{"paper_id":"0b3bbf50-5af7-4147-be53-6d772745b129","body_text":"Modernizing carbon dioxide emissions inventories for action in the United States | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Modernizing carbon dioxide emissions inventories for action in the United States Anastasia Montgomery, Geoff Roest, Jason Zou, Erik Badger, Phil DeCola, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7456083/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract We present 2010–2024 annual model output of a new carbon dioxide emissions model, with a raw resolution at points, lines, and polygons corresponding to emission sources, and a gridded resolution of 1 km² spatial resolution. The underlying model modernizes emissions modeling by incorporating web-scraping and data-fusion methods to update input emissions values as they are ingested, allowing for reliable and consistent updates of the emissions model. From the output, direct CO 2 emissions from fuel combustion for the entire US show that emissions were 5,267 MMTCO 2 in 2024. The largest activity sectors contributing to the national emissions was emissions from electricity production (1,667 MMTCO 2 , 31.6% of national total) and the onroad sector (1,566 MMTCO 2 , 29.7% of national total). By integrating energy forecasts, this model also highlights the potential for enhanced decarbonization policy applications given macroeconomic trends. Further comparison against existing federal datasets such as those of the Environmental Protection Agency’s Greenhouse Gas Emissions Inventory and State Inventory Tool, and independent datasets such as Vulcan, Open-Data Inventory for Anthropogenic Carbon dioxide (ODIAC), and CarbonTracker, demonstrate robust agreement, though variation exists in spatial patterns and the presented dataset comprises relatively higher CO 2 emissions estimates. The flexibility and scalability of the model make it a valuable tool for monitoring CO₂ emissions trends and informing mitigation strategies. Earth and environmental sciences/Climate sciences Physical sciences/Energy science and technology Earth and environmental sciences/Environmental sciences Earth and environmental sciences/Environmental social sciences carbon dioxide emissions greenhouse gas emissions fossil fuel emissions inventory Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction High-resolution carbon emissions data are critical for effective climate change mitigation strategies. Numerous agencies have called for significant reductions in greenhouse gases to limit global temperature rise 1 , 2 . Accurate emissions data enable policymakers to identify specific sources of emissions, assess the effectiveness of existing regulations, and develop targeted interventions to reduce greenhouse gas emissions. Carbon emissions models are integral to this process, allowing for the investigation of improved stewardship of public lands 3 , investigation of local and regional carbon dynamics 3 , and characterization of the effects of urbanization on carbon emissions 4 . Thus, the ability to map emissions at finer spatial resolutions is not merely a technical improvement; it is a fundamental requirement for effective climate action. Numerous CO 2 emissions models have been developed to assess and predict emissions in the U.S., each employing different methodologies and data sources. Models integrating a bottom-up approach use activity data and emission factors to estimate CO 2 at various resolutions 5 – 7 . Other CO 2 emissions models use direct measurements or satellite observations and use atmospheric inversion to get atmospherically-derived fluxes of CO 2 8 . The diversity of these models underscores the necessity for a multi-faceted approach to emissions assessment as different models provide complementary insights into the complexities of carbon emissions in the U.S. However, current CO 2 emissions modeling frameworks in the U.S. require updates to improve spatial resolution and representativeness, and to be made readily accessible for a wide audience of stakeholders. The integration of high-resolution data is increasingly recognized as vital for understanding localized emissions patterns and their drivers 9 . By relying on coarse spatial resolutions, critical variations in emissions can be obscured across different urban landscapes 10 . Furthermore, as the scientific community calls for more robust data to inform climate policy, the need for models that can incorporate real-time data and account for uncertainties in emissions estimates has become apparent 11 . This evolution in modeling approaches is essential not only for compliance with regulatory frameworks, but also for fostering public trust in climate action initiatives. Despite significant progress in CO₂ emissions modeling, critical gaps remain in harmonizing diverse data sources to develop near-real-time estimates with a full spectrum of emission sources. Many current models are constrained by limitations in data availability or consistency, particularly when integrating disparate datasets with different data publication schedules, such as federal inventories, satellite observations, and activity-based emissions estimates. Further, integrating diverse data sources requires harmonization due to inherent differences in methodologies; e.g., many atmospherically-derived CO₂ emissions datasets have shown