Those figures could not always be attributed causally to restrictions because meteorology and background pollution also influenced the observations. The review synthesized work across six domains: detecting abrupt emission changes, mapping pollution across space and time, developing open geospatial workflows, estimating conditions at the surface, validating products and supporting policy-oriented applications. Open workflows based on the Copernicus ecosystem, Google Earth Engine, Python and geographic information system software enabled scalable processing and mapping of satellite observations.

Machine-learning studies also extended column observations toward near-surface estimates of nitrogen dioxide and particulate matter. These outputs are model estimates rather than direct measurements, and the reviewed evidence indicates that their interpretation depends on the data and conditions represented in each study. Validation against multi-axis differential optical absorption spectroscopy instruments, Pandora instruments, airborne measurements and regulatory monitoring networks revealed study-specific biases and constraints of representativeness.

The reviewers did not apply a formal risk-of-bias tool because the eligible corpus combined heterogeneous observational and methodological designs. Instead, they assessed reporting completeness descriptively. The review followed PRISMA 2020 principles and identified records primarily through Scopus and the Web of Science Core Collection, with ScienceDirect, PubMed and SpringerLink used for complementary discovery or full-text access; Crossref and publishers’ official webpages were used only for bibliographic and DOI verification.

Across the reviewed literature, Sentinel-5P/TROPOMI emerged as a flexible decision-support system rather than a substitute for ground monitoring. Its applications included constructing spatiotemporal datasets, assessing changes in emissions and contributing evidence to policy questions, but satellite observations, model-derived surface estimates and ground measurements addressed different aspects of air quality. The findings therefore favor interpreting TROPOMI together with atmospheric information, validation data, estimation models and local measurements.

Subject of Research: Systematic review of Sentinel-5P/TROPOMI products, geospatial workflows, validation and policy applications for air-pollution monitoring

Article Title: Sentinel-5P/TROPOMI for air-pollution monitoring: a systematic review of satellite products, geospatial workflows, and policy-oriented applications

Article References: Canales-Manchuria, G., Canales-Gutiérrez, A., Miranda-Mamani, J., Mendoza, A. C., & Mullisaca-Torres, F. (2026). Sentinel-5P/TROPOMI for air-pollution monitoring: a systematic review of satellite products, geospatial workflows, and policy-oriented applications. Air Quality, Atmosphere & Health, 19(11), Article 243. https://doi.org/10.1007/s11869-026-02130-1

Image Credits: AI Generated

DOI: 10.1007/s11869-026-02130-1

Keywords: Sentinel-5P, TROPOMI, air pollution, nitrogen dioxide, satellite monitoring, geospatial workflows, product validation, machine learning, policy applications, environmental monitoring, air-pollution, monitoring

The stakes of this question are enormous. Methane is a far more potent greenhouse gas than carbon dioxide over short timescales, trapping nearly 30 times more heat per molecule while persisting in the atmosphere for a much shorter period. That combination of high potency and brief atmospheric lifetime makes methane reduction one of the fastest available levers for slowing near-term warming, and it explains why the gas has become a central target for emissions monitoring programs worldwide. Scot Miller, an associate professor of environmental health and engineering at Johns Hopkins University and senior author of the study, emphasized that the flat trend carries both reassurance and disappointment. The good news, he noted, is that emissions have not markedly increased despite the enormous growth in natural gas production that followed the hydraulic fracturing boom. The bad news is that oil and gas operations still account for a large share of U.S. methane emissions, and the research found no detectable evidence that recent federal legislation had an impact on the total.

The regulatory backdrop against which this study unfolds is remarkable. In 2021, the Environmental Protection Agency restarted the process of developing rules to address methane emissions under the Clean Air Act, reviving an effort that had been stalled. Congress then enacted the Inflation Reduction Act of 2022, which went further than any previous federal measure by establishing fees on methane emissions from oil and gas facilities and allocating more than $1 billion in technical assistance to help the industry detect and repair leaks. Although some of these provisions proved short-lived in practice, the researchers describe the package as an unprecedented push toward reducing one of the major contributors to climate change. The central question motivating the study was whether that push produced measurable results in the atmosphere, where it ultimately matters.

Answering that question required a methodological approach capable of seeing past the self-reported inventories that have historically dominated emissions accounting. The research team turned to space. They combined observations from Europe’s TROPOMI instrument and the GOSAT satellite, both of which measure the concentration of methane in the air at countless locations across the globe, with sophisticated atmospheric modeling. Because methane released at the surface does not simply hover above its source but is carried and dispersed by winds and other meteorological forces, the researchers developed an algorithm that traced the gas observed from orbit back to its origins on the ground. This inverse modeling framework allowed them to estimate where methane emissions from the oil and gas sector occurred across the United States and how large those emissions were, year by year, for the period bookending the legislative activity from 2019 through 2024.

The numbers that emerged paint a picture of remarkable stability. Total methane emissions from oil and gas activities across the contiguous United States remained within a narrow band, between 9.3 and 10.4 teragrams per year, throughout the study period. Emissions declined by 0.58 teragrams between 2019 and 2023, a modest dip that was partially offset by a slight rise from 2023 to 2024. Set against the backdrop of rising production, that stability is itself a finding. Ao Chen, a PhD student at Johns Hopkins and the study’s first author, stated plainly that the team does not see a detectable decrease in emissions after the Inflation Reduction Act. She was careful, however, to acknowledge the limits of such an assessment: there is no way to test the counterfactual, and it remains possible that without the legislation, methane emissions would have climbed much higher.

One metric did show meaningful movement. The intensity of emissions, defined as the amount of methane released relative to the volume of oil and gas produced, decreased by approximately 11 percent over the study period. Yet the researchers interpret this decline cautiously, describing it as the continuation of a decades-long trend rather than a sudden improvement attributable to policy. Emissions intensity has been gradually falling for years as the industry has matured, and the new data suggest the recent period simply extends that trajectory. In other words, the sector is becoming somewhat more efficient at delivering energy without leaking methane, but because production keeps growing, the absolute quantity of methane entering the atmosphere has not budged.

The gradual improvement in emissions intensity may reflect a fundamental economic shift in how the industry treats natural gas. Methane, the primary component of natural gas, was once regarded largely as a nuisance byproduct of oil extraction, something to be disposed of as cheaply as possible. When there is no pipeline to carry gas from a wellhead to market, drilling companies have little financial incentive to prevent leaks and often simply flare the gas to get rid of it. But when pipeline infrastructure became available in regions such as West Texas, that same gas moved to market, where consumers burned it for heating, cooking, powering appliances, and generating electricity. Evidence from other studies suggests that leaks decline when gas becomes a valuable commodity, because every cubic foot that escapes is revenue lost. Companies have also been updating and improving their infrastructure, replacing aging equipment and adopting better operational protocols. As Miller put it, better industry protocols and less flaring have likely helped gradually curb emissions intensity even as gas and oil production have increased.

Beneath the national averages, the study uncovered striking regional variation. Texas, home to the largest oil-producing basin in the country, recorded a 14 percent decline in oil and gas methane emissions over the same period, a drop that stands out given the state’s dominant role in American energy production. By contrast, the researchers found that California and Colorado are likely falling short of their own emissions reduction goals, suggesting that state-level commitments have not yet translated into atmospheric reality in those jurisdictions. This geographic heterogeneity underscores a key advantage of the satellite-based approach: rather than relying on bottom-up inventories that may lag or misrepresent actual conditions, the method captures what is genuinely happening on the ground, basin by basin, and can reveal where progress is real and where it remains aspirational.

The findings arrive at a moment of uncertainty for methane policy in the United States. The fee mechanism established under the Inflation Reduction Act and the EPA’s Clean Air Act rules represented the most aggressive federal attempt yet to price and regulate methane leakage, yet the atmosphere, at least through 2024, shows no clear signature of their effect. Chen warned that without long-term regulatory durability, the industry may lack the incentive to further reduce methane emissions, and she argued that while curbing emissions intensity is a step in the right direction, a reduction in total methane emissions is what is truly needed to have an impact on the climate. The study is part of a broader scientific effort to determine whether federal greenhouse gas legislation and mandates have meaningfully decreased methane emissions, and its verdict so far is sobering: stability in the face of growth is preferable to the increases many scientists feared, but holding the line is not the same as bending the curve downward.

