Search

Nationally reported greenhouse gas inventories are a core component of the Paris Agreement’s transparency framework. Comparisons with emission estimates derived from atmospheric observations help identify improvements to reduce uncertainties and increase the confidence in reported values. To facilitate comparisons over the contiguous United States, we present a 0.1° × 0.1° gridded inventory of annual 2012–2018 anthropogenic methane emissions, allocated to 26 individual source categories, with scale-dependent error estimates. Our inventory is consistent with the U.S. Environmental Protection Agency (EPA) Inventory of U.S. Greenhouse Gas Emissions and Sinks (GHGI), submitted to the United Nations in 2020. Total emissions and patterns (spatial/temporal) reflect the activity and emission factor data underlying the GHGI, including many updates relative to a previous gridded version of the GHGI that has been extensively compared with observations. These underlying data are not generally available in global gridded inventories, and comparison to EDGAR version 6 shows large spatial differences, particularly for the oil and gas sectors. We also find strong regional variability across all sources in annual 2012–2018 spatial trends, highlighting the importance of understanding regional- and facility-level activities. Our inventory represents the first time series of gridded GHGI methane emissions and enables robust comparisons of emissions and their trends with atmospheric observations.

Surface and satellite observations of atmospheric methane show smooth seasonal behavior in the Southern Hemisphere driven by loss from the hydroxyl (OH) radical. However, observations in the Northern Hemisphere show a sharp mid‐summer increase that is asymmetric with the Southern Hemisphere and not captured by the default configuration of the GEOS‐Chem chemical transport model. Using an ensemble of 22 OH model estimates and 24 wetland emission inventories in GEOS‐Chem, we show that the magnitude, latitudinal distribution, and seasonality of Northern Hemisphere wetland emissions are critical for reproducing the observed seasonality of methane in that hemisphere, with the interhemispheric OH ratio playing a lesser role. Reproducing the observed seasonality requires a wetland emission inventory with ∼80 Tg a−1poleward of 10°N including significant emissions in South Asia, and an August peak in boreal emissions persisting into autumn. In our 24‐member wetland emission ensemble, only the LPJ‐wsl MERRA‐2 inventory has these attributes. The amount of methane, a powerful greenhouse gas, has been growing in Earth's atmosphere during the last decade, and scientists disagree about which methane sources and sinks are responsible for the growth. One clue into understanding methane's sources and sinks is their seasonality—their month‐to‐month cycles that happen every year. Measurements of atmospheric methane taken at the Earth's surface and using satellite instruments show a steep increase each summer in the Northern Hemisphere that is not replicated when methane is simulated in a global chemical transport model, indicating missing information about source and sink seasonalities. To investigate, we use that model to simulate 24 representations of methane's largest source, emissions from wetlands, and 22 representations of its largest sink, chemical loss by the hydroxyl radical (OH). We find that OH is unlikely to cause the summer increase and model bias, but the amount, spatial distribution, and seasonal cycles of global wetland emissions are the strongest drivers. We suggest that these characteristics are linked to the underlying mechanisms determining wetland area and methane production in wetland models. The results unveil the role of global wetlands in driving methane's seasonality and inform research to analyze methane's long‐term trends. Northern Hemisphere atmospheric methane shows a summer increase not replicated by the GEOS‐Chem model with its default sources and sinksThe summer increase's timing and magnitude is determined by the magnitude, seasonality, and spatial distribution of NH wetland emissionsInversions of atmospheric methane observations should use a suitable wetland emission inventory and optimize hemispheric OH concentrations Northern Hemisphere atmospheric methane shows a summer increase not replicated by the GEOS‐Chem model with its default sources and sinks The summer increase's timing and magnitude is determined by the magnitude, seasonality, and spatial distribution of NH wetland emissions Inversions of atmospheric methane observations should use a suitable wetland emission inventory and optimize hemispheric OH concentrations