that usage-based inventories are missing emissions from natural gas 12 . These challenges are compounded by the need to address complex sectors such as the land-use sector, which can significantly impact regional emissions profiles. While advancements in machine learning and data assimilation techniques offer promising pathways for overcoming these challenges, their application to emissions modeling is still in active development 13 . Bridging these gaps will require a concerted effort to leverage emerging technologies and develop models that are both flexible and transparent. These enhancements are not only essential for more precise emissions assessments but also for aligning modeling frameworks with the evolving needs of policymakers, researchers, and the broader public. In this paper, we will describe the methodology that the Crosswalk emissions model implements to create annual CO 2 emissions, review novel data sources and methods including the use of GPS-derived traffic data for onroad vehicle emissions, describe the output from the Crosswalk CO 2 emissions model, and finally compare the emissions output with other independent CO 2 emissions inventories. Methods Model Concept We present a CO 2 emissions inventory for the year 2010–2024, gridded to a 1 km spatial resolution. We integrate the sectoral definitions set by the National Emissions Inventory, with activities based on the EPA’s Source Classification Codes (SCC) and other independent datasets (e.g., GPS-based traffic data). As such, this inventory follows a similar sectoral aggregation to Vulcan 5 and the Anthropogenic Carbon Emissions System 14 . Our domain covers the entire United States excluding US territories (Fig. 1 ). The Crosswalk emissions model is built in Python with a separate module for each activity sector. Input data from public sources is accessed via APIs, webscraping, or downloading directly from an FTP server, then copied to Amazon S3 for traceable usage within the model. In the case of onroad emissions, private data sources are used as described in that subsequent section. Files are stored as Parquet files for efficient handling of larger datasets. The data are transformed to have a standardized schema so that disparate datasets can be merged. The model routine takes one or more data sources, applies emission factors, and applies a logic hierarchy to select the final emission calculation. The spatialization routine takes the calculated emissions and geolocates the emissions output. The spatial surrogates for the model output are customized for each sector, as described in subsequent sector-specific sections. The raw model output is then available in point, line, and polygon shape formats which then can be gridded. The output is also written as GeoParquet files using the newer GeoParquet 1.1 specification, which allows for more efficient filtering of data during downstream geo-processing routines. Using the development framework presented herein, the Crosswalk emissions model can easily digest multiple data sources while still maintaining a consistent output format. Further, this framework allows for quarterly updating, utilizing the latest available datasets and interpolating data where older datasets have more time lag. To generate the model output, the model is run distributed on a cluster of Amazon EC2 instances using AWS Batch which takes approximately five days real-world runtime for end-to-end national output for 14 years of data. The software is version controlled using Github, and we maintain a consistent execution environment across runs via Docker containers. Sector Modeling Buildings The building emissions sector is defined as a subset and downscaled version of the NEI nonpoint emissions. The NEI nonpoint emissions represent the large number of diffuse, small (< 25 MMT CO 2 e) carbon emitters associated with industrial, commercial, and residential activity, such as commercial cooking emissions. We define the industrial, residential, and commercial buildings based on key words in the level-2 SCC definitions, as defined in Supplementary Table S1 . To estimate building-level CO₂ emissions, the module begins by filtering NEI Nonpoint data to include only relevant sources and applying emission factors (Supplementary Table S2) to convert CO emissions to CO₂ (Eq. 1). EIA SEDS fuel data is mapped to internal fuel types (Supplementary Table S3) and used to estimate CO₂ emissions from heat. Data from NSI, NEI Nonpoint, and EIA SEDS is then reconciled to ensure consistency. Building-level CO₂ emissions are downscaled using county-level NEI CO₂ totals, proportional to each building’s square footage. If county emissions are zero or significantly different from expected values (40x lower or 1.5x higher than state totals), emissions are backfilled using state-level EIA SEDS estimates, scaled by building square footage. To refine backfilled estimates, an adjustment factor based on the ratio of stationary point fuel consumption to EIA state-level fuel consumption is applied, ensuring emissions align with expected fuel use while preserving self-reported NEI data. Onroad Emissions from the onroad sector use a novel dataset - a GPS-based traffic dataset covering every road segment in the contiguous US - to estimate emissions. This sector is defined as segment-level CO 2 emissions representing non-electric vehicles driving