For the growing community of researchers and policymakers focused on methane, the study demonstrates both the power and the promise of atmospheric monitoring from space. By pairing satellite observations with models that trace pollutants back to their sources, scientists can now audit national emissions claims with a degree of independence that was impossible a decade ago. As more satellites come online and inversion techniques grow more refined, the ability to verify whether policies, technologies, and market forces are genuinely reducing methane pollution will only sharpen. For now, the U.S. oil and gas sector presents a paradox worth watching closely: an industry producing more energy than ever, leaking no more methane than before, and yet still releasing roughly ten teragrams of one of the most powerful greenhouse gases into the atmosphere every single year, with the total showing no sign yet of the sustained decline that climate goals demand.

Subject of Research: Satellite-based assessment of U.S. oil and gas methane emissions and the impact of federal climate policy from 2019 to 2024

Article Title: U.S. methane emissions flatline, despite increases in oil and gas production

Article References: U.S. methane emissions flatline, despite increases in oil and gas production. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: methane emissions, oil and gas, greenhouse gases, satellite monitoring, Inflation Reduction Act, EPA regulations, hydraulic fracturing, emissions intensity, TROPOMI, GOSAT, climate policy, Communications Earth & Environment

The research, led by Ao Chen and Scot Miller of Johns Hopkins University together with colleagues at several institutions, used an atmospheric inverse model constrained by satellite observations from the TROPOMI and GOSAT instruments to estimate methane emissions at a daily temporal resolution and a spatial resolution of 0.5 degrees latitude by 0.625 degrees longitude. Inverse modeling works backward from measured concentrations of methane in the air: the model simulates how emissions are transported through the atmosphere and then adjusts emission estimates until the simulated concentrations match what the satellites actually observed. The team blended the two satellite data products, using a machine-learning correction that adjusts TROPOMI retrievals to match the more accurate GOSAT measurements, and ran eight independent modeling scenarios to capture uncertainty arising from different underlying inventories of anthropogenic and wetland emissions.

The headline numbers are striking. Oil and gas methane emissions across the contiguous United States decreased by 0.58 teragrams, with an uncertainty range of 0.30 to 0.96 teragrams, from 2019 to 2023. But then, in the very years following the new federal policy, emissions increased by 0.21 teragrams between 2023 and 2024, with an uncertainty range spanning from a slight decrease of 0.06 teragrams to an increase of 0.53 teragrams. That uptick stands in sharp contrast to what the Environmental Protection Agency had projected: a 0.38 teragram reduction in oil and gas methane emissions from 2023 to 2024 from the combined influence of the Waste Emissions Charge, expected to contribute 0.15 teragrams, and the new methane rules, expected to contribute 0.23 teragrams.

The disconnect between projection and reality, the authors argue, comes down to implementation. The Inflation Reduction Act, enacted in August 2022, represented the first time Congress established a direct statutory framework and fee specifically targeting methane emissions. Its Waste Emissions Charge was designed to create a financial penalty for large emitters, and the law also created the Methane Emissions Reduction Program, allocating 1.36 billion dollars in technical and financial support for the oil and gas sector. In parallel, the EPA finalized rules in 2024 under the Clean Air Act covering both new and existing oil and gas sources. But most implementation details of the charge — conditional exemptions, monitoring requirements, and the EPA’s final regulations — remained unsettled during the study period, and subsequent political developments delayed the measures entirely, keeping the fee effectively inactive through 2026. Congress nullified the EPA’s Waste Emissions Charge rule in early 2025, and the rules for existing facilities did not come into effect until 2024, meaning states were not required to submit implementation plans until 2026.

Industry behavior compounded the regulatory limbo. While a subset of large, forward-looking operators proactively initiated enhanced monitoring and assessment protocols when the law passed, many others maintained a cautious wait-and-see posture as the technicalities of the fee and its alignment with EPA rules evolved. Even some proactive firms implemented strategic pauses in late 2023 to assess how the charge would interact with the new regulations. The authors are careful to note a caveat: they cannot rule out a counterfactual scenario in which emissions would have risen more than 0.21 teragrams in the absence of the legislation and other regulatory activity. But the data show no detectable signature of the 0.38 teragram annual reduction the EPA had anticipated.

Yet the longer arc of the data tells a genuinely encouraging story. Methane emissions intensity — total oil and gas methane emissions normalized by the combined energy equivalent of oil and natural gas production — declined by approximately 11.1 percent, with an uncertainty range of 9.3 to 13.6 percent, from 2019 to 2024, falling from 10.1 to 9.0 in units of ten-to-the-minus-four teragrams per million barrels of oil equivalent. Most of that decline occurred between 2020 and 2022, averaging 2.2 percent per year, and it happened alongside an 8.5 percent increase in production and a 3.5 percent decline in absolute emissions. The study builds on earlier work that found a 32 percent reduction in emissions intensity between 2010 and 2019, suggesting a sustained, multi-year downward trend that predates the recent federal push.

The researchers propose three explanations for why emissions stayed relatively flat even as production grew. First, earlier federal regulations — the New Source Performance Standards under Subpart OOOOa — instituted systematic leak detection and repair requirements for any facility constructed, modified, or reconstructed after September 18, 2015. These newer facilities are often built with cleaner equipment such as air-driven pneumatics and controlled tanks, yet have very high production volumes, contributing to lower overall intensity. Second, methane measurement technologies have improved dramatically over the last decade, making it easier to detect large emission events, which matter enormously because the biggest leaks account for an outsized share of total emissions. Because these advancements accumulated incrementally, they produced a gradual improvement rather than a precipitous drop in any single year. Third, rapid growth in US liquefied natural gas exports since 2019 has driven a steady increase in the outflow of Permian Basin natural gas to global markets, and the corresponding expansion of regional pipeline capacity has significantly reduced flaring and direct venting of associated gas.

The satellite-based estimates also expose a persistent gap with official accounting. Total contiguous US oil and gas methane emissions from 2019 to 2022 remained, on average, 17 percent higher than the EPA’s Greenhouse Gas Inventory, consistent with a decade of independent top-down assessments that have ranged from 12 to 100 percent above the official figures. The study concludes that the inventory captures the right directional trend in intensity but likely underestimates the absolute magnitude of emissions — capturing, as the authors put it, the right trend for the wrong reason.

Perhaps the most provocative finding concerns the states. In the three highest-emitting states with statutory greenhouse gas reduction targets — New Mexico, California, and Colorado — official oil and gas inventories aligned closely with the satellite-derived estimates, with a group average ratio of 1.0. New Mexico showed near-perfect agreement, while Colorado and California remained closely aligned. In the three highest-emitting states without such targets — Texas, Louisiana, and Oklahoma — inventories were substantially lower than the inverse model estimates, with a group average ratio of 3.7, reaching 5.8 in Oklahoma. States with targets tend to fund methane measurement campaigns and use custom emission factors, whereas states like Texas and Oklahoma often extrapolate emission factors from large, well-regulated facilities to smaller operators, even though low-producing wells have disproportionately high methane intensity.

But accurate accounting did not translate into faster reductions. The inverse model ensemble revealed no substantial difference in oil and gas methane emission trends between states with and without greenhouse gas targets during 2019 to 2024, with overlapping reduction rates. Texas — the largest emitting state, with no target — showed a 14 percent decline in oil and gas methane emissions, from 3.50 teragrams in 2019 to 3.02 teragrams in 2024, likely aided by new pipelines such as the Gulf Coast Express, Whistler Express, and Matterhorn Express that move Permian gas to market instead of flaring it, and by a rising gas-to-oil ratio that makes captured gas more valuable. Meanwhile California, despite its aggressive statutory mandate of a 40 percent reduction by 2030, would need to cut an additional 0.51 teragrams — dwarfing the 0.12 teragram reduction achieved over the preceding decade. The lesson, the authors suggest, is that emissions reduction is a complex process: infrastructure, technology, and economics may matter as much as the laws written on paper, and simply dividing states into those with targets and those without is insufficient for predicting who will actually cut methane fastest.