In this paper, we integrated multiple types of predictor variables and three types of machine learners (neural network, random forest, and gradient boosting) into a geographically weighted ensemble model to estimate the daily maximum 8 h O3with high resolution over both space (at 1 km × 1 km grid cells covering the contiguous United States) and time (daily estimates between 2000 and 2016). We further quantify monthly model uncertainty for our 1 km × 1 km gridded domain. The results demonstrate high overall model performance with an average cross-validated R2(coefficient of determination) against observations of 0.90 and 0.86 for annual averages. Overall, the model performance of the three machine learning algorithms was quite similar. The overall model performance from the ensemble model outperformed those from any single algorithm. The East North Central region of the United States had the highest R2, 0.93, and performance was weakest for the western mountainous regions (R2of 0.86) and New England (R2of 0.87). For the cross validation by season, our model had the best performance during summer with an R2of 0.88. This study can be useful for the environmental health community to more accurately estimate the health impacts of O3over space and time, especially in health studies at an intra-urban scale.

We use 2005–2016 observations of formaldehyde (HCHO) columns over China from the OMI, GOME‐2, and SCIAMACHY satellite instruments to evaluate long‐term trends in emission inventories of volatile organic compounds (VOCs) that affect air quality. The observations show large increases over 2005–2016 in the North China Plain (+1.1 ± 0.5% a−1relative to 2005) and the Yangtze River Delta region (+1.5 ± 0.4% a−1relative to 2005), consistent with the trend of anthropogenic VOC emissions in the Multi‐resolution Emission Inventory for China (MEIC). Unlike other pollutants, VOC emissions have not been decreasing in recent years. An exception is the Huai River Basin in rural eastern China where the satellite data show rapidly decreasing VOC emissions since the early 2010s that appear to reflect bans on agricultural fires. Satellite observations of formaldehyde can detect emissions of volatile organic compounds (VOCs), which are of concern for air quality. Here we use 2005–2016 satellite observations of formaldehyde over China to infer trends in VOC emissions. We find increases of 13% in the North China Plain and 18% in the Yangtze River Delta. Unlike other pollutants, emissions of VOCs in China have not decreased in recent years. An exception is in rural areas where post‐2010 restrictions on crop burning to improve air quality have resulted in sharp decreases of VOC emissions. 2005 to 2016 satellite observations of formaldehyde show increases in Chinese anthropogenic emissions of volatile organic compoundsThe increases are largest in the North China Plain and Yangtze River Delta and are consistent with the MEIC emission inventoryDecreases are observed in rural eastern China since the early 2010s that appear due to new restrictions on crop burning

Peroxyacetyl nitrate (PAN) is produced in the atmosphere by photochemical oxidation of non-methane volatile organic compounds in the presence of nitrogen oxides (NOx), and it can be transported over long distances at cold temperatures before decomposing thermally to release NOxin the remote troposphere. It is both a tracer and a precursor for transpacific ozone pollution transported from East Asia to North America. Here, we directly demonstrate this transport with PAN satellite observations from the infrared atmospheric sounding interferometer (IASI). We reprocess the IASI PAN retrievals by replacing the constant prior vertical profile with vertical shape factors from the GEOS-Chem model that capture the contrasting shapes observed from aircraft over South Korea (KORUS-AQ) and the North Pacific (ATom). The reprocessed IASI PAN observations show maximum transpacific transport of East Asian pollution in spring, with events over the Northeast Pacific offshore from the Western US associated in GEOS-Chem with elevated ozone in the lower free troposphere. However, these events increase surface ozone in the US by less than 1 ppbv because the East Asian pollution mainly remains offshore as it circulates the Pacific High.

