinel-5P-Carbon-Monoxide-Level-2-Product-Readme-File.pdf/f8942626-ffb6-\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e4951-90fc-a16b6589e39e?t=1610561347131) (2023).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4604393/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4604393/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eA space-based mass-conserving framework using observed carbon monoxide (CO) and formaldehyde (HCHO) columns quantifies day-to-day and grid-to-grid CO emissions over energy-consuming Shanxi. Annualized total emissions are 8 times higher than a priori datasets, especially over low emission areas, resulting in an at least 7% increase in CO\u003csub\u003e2\u003c/sub\u003e emissions. Significant forcings include atmospheric lifetime of CO (0.3\u0026ndash;16.5 d) and HCHO (0.1-6.5h), and transport. Annual CO emissions decreased year-by-year, although this is only obvious when considering the two to three highest months. The ratio of top-down CO to NO\u003csub\u003ex\u003c/sub\u003e emissions show source attribution is possible over rural, urban, and five industrial areas (including power, iron/steel, and coke). Cross-border transport of CO is important in the peak emission months, including evolving sources from central Shaanxi and western Hebei. The major reason for the significant increase CO emissions is the fractional increase in non-high emitting area\u0026rsquo;s energy consumption, resulting in a spatial mis-alignment.\u003c/p\u003e","manuscriptTitle":"Space-based inversion tracks and attributes Shanxi's under-estimated carbon monoxide emissions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-17 09:26:02","doi":"10.21203/rs.3.rs-4604393/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-earth-and-environment","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsenv","sideBox":"Learn more about [Communications Earth and Environment](https://www.nature.com/commsenv/)","snPcode":"","submissionUrl":"","title":"Communications Earth \u0026 Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"75ef1eae-f50f-42ad-9933-80871d735847","owner":[],"postedDate":"July 17th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":33728037,"name":"Earth and environmental sciences/Environmental sciences/Environmental chemistry/Atmospheric chemistry"},{"id":33728038,"name":"Earth and environmental sciences/Climate sciences/Atmospheric science/Atmospheric chemistry"},{"id":33728039,"name":"Earth and environmental sciences/Climate sciences/Climate change/Attribution"}],"tags":[],"updatedAt":"2025-05-09T07:08:13+00:00","versionOfRecord":{"articleIdentity":"rs-4604393","link":"https://doi.org/10.1038/s43247-025-02301-5","journal":{"identity":"communications-earth-and-environment","isVorOnly":false,"title":"Communications Earth \u0026 Environment"},"publishedOn":"2025-05-08 04:00:00","publishedOnDateReadable":"May 8th, 2025"},"versionCreatedAt":"2024-07-17 09:26:02","video":"","vorDoi":"10.1038/s43247-025-02301-5","vorDoiUrl":"https://doi.org/10.1038/s43247-025-02301-5","workflowStages":[]},"version":"v1","identity":"rs-4604393","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4604393","identity":"rs-4604393","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}