Subject of Research: Satellite-based estimation of US oil and gas methane emission trends and the impact of federal methane policy from 2019 to 2024

Article Title: U.S. oil and gas methane emission intensity declined 2019–2024 with little indication of federal policy impact

Article References: Chen, A., Daniels, W. S., Huang, Z., Ravikumar, A. P., Jordaan, S. M., Zhang, Z., Feng, L., Liu, H., Murray, L. T., Yu, X., Gaeta, D. C., Morgan, K. L., Yuan, J., & Miller, S. M. (2026). U.S. oil and gas methane emission intensity declined 2019–2024 with little indication of federal policy impact. Communications Earth & Environment, 7(1), Article 813. https://doi.org/10.1038/s43247-026-04039-0

Image Credits: AI Generated

DOI: 10.1038/s43247-026-04039-0

Keywords: methane emissions, oil and gas, Inflation Reduction Act, Waste Emissions Charge, satellite observations, inverse modeling, TROPOMI, GOSAT, emissions intensity, EPA regulations, Permian Basin, state climate policy

A new study published in Atmospheric Chemistry and Physics by Francisco J. Pérez-Invernón of the Instituto de Astrofísica de Andalucía in Granada and an international team of collaborators has now tackled that problem with an unusually powerful combination of instruments. The researchers focused on the Ebro Valley in northeastern Spain, a region of notably high lightning activity where airflows shaped by the Mediterranean Sea, the Atlantic Ocean, the Pyrenees and the Iberian thermal low frequently converge to spark thunderstorms. By pairing optical lightning observations from the Meteosat 12 Lightning Imager, the first lightning detector of Europe’s Meteosat Third Generation fleet in geostationary orbit, with ground-based radio measurements from a 24-station Lightning Mapping Array, the team was able to characterize individual flashes in unprecedented detail and connect their properties to the amount of NOx they generate.

The technical challenge is formidable. Satellites such as the TROPOspheric Monitoring Instrument, TROPOMI, flying on the Sentinel-5 Precursor spacecraft, can measure nitrogen dioxide columns from orbit with a horizontal resolution of about 3.5 by 5.5 kilometers at nadir. But the NOx they see over a storm is a mixture: some of it was freshly injected by lightning, some was lofted from the boundary layer, some arrived from distant sources, and some is stratospheric. Disentangling the lightning contribution requires knowing both the vertical distribution of the lightning-produced gases and the background against which they appear. The team addressed this by running high-resolution simulations with the WRF-Chem atmospheric model, nudged toward the actual lightning flashes observed by Meteosat 12, so that the model reproduced where and when the storms deposited their chemical payload. Comparing paired simulations with and without fresh lightning emissions allowed the researchers to calculate the fraction of the measured column attributable to lightning in each pixel, as well as the air mass factors needed to convert measured slant columns into vertical column densities.

The lightning observations themselves came from two complementary systems. The Lightning Imager aboard Meteosat 12, launched in December 2022 and declared operational in late 2024, stares continuously at the full Earth disk from geostationary orbit, detecting the 777-nanometer optical emission of atomic oxygen in lightning discharges with a CMOS detector sampling at one kilohertz. It records for each flash a suite of properties: the illuminated footprint, the total radiance, the duration, and the temporal evolution of the optical signal. On the ground, the recently upgraded eXtended Catalonia Lightning Mapping Array, XCALMA, tracks the very high frequency radio impulses emitted by stepped leaders as they carve lightning channels through the cloud, reconstructing the three-dimensional geometry of each flash without any distortion from intervening cloud layers. From XCALMA data the team extracted total flash channel length, flash volume and flash duration, quantities that no space-based optical sensor can measure directly.

Between July 2024 and May 2025, the researchers identified eighteen thunderstorm days that satisfied strict quality criteria, including at least 650 flashes detected by the Lightning Imager within the well-covered region of the mapping array and a mean flash age of less than 3.5 hours before the TROPOMI overpass, which occurs at roughly 13:30 local solar time. For each storm they computed the lightning NOx production efficiency, defined as the moles of NOx produced per flash, using a retrieval that accounts for the detection efficiency of each lightning system and the exponential decay of NOx during the hours between flash occurrence and satellite observation. The results were striking in their internal consistency: using Meteosat 12 flashes with an assumed five-hour NOx lifetime yielded 37 plus or minus 25 moles of NOx per flash, while using the ground-based mapping array yielded 31 plus or minus 24 moles per flash, with slightly higher values of 41 and 38 moles when a three-hour lifetime was assumed.

The heart of the study lies in how these production efficiencies relate to the physical characteristics of the flashes themselves. Across the eighteen storms, the team found consistent positive correlations between lightning-produced NOx per flash and nearly every flash property they examined. Longer flash channels produced more NOx, as expected from laboratory and modeling work showing that production scales with the volume of air heated by the discharge. Larger flash volumes, longer flash durations, brighter optical radiances and bigger illuminated footprints all correlated positively with per-flash production. The strongest linear relationships emerged when the ground-based channel length, volume and duration measurements were used, with the mapping array metrics proving to be better proxies for NOx production than the satellite optical properties, likely because cloud cover attenuates and distorts the light that reaches geostationary orbit.

One of the most intriguing findings concerns continuing currents, the slow transfer of charge that can persist for tens to hundreds of milliseconds after a cloud-to-ground stroke. Earlier work on rocket-triggered lightning suggested that these slow discharge processes account for roughly eighty percent of the total NOx produced per flash, but confirming this in natural storms has been nearly impossible because continuing currents hide below cloud and are poorly detected by optical sensors from space. The team devised new variables from the Lightning Imager data, counting the number of time-contiguous optical groups per flash and the fraction of flashes showing at least five milliseconds of uninterrupted optical signal. These continuity measures correlated positively, though weakly, with NOx production, and less strongly than total flash duration measured on the ground. The gap between the two correlations hints that significant lightning activity below the cloud, invisible to the satellite, contributes to the chemical budget, and that radio-frequency mapping remains essential for a complete picture.

The study also clarified how the detection efficiency of the new geostationary instrument behaves. The mapping array detected between 1.3 and 6.5 times more flashes than Meteosat 12 over the same storms, and the ratio of the two counts grew for more active thunderstorms, confirming that the satellite’s detection efficiency drops as flash rates rise. The satellite performed relatively better for brighter flashes, particularly cloud-to-ground flashes, which the analysis showed tend to be longer, larger and more luminous than intracloud flashes. Flashes initiating at lower altitudes, where cloud-to-ground activity predominates, were also more extensive and more energetic. These calibration insights matter beyond Spain: as the global constellation of geostationary lightning mappers expands, understanding their biases is essential for anyone using their data to estimate lightning’s chemical or climatic influence.

The broader implications reach into climate modeling and air quality forecasting. Lightning-produced NOx is roughly six times more efficient at driving ozone production than anthropogenic NOx, generating about a hundred ozone molecules per NOx molecule emitted, and it influences the tropospheric budgets of methane and carbon monoxide. Because climate change is expected to alter thunderstorm frequency and intensity, models need accurate, flash-property-aware parameterizations of lightning emissions. This study moves the field toward that goal by demonstrating that flash geometry, duration and brightness, measurable from space, carry real information about per-flash chemistry. The team estimates that their refinements, including lightning-nudged simulations and the upgraded mapping array, reduce the total uncertainty of satellite-based production estimates from the previously reported range of twenty to fifty percent down to roughly twelve to forty percent. With the upcoming Sentinel-4 spectrometer and the Meteosat Third Generation Infrared Sounder set to provide coincident air composition and humidity profiles over Europe, the Ebro Valley is poised to remain a natural laboratory where the chemistry of thunderstorms is finally being written down, flash by flash.

Subject of Research: Lightning-produced nitrogen oxide emissions measured with satellite and ground-based lightning observations over the Ebro Valley, Spain

Article Title: Characterizing lightning NOx production in the Ebro Valley with meteosat 12 LI and LMA observations

Article References: Pérez-Invernón, F. J., Gordillo-Vázquez, F. J., van der Velde, O., Montanyà, J., López Trujillo, J. A., Pineda, N., Loyola, D., Seo, S., Knote, C., Fierro, A. O., & Huntrieser, H. (2026). Characterizing lightning NO x production in the Ebro Valley with meteosat 12 LI and LMA observations. Atmospheric Chemistry and Physics, 26(19), 14277-14301. https://doi.org/10.5194/acp-26-14277-2026

Image Credits: AI Generated

DOI: 10.5194/acp-26-14277-2026

Keywords: lightning, nitrogen oxides, NOx, Meteosat 12, Lightning Imager, Lightning Mapping Array, TROPOMI, WRF-Chem, thunderstorms, atmospheric chemistry, ozone, Ebro Valley

The central trick is a well-known chemical companionship. When fossil fuels burn, they release carbon dioxide and nitrogen oxides, collectively known as NOx, in proportions that depend on what is being burned and how. Carbon dioxide is notoriously difficult to measure from orbit because it mixes into a background already saturated with natural carbon-cycle fluxes; fresh anthropogenic emissions represent at most a few percent of the total column of gas above any location. Nitrogen oxides, by contrast, are short-lived, surviving only hours to a day in the atmosphere, so a satellite snapshot of NO2 reveals essentially fresh combustion activity. The researchers exploited this by measuring NO2 from the TROPOspheric Monitoring Instrument, or TROPOMI, aboard the Copernicus Sentinel-5 Precursor satellite, and then translating those observations into CO2 estimates using sector-specific NOx to CO2 emission ratios.