on roadways. Emissions from other off-highway vehicles are not included in this sector (e.g. construction equipment, agriculture/forestry vehicles, golf carts, etc.). The onroad emissions module provides a comprehensive framework for calculating on-road vehicle miles traveled (VMT) and associated fuel consumption and CO₂ emissions, integrating multiple datasets across various spatial scales. The onroad modeling process integrates multiple sources of vehicle miles traveled (VMT) data and calculates the associated fleet mix, fuel consumption and CO 2 emissions. For CONUS, VMT is sourced from GPS-based traffic data for 2023 and is merged with fleet mix information (FHWA Highway Statistics VM1, VM4, and VM2 tables) and the fuel economies (Transportation Energy Data Book), aligning on common dimensions such as year, state, road class, urban/rural land class, and vehicle class. A separate approach is used for Alaska and Hawaii, where VMT is calculated using FHWA HPMS AADT shapefiles for 2022 and aggregated at the state level. Discrepancies between measured HPMS VMT and state-level totals are allocated across road segments based on length, ensuring accurate distribution. Vehicle-type fractions are applied to derive specific VMT estimates, with gaps in truck data filled using HPMS sources. For years before and after the years corresponding to the VMT data (2023 for the contiguous US and 2022 for Alaska and Hawaii), VMT was estimated using a regression model that was trained on 2023 county-level VMT for the CONUS and 2022 VMT for Alaska and Hawaii using road length and population as predictors. This model estimates past VMT, distributing values across road segments proportionally to 2023 (2022 for Alaska and Hawaii) county-level VMT. Estimates are constrained using state-level FHWA VMT and SEDS/EIA fuel data. CO₂ emissions are then computed using fleet mix, fuel economy, and year-specific fuel consumption, following the same methodology as 2023 (2022 for Alaska and Hawaii). Hoteling Hoteling emissions are created by long-haul combination truck vehicles idling during mandated rest times for periods of longer than one hour duration. This definition follows that of NEI. Hoteling hours for all four NEI years were read in as a long table and combined with data from the FHWA Highway Statistics Series (VM1 and VM4 tables) and the Transportation Energy Data Book (Tables A.1, A.3, and A.6). The EPA surrogate shapefile for hoteling locations was also incorporated. NEI and FHWA vehicle miles traveled (VMT) data were merged using state FIPS codes, and annual hoteling hours were interpolated for years between or beyond NEI years using VMT data for combined trucks. The interpolated NEI data were spatially joined with EPA hoteling location shapes, and the annual number of hoteling hours at each location was calculated as: $$\\:{hours}_{i}\\:=\\:{hours}_{county}\\times\\:\\frac{{p}_{i}}{{p}_{county}}$$ 2 where hours i represents the hoteling hours at a specific location, hours county is the total county-wide hoteling hours, p i is the number of parking spaces at the location, and p county is the total parking spaces in the county. Finally, CO₂ emissions at each location were estimated using an emission factor and fuel consumption rate, calculated as: $$\\:{CO2}_{i}\\:=\\:{hours}_{i}\\times\\:\\frac{0.71\\:gallons}{hour}\\times\\:\\frac{10.21\\:kg\\:CO2}{gallon}$$ 3 EGU The Energy Generating Unit (EGU) sector is defined as point-level emissions representing emissions from EGUs, aka power plants, associated with Scope 1 electricity generation using data from the U.S. Energy Information Administration (EIA) and Clean Air Markets (CAM). The EGU module focuses on integrating emissions and facility data from the EIA 923 and EIA 860 datasets with CAM emissions data to create a comprehensive and harmonized dataset for CO₂ emissions analysis. The methodology involves several key steps to ensure accurate geospatial tagging, data consolidation, and emissions estimation. First, the EIA 923 and EIA 860 datasets are merged to tag each facility with latitude and longitude coordinates, enabling spatialization. Any missing geolocation data is supplemented using an external reference table to ensure all facilities are mapped accurately. CAM emissions data, which includes Department of Energy fuel codes, are then aligned with Supplementary Table S1 fuel IDs using a standardized crosswalk, ensuring consistency in fuel classifications across datasets. Both EIA and CAM datasets provide heat output data, which is used to estimate CO₂ emissions by applying emission factors specific to each fuel type, measured in metric tons of CO₂ per million BTUs (tCO₂/mmBTU). This approach ensures consistency in emissions estimation across the datasets while leveraging the detailed heat output information available in the source data. The EIA and CAM datasets are aggregated by facility, fuel type, and year to create a unified view of emissions at the facility level. These aggregated datasets are then merged using a facility/unit ID crosswalk, which resolves discrepancies between the two sources. Data from the EIA is prioritized over CAM where overlaps occur, as the EIA datasets provide a more comprehensive list of facilities and corresponding data. This ensures the resulting dataset is both complete and accurate. The final dataset provides a comprehensive