Arctic observations show large decreases in the concentrations of sulfate and black carbon (BC) aerosols since the early 1980s. These near‐term climate‐forcing pollutants perturb the radiative balance of the atmosphere and may have played an important role in recent Arctic warming. We use the GEOS‐Chem global chemical transport model to construct a 3‐D representation of Arctic aerosols that is generally consistent with observations and their trends from 1980 to 2010. Observations at Arctic surface sites show significant decreases in sulfate and BC mass concentrations of 2–3% per year. We find that anthropogenic aerosols yield a negative forcing over the Arctic, with an average 2005–2010 Arctic shortwave radiative forcing (RF) of −0.19 ± 0.05 W m−2at the top of atmosphere (TOA). Anthropogenic sulfate in our study yields more strongly negative forcings over the Arctic troposphere in spring (−1.17 ± 0.10 W m−2) than previously reported. From 1980 to 2010, TOA negative RF by Arctic aerosol declined, from −0.67 ± 0.06 W m−2to −0.19 ± 0.05 W m−2, yielding a net TOA RF of +0.48 ± 0.06 W m−2. The net positive RF is due almost entirely to decreases in anthropogenic sulfate loading over the Arctic. We estimate that 1980–2010 trends in aerosol‐radiation interactions over the Arctic and Northern Hemisphere midlatitudes have contributed a net warming at the Arctic surface of +0.27 ± 0.04 K, roughly one quarter of the observed warming. Our study does not consider BC emissions from gas flaring nor the regional climate response to aerosol‐cloud interactions or BC deposition on snow. Observed sulfate and BC mass concentrations at Arctic surface sites and Greenland ice cores show decreases of 2–3%/yr between 1980 and 2010Anthropogenic aerosol RF is negative in the Arctic troposphere due to a large negative RF from sulfate in spring (−1.17 ± 0.10 W m−2)The 1980–2010 trends in aerosol‐radiation interactions over the Arctic and NH midlatitudes contributed 0.27 ± 0.04 K warming at Arctic surface

Supplemental Digital Content is available in the text.

Hydroelectric reservoirs can emit significant quantities of methane, particularly through degassing at turbine outlets. Improved understanding of processes affecting hydroelectric reservoir CH4emissions is thus important as the world economy transitions to renewable forms of energy production. Here we develop and evaluate a new mechanistic model of CH4emissions: ResME ([Res]ervoir [M]ethane [E]missions), which estimates carbon inputs and methanogenesis to predict CH4release via ebullition and diffusion, plant emissions, and downstream emissions. ResME results demonstrate that the relative importance of allochthonous and autochthonous carbon input to methane emissions varies by latitude, with allochthonous carbon contributions typically being higher in tropical reservoirs. Results also demonstrate that total reservoir emissions are highly dependent on turbine intake depths, which are not typically reported. Potential maximum degassing emissions from existing hydroelectric reservoirs are estimated as 11 ± 4 Tg C/yr, if all reservoirs had deep turbine intakes and stratified for 5 months per year. In comparison, the estimated diffusive, ebullitive, and plant CH4emissions are estimated to be 2.8 ± 0.2 Tg C/yr (where the true uncertainty is much higher than the model standard error). Future work should focus on improving estimates of reservoir carbon inputs and decomposition rates, as well as surveying turbine intake depths. Satellite measurements from missions such as TROPOMI may also help constrain hydropower methane emissions. Methane is an important greenhouse gas that is naturally produced in lake and reservoir sediment, among other sources. Hydroelectric power reservoirs produce renewable energy, yet also emit methane at their surfaces, and from turbines and downstream reaches. To better understand drivers and pathways of methane emissions, we have developed a new mechanistic model for methane emissions as a function of carbon inputs, chemical decomposition, and physical processes. Results also show that downstream methane emissions have the potential to exceed surface emissions if turbines pull from stratified, anoxic waters. Large uncertainties remain in model inputs, and future work should focus on improved understanding of carbon loading to reservoirs, as well as decomposition rates and turbine intake depths. Our mechanistic model (Reservoir Methane Emissions) illuminates the main drivers of hydropower methane emissionsEmissions from downstream degassing are comparable to surface emissions when turbines are parameterized with deep water intakesWe estimate global emissions from hydropower surfaces as 2.8 ± 0.2 Tg C/yr, plus 11 ± 4 Tg C/yr from downstream degassing Our mechanistic model (Reservoir Methane Emissions) illuminates the main drivers of hydropower methane emissions Emissions from downstream degassing are comparable to surface emissions when turbines are parameterized with deep water intakes We estimate global emissions from hydropower surfaces as 2.8 ± 0.2 Tg C/yr, plus 11 ± 4 Tg C/yr from downstream degassing