The technical machinery behind the study is an ensemble data assimilation system built around the GEOS-Chem atmospheric transport model. The team assimilated a full year of TROPOMI NO2 observations from 2021 into an Ensemble Kalman Filter, a statistical framework that uses an ensemble of 100 perturbed model states to represent and propagate uncertainty in emissions. Each day, the filter compared simulated NO2 columns, sampled at the time and location of satellite overpasses and convolved with scene-dependent averaging kernels, against the actual retrievals. It then adjusted emission scaling factors to reduce the mismatch, updating the model state before the next day’s simulation began. Because NOx is so short-lived, a one-day assimilation window proved sufficient, since observations are largely insensitive to emissions from previous days.

A key innovation made this computationally feasible. Running full atmospheric chemistry for 100 ensemble members, day after day, would be prohibitively expensive. Instead, the team used a lightweight offline treatment of NOx chemistry, developed in earlier work, in which the chemical rates of change are estimated by scaling a baseline rate according to the relative change in local NOx concentration. This parameterisation reproduces the full-chemistry results with remarkable fidelity, and the team extended its validation to the larger perturbations encountered during the inversion, finding correlation coefficients above 0.94 for chemical loss rates and near-perfect stability in the NO2 to NOx partitioning ratio. The result is a pathway that retains critical chemical feedbacks while enabling large-ensemble inversions that would otherwise be out of reach.

The results show that assimilating TROPOMI data systematically sharpens the picture of European emissions. The mean uncertainty in total European fossil fuel CO2 emissions fell from 5.6 percent in the prior estimate to 3.3 percent after assimilation, a reduction of roughly 40 percent. Model agreement with the satellite observations also improved, with the annual correlation between all daily model-observation pairs rising from 0.42 to 0.54. The largest uncertainty reductions appeared over major source regions such as London, Paris, and Madrid, where dense satellite coverage and strong emission signals combine to give the filter the most information to work with.

Perhaps the most eye-catching finding is the seasonal pattern that emerges when the posterior NOx fluxes are converted into CO2 estimates. The inferred emissions exhibit a far more pronounced seasonal cycle than the prior inventories, with elevated values in autumn and winter, most notably in February, November, and December, and a close correlation with surface temperature variability. During colder months, the posterior adjustments implied widespread increases in total European daily fluxes, ranging from 10 percent to as much as 177 percent, while warmer months showed decreases of up to 22 percent, particularly over northwestern Europe. The team attributes the winter amplification partly to the difficulty that bottom-up inventories face in capturing domestic heating, which is highly seasonal and depends heavily on meteorology, fuel use, and assumed temporal profiles.

The national-level adjustments were substantial. Annual combustion CO2 emissions increased relative to the prior in every one of the twelve high-emitting countries analysed, ranging from 20 percent in the United Kingdom to 91 percent in Turkey, with the changes exceeding the stated prior uncertainties by factors of roughly six to seventy. In Sweden, where February temperatures stayed below minus five degrees Celsius for the entire month, the satellite-derived estimates captured a distinct spike in emissions. Comparisons against the independent EDGAR inventory showed improved agreement for five countries, including Sweden, Romania, Turkey, Spain, and the United Kingdom, but deteriorated for the remaining seven, suggesting that some of the inferred increases may reflect limitations of the inversion framework rather than genuine biases in national reporting.

Independent evaluation against ground-based measurements lent cautious support to the approach. Comparisons with 202 monitoring stations from the European Environment Agency network showed consistent improvements in correlation, mean absolute error, and bias, with the largest gains during the colder months when the posterior flux adjustments were greatest. Evaluations against 47 in situ CO2 sites from the ICOS and UK DECC networks showed more mixed results, with roughly half to three-quarters of sites improving depending on the season, reflecting the weaker sensitivity of rural background stations to localised fossil fuel plumes and the strong influence of biogenic fluxes. Comparisons with NASA’s OCO-2 satellite column measurements of CO2 showed negligible improvement, underscoring just how difficult it remains to isolate the fossil fuel signal from the vast natural carbon cycle using column CO2 observations alone.

The authors are careful to frame the work as a proof of concept rather than a definitive correction to national inventories. A central limitation is the assumption of temporally fixed NOx to CO2 emission ratios, which in reality vary across sectors, regions, and seasons due to differences in combustion conditions, fuel composition, and emissions control technology. Non-combustion and biogenic NOx sources, particularly soil emissions in summer, were held fixed at prior values, meaning their uncertainties may have been compensated for through adjustments to combustion emissions. The reliance on a single daily satellite overpass also requires prescribed diurnal emission profiles. These caveats mean the inferred increases should be interpreted as initial estimates that warrant further validation and multi-species observational constraints.

Even so, the study arrives at a pivotal moment for climate accountability. As the Copernicus CO2M mission and GOSAT-GW prepare to deliver co-located CO2 and NO2 retrievals at high resolution, and geostationary platforms such as TEMPO add constraints on the diurnal cycle, the Edinburgh team’s reduced-complexity chemistry approach suggests that the historical trade-off between chemical accuracy and computational feasibility need not stand in the way of operational monitoring. The next step is a fully coupled NO2-CO2 inversion in which both species are assimilated simultaneously, allowing emission ratios themselves to be dynamically constrained. Such systems could move the world closer to a transparent, satellite-based verification capability, one capable of independently checking the emission pledges that underpin international climate accords and national net-zero targets.

Subject of Research: Satellite-based estimation of European fossil fuel CO2 emissions using TROPOMI nitrogen dioxide observations and sector-based NOx to CO2 emission ratios

Article Title: Inferring European fossil fuel CO2 emissions using TROPOMI NO2 data and sector-based NOx : CO2 emission ratios

Article References: Schooling, C. N., Feng, L., Super, I., & Palmer, P. I. (2026). Inferring European fossil fuel CO 2 emissions using TROPOMI NO 2 data and sector-based NO x : CO 2 emission ratios. Atmospheric Chemistry and Physics, 26(19), 13743-13766. https://doi.org/10.5194/acp-26-13743-2026

Image Credits: AI Generated

DOI: 10.5194/acp-26-13743-2026

Keywords: fossil fuel CO2 emissions, TROPOMI, nitrogen dioxide, NOx, data assimilation, Ensemble Kalman Filter, GEOS-Chem, satellite monitoring, Europe, carbon emissions verification, emission ratios, climate mitigation

The study, published in the journal Biogeosciences by Pierre Vanderbecken of Météo-France and colleagues from institutions including the European Centre for Medium-Range Weather Forecasts, the Laboratoire des Sciences du Climat et de l’Environnement, and Forschungszentrum Jülich, tackles a long-standing problem in Earth system science. Land surface models, which simulate the exchange of water, energy and carbon between the ground, vegetation and the atmosphere, are the backbone of climate monitoring and carbon budget accounting. Yet the natural carbon flux from the land surface remains the most uncertain component of the global carbon budget, a gap that matters enormously for tracking whether countries are meeting their emission reduction commitments under the Paris Agreement.

The satellite at the heart of the work is the Copernicus Sentinel-5 Precursor, which carries an instrument called TROPOMI. With a vast 2,600-kilometre-wide swath and a heliosynchronous orbit, TROPOMI passes over any given region at least once a day, retrieving SIF in the near-infrared part of the spectrum within a narrow window between 743 and 753 nanometres. The retrieval works by statistically modelling the top-of-atmosphere radiance over vegetation and subtracting the contribution of the underlying bare surface, which means the resulting distribution of observations is centred on zero over barren ground and can occasionally dip into small negative, non-physical values. The instrument’s nadir measurement crosses the equator at 13:30 local solar time with a resolution of 7 by 3.5 kilometres, offering a daily, near-global picture of photosynthetic activity that no previous instrument could match.

There was, however, a catch. The ISBA land surface model, developed at Météo-France and distributed through the SURFEX modelling platform, simulates variables such as leaf area index, soil moisture and gross primary production, but it does not simulate SIF itself. To assimilate the satellite observations into the model, the researchers needed an observation operator: a mathematical bridge that translates model variables into the space of the observations. A physically based SIF operator would have to solve complex radiative transfer processes and would demand heavy computational resources and careful calibration, making it impractical for an operational system. The team’s alternative was a machine learning approach, adapting a methodology previously used to assimilate radar backscatter observations from the ASCAT satellite.

The resulting neural network is elegantly simple in concept but sophisticated in its training. It is a feed-forward network with two hidden layers of 128 neurons each, using ReLU activation, and a single linear output neuron. A Gaussian noise layer, active only during training, was inserted before the first activation to reproduce the effect of noisy inputs and to prevent overfitting, while batch normalisation layers at the end of each hidden layer regularised the learning process. Because the distribution of TROPOMI SIF values resembles a log-normal distribution with a mode close to zero, the network was trained on a transformed version of the signal to avoid non-physical negative predictions. The loss function was a Huber loss, which behaves like a mean-square loss for small errors but switches to a more linear behaviour for extreme differences, making the training less sensitive to outliers in the noisy satellite data.

Crucially, the network takes only four inputs: the day of the year, the latitude, the longitude, and the observed leaf area index from the Copernicus Land Monitoring Service. This minimal design means the operator can be plugged into any land surface model and will remain robust as those models improve. It was trained on a vast database covering Europe from 26 degrees west to 46 degrees east and 28 to 72 degrees north on a regular 0.1-degree grid, drawing on TROPOMI SIF retrievals from the TROPOSIF project spanning 2018 to 2021. After screening out oceans, lakes, snow, urban areas, high-altitude regions and frozen ground, the network learned from roughly 14 million input-output pairs, trained on one year of data and tested on a completely different year to verify that it could generalise. The results were strong: a Pearson correlation above 0.8 between the network’s estimates and the observed SIF, and a root mean square error of around 0.15 milliwatts per square metre per steridian per nanometre, with no evident overfitting between the training and test years.

With the trained operator in hand, the team embedded it in a land data assimilation system that uses a simplified extended Kalman filter to nudge the ISBA model towards the observations on a daily cycle. The control vector comprises the leaf area index and soil moisture in different soil layers, and the Kalman gain weighs the background error against the observation error to compute each day’s analysis. The proving ground was the Ebro basin in Spain, a region famous for its heavily irrigated croplands, which conventional land surface models struggle to represent because irrigation is a human decision, not a natural process the model knows about. The researchers ran a suite of experiments: an open-loop simulation with no assimilation, a baseline assimilating the standard 10-day leaf area index product with a fixed 20 percent relative error, several experiments assimilating TROPOMI SIF alone with different observation error assumptions, and co-assimilation experiments combining both data streams.

The results were striking. After SIF assimilation, the leaf area index improved across the domain, with analysis increments reaching up to 0.2 square metres per square metre over the C3 croplands. Most tellingly, the assimilation revealed a confined area around the Ebro basin where the leaf area index increased during the summers, a pattern repeated in every year of the experiment and corresponding to artificially irrigated land. Irrigation sustains both the fluorescence and the foliage while surrounding areas dry out, and the satellite observations captured this signature even though the model never simulated it. The residuals, meaning the differences between the analysed SIF and the observations, were consistently smaller than the innovations, the differences before assimilation, confirming that the filter was genuinely pulling the model towards reality. Over irrigated pixels, the late-summer regrowth of vegetation that the open-loop simulation completely missed was faithfully reproduced in the analysis.

Perhaps the most surprising finding concerned gross primary production, the gross carbon flux that SIF is famous for tracking. Although assimilating SIF improved the correlation with the independent FLUXCOM-X GPP product on average across the domain, the local gains were modest, and a marked difference remained between simulated and observed values. The reason lies in the internal logic of the model: simulated GPP depends not only on the vegetation state but also on soil water content. When SIF assimilation added extra leaf area in summer, the model’s soil dried out faster and evapotranspiration increased, which prevented the computed GPP from rising, even though the observed product captured a second late-summer surge driven by irrigation. Because the assimilation cannot infer an increase in soil water supply that the model does not represent, SIF alone could not close that final gap. The comparison with airborne SIF measurements from the HyPlant campaign also showed that the neural network is an emulator, not a substitute for a physically based fluorescence model.

The broader lessons are twofold. First, assimilating daily TROPOMI SIF alone delivered benefits comparable to the standard assimilation of the 10-day leaf area index product, while providing roughly five times more observations to the system, and an observation error of 20 percent proved close to optimal. Second, and most powerfully, co-assimilating SIF with the leaf area index product outperformed either observation alone, combining the high-frequency, day-to-day sensitivity of fluorescence with the robust, low-noise anchor of the 10-day vegetation product. In the co-assimilation framework, a more complex error description for SIF even improved the results further. The authors suggest the approach could be extended to other SIF missions and to additional satellite products such as land surface temperature. For a scientific community racing to build a near-real-time monitoring and verification capacity for the carbon cycle, the message is clear: the faint glow of photosynthesis, filtered through a modest neural network, can now help models see the human fingerprints on the land surface, one irrigated field at a time.

Subject of Research: Deep learning-based assimilation of satellite sun-induced chlorophyll fluorescence into a land surface model to improve vegetation and carbon flux simulation

Article Title: Using deep learning to assimilate sun-induced fluorescence satellite observations in the ISBA land surface model

Article References: Vanderbecken, P., Vural, J., Rojas-Muñoz, O., Garrigues, S., Bonan, B., Bacour, C., Rascher, U., Siegmann, B., de Rosnay, P., & Calvet, J.-C. (2026). Using deep learning to assimilate sun-induced fluorescence satellite observations in the ISBA land surface model. Biogeosciences, 23(18), 6687-6703. https://doi.org/10.5194/bg-23-6687-2026

Image Credits: AI Generated

DOI: 10.5194/bg-23-6687-2026

Keywords: sun-induced fluorescence, TROPOMI, Sentinel-5P, data assimilation, deep learning, neural network, land surface model, ISBA, leaf area index, gross primary production, irrigation, carbon cycle

Researchers Ruizhe Huang and Sherrie Wang of the Massachusetts Institute of Technology analyzed more than 1.5 million satellite observations of over 6,000 power plants worldwide, drawing on data from the TROPOspheric Monitoring Instrument, or TROPOMI, aboard the Copernicus Sentinel-5P satellite. Their findings, published in the journal Atmospheric Measurement Techniques, represent the first global, data-driven account of when and where power plant nitrogen dioxide plumes become visible from space. The work reveals that detectability is not simply a function of how much a plant pollutes. Instead, it emerges from an intricate interplay of wind, terrain, surface brightness, and the geometry of the satellite’s view.

The scale of the analysis is unprecedented. For the United States, the team paired TROPOMI observations from 2019 through 2024 with hourly emissions records from the Environmental Protection Agency’s Continuous Emissions Monitoring Systems, covering the 500 largest power sector emitters of nitrogen oxides. Globally, they used annual emission estimates from the CoCO2 catalog for the 6,000 highest-emitting plants, spanning coal, natural gas, oil, biomass, and waste facilities. After quality filtering, the dataset contained 666,222 high-quality US observations and 875,686 global ones, each capturing a single satellite overpass of a specific plant.

A central challenge was teaching a computer to recognize a plume and correctly attribute it to its source. The researchers developed an automated algorithm that masks potential interference from nearby cities and other power plants, uses wind direction to define a downwind search zone, calculates the background nitrogen dioxide concentration from upwind air, and then applies a dual threshold requiring both statistical significance and an absolute minimum signal strength. Validated against 800 manually labeled observations stratified by continent and emission level, the algorithm achieved 98 percent accuracy in both US and global tests, a level of reliability that made the subsequent analysis possible.

Even with this powerful detection tool, the researchers found that geography matters enormously. Among 171 US plants that survived the study’s strict interference filtering, per-plant detectability ranged from 6 percent to 85 percent of overpasses. Globally, among 1,065 plants, the range stretched from zero to 100 percent, with a median of 27.8 percent. The pattern follows a striking geographic logic: detectability is highest in arid and semi-arid regions such as the American interior West, Iberia, North Africa, Greece and Turkey, and eastern Australia, and lowest in humid, cloudy regions where bright surfaces are rare and cloud cover frequently spoils the retrieval.

To untangle the physical causes, the team trained a machine learning model, a multi-layer perceptron neural network, to predict whether a plume would be detectable from 25 variables describing the sensor, the plant, the meteorology, and the environment. The model performed well, achieving F1 scores above 0.66 and areas under the curve exceeding 0.8. Feature importance analysis using permutation scores and SHAP values identified six dominant factors: the nitrogen oxide emission rate, surface altitude, surface albedo in the nitrogen dioxide spectral window, the satellite’s viewing zenith angle, the plant’s primary fuel type, and wind speed.

Some of these relationships confirm long-standing theory. Higher emissions produce more concentrated plumes and higher detection probability, and brighter surfaces improve the signal-to-noise ratio of the retrieval by increasing the amount of light available for absorption spectroscopy. Wind speed works in the opposite direction: strong winds stretch and dilute a plume, lowering the concentration contrast that the satellite must detect. The dependence on viewing angle proved more subtle, with detectability peaking at intermediate off-nadir angles, near 20 degrees globally and 35 degrees for US plants, where the longer optical path through the lower atmosphere amplifies the absorption signal before oblique geometry begins to degrade the retrieval.

The numbers offer a practical benchmark. For US power plants, an average hourly emission rate of roughly 400 kilograms of nitrogen oxides corresponds to about a 50 percent chance of detection, but under different combinations of wind, albedo, and viewing geometry, the same emission rate can yield detection probabilities anywhere from below 20 percent to above 60 percent. The researchers illustrate this with individual facilities: a low-emitting plant with dark surfaces and modest winds reached only 18 percent detectability, while a plant emitting eight times as much, blessed with bright terrain and calm air, achieved 64 percent. Identical pollution, in short, can be seen or missed depending entirely on local conditions.

The study also exposed a sobering limitation of current satellite monitoring. To ensure that detected plumes could be attributed to a single source, the researchers excluded any plant within 20 kilometers of another major power plant or within 45 to 90 kilometers of a large city, depending on the city’s size. This filtering removed a staggering share of the world’s monitoring targets: 82.3 percent of the 6,000 global plants fell inside interference zones, leaving just 21.1 percent of their combined emissions in the analysis. In the United States, where emission sources are less clustered, 45 percent of emissions survived the filter. The finding underscores how difficult it remains to monitor pollution from the densely industrialized regions where it matters most.

The implications reach well beyond this single study. The same variables that govern whether a plume is visible also shape how it appears, meaning that emission-quantification methods could become more accurate by explicitly accounting for wind, albedo, and viewing geometry as auxiliary variables or priors. And as a new generation of satellites comes online, including missions targeting pixel sizes of a few kilometers or even 300 meters, finer spatial resolution should reduce plume dilution and interference, potentially bringing many more of the world’s power plants within reach of space-based monitoring. For now, the study provides the first empirical map of where today’s satellites can and cannot see, a foundation for turning orbital observations into trustworthy global emission accounts.

Subject of Research: Satellite detectability of nitrogen dioxide plumes from power plants

Article Title: Global variability in the detectability of power plant NO2 plumes from space

Article References: Huang, R., & Wang, S. (2026). Global variability in the detectability of power plant NO 2 plumes from space. Atmospheric Measurement Techniques, 19(18), 6099-6124. https://doi.org/10.5194/amt-19-6099-2026

Image Credits: AI Generated

DOI: 10.5194/amt-19-6099-2026

Keywords: nitrogen dioxide, power plants, TROPOMI, satellite remote sensing, air pollution, machine learning, emissions monitoring, plume detection, Sentinel-5P, NOx, surface albedo, detectability

The first instrument, the TROPOspheric Monitoring Instrument known as TROPOMI, flies aboard the European Space Agency’s Sentinel-5 Precursor satellite and measures the dry-air column-volume mixing ratio of methane, abbreviated XCH4, across swaths of the planet every day. Because TROPOMI observes the entire atmospheric column between the ground and the satellite, it excels at revealing broad regional patterns: which provinces accumulate the most methane, how those concentrations shift with the seasons, and whether the overall burden is rising or falling over time. What it cannot do, given pixels that span several kilometers, is pinpoint a single leaking pipeline valve or a specific corner of a landfill.

That is where the second instrument comes in. EMIT, the Earth Surface Mineral Dust Source Investigation, was installed on the International Space Station originally to map the minerals in desert dust, but its imaging spectrometer turned out to be remarkably good at something else entirely: detecting the narrow absorption fingerprints that methane plumes leave in reflected sunlight. With spatial resolution fine enough to distinguish individual facilities, EMIT can spot a plume drifting downwind of a gas processing plant or a waste dump and, by modeling how much gas must be present to produce the observed spectral signal, estimate the emission rate in tonnes per hour. The Pakistani team leveraged both instruments in tandem, using TROPOMI to establish the big picture and EMIT to zoom in on the culprits.

The regional picture that emerges from the TROPOMI record is sobering. Between 2019 and 2024, the average atmospheric methane concentration over Pakistan climbed from 1878 ± 13 parts per billion to 1933 ± 14 parts per billion, an increase of roughly 55 parts per billion in just six years. That trajectory mirrors the global rise in methane, which has been accelerating since 2007, but it underscores that a country with rapidly growing energy demand, expanding cities, and a large agricultural sector is tracking at the high end of concern. Methane is more than eighty times more effective than carbon dioxide at trapping heat over a twenty-year horizon, so changes of this magnitude translate into meaningful warming pressure in the near term.

Geographically, the highest methane columns appeared repeatedly over the eastern parts of the country. The Rann of Kutch, the Larkana district, and major urban and industrial centers including Islamabad, Karachi, Lahore, Hyderabad, Multan, Faisalabad, and Gujranwala all registered elevated values. The researchers attribute these hotspots to a combination of natural and anthropogenic activity: dense populations, heavy traffic, industry, agriculture, and in the case of the Rann of Kutch, wetland and seasonal flooding dynamics that foster microbial methane production. The clustering of hotspots over Punjab’s cities and Sindh’s industrial corridor aligns with earlier studies of Pakistan’s air quality, which have ranked several of these cities among the most polluted in the world for particulate matter as well.

Seasonality added another layer to the analysis. Methane concentrations peaked in autumn at 1937 ± 17 parts per billion, followed by summer and then winter. This seasonal rhythm likely reflects a mix of factors, including the timing of rice cultivation and crop residue burning, temperature-driven microbial activity in wetlands and landfills, and atmospheric circulation patterns that either ventilate or trap emissions over the region. The Asian monsoon in particular is known to lift pollution from South Asia into the upper atmosphere, and seasonal wind reversals can redistribute methane across national borders, complicating any attempt to attribute concentrations to sources within a single country.

The most consequential findings, however, came from the EMIT observations, which allowed the team to move from regional averages to named facilities. Between January 2023 and December 2024, the researchers identified and quantified methane plumes originating from three sectors: oil and gas infrastructure, landfills, and electricity generation. The oil and gas sector dominated the tally. The study documented 26 plumes with emission rates exceeding 0.5 tonnes of methane per hour from oil and gas sites, together releasing approximately 78.3 ± 14.7 tonnes of methane every hour. Sustained over a year, emissions of that scale represent millions of tonnes of carbon dioxide-equivalent warming potential, all from a set of facilities that could, in principle, be repaired.

One location stood out above all the rest. Daharki, a town in Sindh province at the heart of Pakistan’s gas-producing region, accounted for the largest number of plumes, with 12 separate detections and total emissions approaching 53.2 ± 9.2 tonnes of methane per hour. In other words, a single cluster of facilities near Daharki was responsible for roughly two-thirds of the oil and gas emissions quantified in the entire study. This pattern echoes what researchers have found in other major producing basins around the world, from the Permian Basin in Texas to offshore platforms in the Gulf of Mexico: a small number of so-called super-emitters, often operating under abnormal process conditions such as equipment malfunctions or venting events, contribute a disproportionate share of total emissions. The implication is that targeted leak detection and repair at a handful of sites can deliver outsized climate benefits.

Landfills emerged as the second major source category, and the detections were unambiguous. The team identified strongly emitting landfill sites releasing between 1 and 5.3 tonnes of methane per hour at Jam Chakro in Karachi, Losar in Rawalpindi, and Lakhodair in Lahore. These are the principal waste disposal sites serving three of Pakistan’s largest cities, and their methane output is the natural consequence of organic waste decomposing anaerobically in massive open dumps. The finding carries a double significance. First, it confirms that unmanaged waste in rapidly urbanizing South Asian cities is a methane source of global relevance, consistent with satellite surveys that have uncovered large landfill emissions worldwide. Second, it points to an opportunity: capturing landfill gas for energy generation, a strategy long discussed in Pakistan, would simultaneously reduce emissions, improve local air quality, and produce usable fuel.

The study arrives at a politically charged moment. Under the Global Methane Pledge, more than 150 countries have committed to cutting global methane emissions by at least 30 percent from 2020 levels by 2030, and Pakistan’s biennial transparency reporting obligations under the UN climate convention require it to account for its emissions with increasing rigor. Until now, the country’s methane inventories have relied heavily on bottom-up estimates and global models that may poorly represent local conditions. By demonstrating that freely available satellite data can both track national trends and expose specific leaking facilities, the researchers have effectively handed Pakistani regulators, and the international community, a verification toolkit. The technical approach, pairing a wide-swath column instrument with a point-source imaging spectrometer, is the same two-tier strategy now being applied in China, India, and the United States, and the Pakistani analysis shows it works even in data-scarce environments. Whether the plumes over Daharki, Jam Chakro, and Lakhodair translate into repairs and policy action remains to be seen, but the era of invisible methane over South Asia is clearly coming to an end.

Subject of Research: Satellite detection and quantification of methane hotspots and emission sources over Pakistan using TROPOMI and EMIT observations

Article Title: Detecting methane hotspots and anomalies from space over Pakistan: analysis using sentinel-5P TROPOMI and EMIT observations

Article References: Mariam, A., Haq, Z. U., & Batool, S. A. (2026). Detecting methane hotspots and anomalies from space over Pakistan: analysis using sentinel-5P TROPOMI and EMIT observations. Air Quality, Atmosphere & Health, 19(10), Article 230. https://doi.org/10.1007/s11869-026-02108-z

Image Credits: AI Generated

DOI: 10.1007/s11869-026-02108-z

Keywords: methane, TROPOMI, EMIT, Pakistan, satellite remote sensing, greenhouse gases, oil and gas emissions, landfills, super-emitters, air quality, climate change, Sentinel-5P

The technical foundation of the study is the TROPOspheric Monitoring Instrument, or TROPOMI, which flies aboard the European Space Agency’s Sentinel-5P satellite launched in 2017. TROPOMI measures the sunlight reflected from Earth’s atmosphere in narrow spectral bands, allowing researchers to retrieve column concentrations of gases such as nitrogen dioxide and carbon monoxide with spatial resolution fine enough to distinguish one urban district from another. Because ground-based air quality monitoring networks in rapidly industrializing countries are often sparse, satellite retrievals have become an indispensable tool for environmental scientists. The approach has already proven its worth in contexts ranging from tracking the European lockdown-related drop in nitrogen dioxide in 2020 to mapping pollution inequality in New York City, and the Bangladeshi team adapted the same logic to a region where industrial growth has dramatically outpaced monitoring infrastructure.

The study’s central finding is a pronounced seasonal cycle. Pollutant concentrations reach their lowest levels during the monsoon, when heavy rainfall scavenges aerosols and gases from the atmosphere and strong convective mixing disperses emissions vertically. In contrast, the winter and post-monsoon periods emerge as the most hazardous times of year. During winter, cooler temperatures and temperature inversions trap pollutants near the surface, while reduced wind speeds limit horizontal transport. The researchers’ hotspot analysis—built on spatial autocorrelation statistics that identify statistically significant clusters of high values—pinpointed severe carbon monoxide pollution in winter across central, northern, and southern Dhaka. Nitrogen dioxide, a gas closely tied to vehicle exhaust, brick kilns, and industrial combustion, showed notably elevated levels in Dhaka, Narayanganj, and Gazipur during both the post-monsoon and winter seasons. Some pollutants also peaked during the pre-monsoon months, adding a third seasonal window of concern.

What distinguishes this work from a simple pollution inventory is its use of Geographically Weighted Regression, a spatial statistical technique that allows the relationship between pollutant concentrations and meteorological drivers to vary from place to place rather than assuming a single city-wide average. Conventional global regression would collapse these local dynamics into one coefficient, masking the fact that the same weather variable can suppress pollution in one district while amplifying it in another. The GWR analysis revealed precisely such spatial non-stationarity. Ozone displayed a strong positive relationship with precipitation in Gazipur during the monsoon, with local coefficients ranging from 0.79 to 2.15, while showing a negative relationship with temperature in Narsingdi during the pre-monsoon season, with coefficients between −2.23 and −1.88.

Nitrogen dioxide told an equally nuanced story. In Gazipur and Dhaka during the monsoon, the gas showed a positive local coefficient with temperature, ranging from 1.02 to 1.47, suggesting that warmer conditions in those industrial cores coincided with higher nitrogen dioxide levels—possibly reflecting enhanced photochemical activity and industrial output. In Narsingdi during the pre-monsoon period, however, nitrogen dioxide was negatively associated with precipitation, with coefficients between −1.66 and −1.38, consistent with rainfall washing the pollutant out of the boundary layer. Formaldehyde, sulfur dioxide, and carbon monoxide each exhibited their own patchwork of relationships with humidity, atmospheric pressure, temperature, wind speed, and precipitation, shifting in sign and magnitude across seasons and locations. For policymakers, this means that a single national control strategy calibrated to average conditions will inevitably miss the mark in specific cities and specific months.

The health implications of these patterns are far from abstract. Carbon monoxide binds to hemoglobin in the blood with an affinity hundreds of times greater than oxygen, reducing the body’s capacity to deliver oxygen to tissues and organs; chronic exposure is linked to cardiovascular strain and neurological effects. Nitrogen dioxide inflames the airways and aggravates asthma, particularly in children. Ozone at ground level is a potent oxidant that damages lung tissue, and formaldehyde is classified as a hazardous air pollutant with carcinogenic potential. Sulfur dioxide, emitted largely from industrial boilers and brick kilns burning sulfur-containing fuel, contributes to both respiratory disease and acid deposition. In the industrial cities examined here, the people living closest to factory clusters—often the workers themselves and low-income communities who cannot afford to move away—face the highest cumulative exposures, a textbook case of environmental injustice.

This justice dimension is where the study’s framing becomes most pointed. Environmental justice research has long documented that the burdens of industrialization fall disproportionately on those who benefit least from it, and satellite data have increasingly been used to document such disparities in the United States and elsewhere. The Bangladeshi analysis extends that evidence base to South Asia, where the scale of industrial expansion—textile dyeing, steel re-rolling, brick manufacturing, and chemical processing—has transformed the urban atmosphere within a single generation. The authors emphasize that workers and residents near industrial zones face significant health risks, and that this unequal exposure obstructs progress toward Sustainable Development Goal 3 on health and well-being, Goal 7 on affordable and clean energy, and Goal 11 on sustainable cities and communities. Clean air, in other words, is not merely an environmental issue but a prerequisite for the entire development agenda.

The meteorological findings also carry a warning about climate change. Previous research has established that a warming climate can alter air quality through changes in temperature, precipitation, and circulation patterns, and the strong local coefficients documented here suggest that such effects will not be uniform. If monsoon rainfall patterns shift, the natural cleansing mechanism that currently keeps summer pollution in check could weaken. If winter inversions intensify or become more prolonged, the seasonal pollution peaks that already endanger public health could grow worse. Understanding the precise, location-specific sensitivity of each pollutant to each weather variable—as the GWR framework makes possible—is therefore a critical input for any forward-looking air quality management plan in Bangladesh or in other rapidly industrializing nations facing similar conditions.

The practical value of the study lies in its ability to guide targeted intervention. Because the hotspot analysis identifies exactly where and when pollution clusters form—central, northern, and southern Dhaka for winter carbon monoxide, for example, and the Dhaka–Narayanganj–Gazipur corridor for post-monsoon and winter nitrogen dioxide—regulators can prioritize inspections, fuel switching, and emission controls in those areas during those seasons rather than spreading resources thinly across the year and the map. Brick kiln modernization, stricter enforcement of effluent and emission treatment requirements in the textile sector, and expansion of the ground-based monitoring network to validate satellite retrievals all follow naturally from the evidence. The authors note that their findings can inform Bangladesh and other industrialized nations in identifying pollution sources and implementing effective mitigation strategies for healthier, more sustainable urban environments.

Ultimately, the study is a demonstration of how a single satellite instrument, combined with rigorous spatial statistics, can illuminate problems that ground-level monitoring alone would miss. It transforms air pollution in Bangladesh’s industrial cities from an invisible, ambient grievance into a mapped, measured, and seasonally predictable phenomenon—one whose burdens can be traced to specific neighborhoods and specific months. That transformation matters, because pollution that can be measured precisely can be regulated precisely, and regulation that respects the geography of exposure is the first step toward the environmental justice that the Sustainable Development Goals promise. For the millions of people living in the shadow of Dhaka’s factories, the winter air they breathe is now documented in numbers, and those numbers make a compelling case for action.

Subject of Research: Seasonal air pollution patterns, meteorological drivers, and environmental justice in the industrial cities of Bangladesh

Article Title: Impact of air pollution on sustainable development and environmental justice: Insights from industrial cities of Bangladesh

Article References: Laskor, M. A. H., Kadir, A., Ahmed, Z., Zhang, F., & Al-Rezoan, H. (2026). Impact of air pollution on sustainable development and environmental justice: Insights from industrial cities of Bangladesh. Theoretical and Applied Climatology, 157(9), Article 604. https://doi.org/10.1007/s00704-026-06555-7

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06555-7

Keywords: air pollution, Bangladesh, Sentinel-5P, TROPOMI, environmental justice, Sustainable Development Goals, nitrogen dioxide, carbon monoxide, geographically weighted regression, industrial cities, Dhaka, seasonal variation

The stakes could hardly be higher. Methane is the second most significant anthropogenic greenhouse gas after carbon dioxide and is responsible for roughly one third of the rise in global mean surface air temperature between 1750 and 2019. Because it has a short atmospheric lifetime of about nine years but a global warming potential nearly 82 times that of carbon dioxide over a 20-year period, cutting methane emissions is one of the most powerful levers available for slowing near-term climate change. More than 150 countries have now committed, through the Global Methane Pledge and their Nationally Determined Contributions, to reduce methane emissions across all sectors by at least 30 percent below 2020 levels by 2030. Every credible pathway to limiting warming to 1.5 degrees Celsius depends on achieving that reduction.

Yet cities, where more than half of humanity already lives and where the proportion is projected to reach 70 percent by 2050, remain stubbornly difficult to audit. Urban methane comes from a tangle of sources: natural gas consumption, leaks from aging pipeline networks, landfills and wastewater treatment systems. Field studies have repeatedly found that official bottom-up inventories, which rely on activity data and emission factors, can underestimate urban methane emissions by a factor of two to three. Aircraft surveys, ground-based mobile monitors and tower networks can deliver accurate snapshots of a city’s emissions, but they cannot provide the continuous, repeatable monitoring needed to track progress year after year, and their spatial sensitivity is often limited.

Satellites promise to close that gap. The TROPOspheric Monitoring Instrument, or TROPOMI, flies aboard the polar sun-synchronous Sentinel-5 Precursor satellite, launched in October 2017 at an altitude of 824 kilometers with a consistent 13:30 local overpass time. The instrument retrieves the methane column by detecting solar backscattered light in the shortwave infrared absorption band at 2.3 micrometers, achieving daily global coverage at a spatial resolution of 7 by 5.5 square kilometers at nadir since August 2019. Previous studies have used TROPOMI to constrain urban methane inventories across 61 cities worldwide and to conduct a preliminary national analysis of urban emissions across North America, but a transparent, widely applicable framework with detailed error accounting has remained elusive.

The Manchester team, led by Huihui Long together with Maria Tsivlidou, Hugo Ricketts and Grant Allen, built their method on the concept of mass balance, refining the established source pixel approach. First, they remapped the heterogeneous Level-2 satellite pixels onto a regular 0.1 by 0.1 degree grid, treating each observation footprint as a surface polygon and calculating the fractional area of overlap with each grid cell, weighted by pixel area and retrieval error. This regularization allows orbit-level emission estimates that are directly comparable with gridded bottom-up inventories over the same spatial domain. Strict quality filters followed: only cloud-free retrievals with quality assurance values above 0.5, solar zenith angles below 70 degrees, smooth topography, low aerosol optical thickness and retrieval precisions better than 10 parts per billion were retained.

Two innovations distinguish the new framework. The first concerns wind. Conventional source pixel methods rely on 10-meter near-surface winds bundled with the satellite retrievals, but methane emitted at the surface is ventilated by winds throughout the planetary boundary layer, the turbulent lowest kilometer or so of the atmosphere where surface emissions mix and are advected across urban scales. The team therefore computed pressure-weighted average wind speed and direction across all ERA5 reanalysis pressure levels within the boundary layer, using hourly-resolved meteorology. The difference matters: over Los Angeles, the near-surface winds pointed from the north-northeast to east-northeast at speeds below 6 meters per second, while the boundary-layer-weighted winds blew predominantly from the west-southwest to south-southwest at up to about 10 meters per second. Since emission rate is conceptually proportional to mean boundary-layer wind speed, and wind direction determines where the upwind background lies, these discrepancies directly reshape the calculated emissions.

The second innovation is an upwind-based approach to defining the background methane concentration. Instead of assuming that the air surrounding a city is clean, the method defines a background region of equal area located immediately adjacent to the city along the boundary-layer-pressure-weighted mean upwind direction, with both boxes oriented so their sides lie perpendicular to the wind. Emissions in the background are implicitly accounted for and subtracted, so the background region need not be free of methane sources. The methane enhancement over the city is then the difference between the mean column mixing ratio over the source region and that over the upwind background, and the emission rate follows from a mass balance equation combining that enhancement, the mean boundary-layer wind, surface pressure and the dimensions of the source box. Coverage thresholds, requiring at least 25 percent valid data over the source region and at least 10 background observations, guard against statistically unreliable single-orbit estimates.

Applied to the three megacities, the method produced three-year mean emissions of 7.78 plus or minus 4.84 tonnes per hour for London, 47.19 tonnes per hour for Los Angeles, and 42.94 tonnes per hour for New York. London showed only a modest urban enhancement, averaging between 1.47 and 1.98 parts per billion across the study years, which made the estimate highly sensitive to background variability; roughly 80 percent of the total emission uncertainty there was attributable to the methane enhancement, and an anomalously low October in 2021, when background concentrations exceeded urban ones, dragged the annual figure down until it was excluded as an outlier. Los Angeles, by contrast, displayed a pronounced and persistent hotspot, with emissions rising from 26.21 tonnes per hour in 2021 to 62.77 tonnes per hour in 2023, though the large uncertainties mean a previously reported declining trend cannot be ruled out. In New York, emissions ranged from 30.85 to 44.77 tonnes per hour, with a more balanced uncertainty structure in which wind variability contributed about 40 percent.

The comparison with official inventories is sobering. Over identical spatial domains, EDGAR reported a mean of 20.07 tonnes per hour for Los Angeles, where landfills account for more than 90 percent of inventory emissions, and just 4.38 tonnes per hour for New York, roughly 11 percent of the satellite-derived figure. The UK’s National Atmospheric Emission Inventory underestimated London’s emissions by up to 45 percent in 2022, while the US EPA inventory fell short by approximately 56 percent for Los Angeles and 84 percent for New York. Encouragingly, when the team rescaled estimates from six previous top-down studies to the same urban domains, most fell within their uncertainty bounds, suggesting the new method is consistent with independent measurement-led approaches while being far simpler to repeat.

The authors are candid about the limitations. Cloud cover, coastal albedo heterogeneity and aerosol scattering eliminated the majority of overpasses, leaving as few as five valid orbits over New York in 2021 and almost no usable summer data for Los Angeles, where persistent marine boundary-layer clouds dominate. Such uneven sampling can introduce seasonal and inter-annual biases that must be weighed when comparing annualized emissions with inventories. Nevertheless, the team argues that the trade-off is necessary to preserve statistical reliability, and that the framework’s transparent uncertainty budgeting, combining retrieval, albedo, aerosol and wind uncertainties in quadrature, offers a repeatable path forward. With the Global Methane Pledge’s 2030 deadline approaching, the ability to audit cities from space, anywhere on Earth including poorly observed regions such as India and China where most of the global population resides, may prove essential for verifying whether stated methane targets are actually being met.

Subject of Research: Satellite-based quantification of urban methane emissions using TROPOMI observations and an improved source pixel method

Article Title: Satellite-based global monitoring of urban-scale methane emissions

Article References: Long, H., Tsivlidou, M., Ricketts, H., & Allen, G. (2026). Satellite-based global monitoring of urban-scale methane emissions. Atmospheric Measurement Techniques, 19(18), 6171-6191. https://doi.org/10.5194/amt-19-6171-2026

Image Credits: AI Generated

DOI: 10.5194/amt-19-6171-2026

Keywords: methane, TROPOMI, satellite monitoring, urban emissions, greenhouse gases, London, Los Angeles, New York, planetary boundary layer, EDGAR inventory, climate change, Global Methane Pledge