and geospatially tagged inventory of CO₂ emissions at the facility level, categorized by fuel type and year. For edge cases, the crosswalk process aims to reconcile duplicate facilities from CAM and EIA using facility/unit IDs. Some CAM facility/unit IDs repeat across multiple generator matches in EIA, leading to situations where EIA boiler IDs are either empty or repeated. To address this, rows with empty EIA boiler IDs are filtered out first, followed by a deduplication process to resolve one-to-many mappings. This manual adjustment ensures consistency and avoids data duplication. Stationary Point The stationary point sector is defined as the facilities that emit greater than 25,000 metric tons CO 2 e per year (i.e., large CO 2 emitting facilities), which are required to report emissions to the EPA GHGRP. These facilities encompass the direct emissions from fossil fuel combustion and process emissions from the activities defined in Envirofacts. For the stationary point module, self-reported, facility-level CO 2 emissions are aggregated to consistent temporal and categorical dimensions, using data from the annual GHGRP, subpart GHGRP, and Envirofacts. Since the three input datasets—GHGRP Annual, Envirofacts, and GHGRP Subparts—report information for an overlapping set of facility ids, a priority order is applied to resolve discrepancies and duplicated data and retain the most reliable data. Specifically, GHGRP Annual data is prioritized over Envirofacts data, which is further prioritized over GHGRP Subparts data. Additionally, where overlapping facilities are identified, Envirofacts fuel information is appended to enrich the dataset with detailed fuel-related attributes. The data is further cleaned by excluding subpart D (electricity production) and by excluding basin-level oil and gas production emissions in subpart W, as these emissions are covered in other sectors in this model. Oil and Gas The Oil and Gas emissions sector is defined as a subset of the NEI nonpoint emissions that relate to oil and gas production. SCCs for this sector are matched using the SCC level two descriptions matching “Oil and Gas production”. To estimate CO₂ emissions from oil and gas production, the NEI Nonpoint datasets are processed through a series of filtering and calculation steps. First, the dataset was filtered to include only records that meet the definition, ensuring that only relevant nonpoint sources were included in the analysis. From the filtered data, CO emissions are isolated, and CO₂ emissions were calculated using the emissions factors specified in Supplementary Table S2. Finally, the emissions are aggregated by fuel ID, county, and state to create a spatially resolved inventory. Emissions are spatialized by identifying oil and gas production wells outlined in the USGS oil and gas production dataset 15 , placing emissions based on production at the well site. Airport The airport sector contains emissions for all aircraft (planes and helicopters) during the landing/take-off (LTO) cycle. This sector models CO 2 emissions by using data on the number of aircraft and aircraft/engine specific emission factors. To estimate CO₂ emissions using the airport module, there are several steps of data filtering, aggregation, and emissions calculations using both FAA 5010 and TFMSC datasets. First, the FAA 5010 dataset is filtered to include only data for years when airports were operational. The resulting data is aggregated by airport, user class, and year, with user class grouping following the rules outlined in Supplementary Table S4, which describes how native FAA 5010 LTO columns are pooled into general user classes. Similarly, the TFMSC dataset is aggregated by airport, user class, engine type, and aircraft type, with user class and engine type mappings defined in Supplementary Table S5 and Supplementary Table S6 respectively. As both FAA 5010 and TFMSC datasets report data for overlapping airports and user classes, FAA 5010 data is prioritized for these cases. The assumption is made that all FAA 5010 reported LTOs are for propeller-type aircraft, as the dataset does not include commercial operations. For non-overlapping TFMSC data, FAA 5010 location data is used to gapfill missing location information, and any remaining TFMSC data lacking location details is dropped, as it primarily represents foreign airports. Emission factors are then applied to calculate CO₂ emissions from LTO operations, with IPCC aircraft emission factors (Supplementary Table S7) being prioritized over EPA engine type emission factors (Supplementary Table S8). In cases where engine type or aircraft type is not reported, default user-class emission factors from the EPA (Supplementary Table S9) are used for CO₂ calculations. Mobile The mobile emissions sector is defined as a subset of the NEI nonpoint, point, and nonroad emissions that relate to Railroad and Shipping activity. SCCs for this sector are matched using the SCC level two descriptions shown in Supplementary Table S10. This module processes the NEI nonpoint, point, and nonroad datasets to estimate CO₂ emissions by applying specific filtering criteria, calculation methods, and aggregation steps. First, the NEI datasets are filtered to include only data meeting the definitions outlined in Supplementary Table S10. From these filtered datasets, CO emissions are isolated, and CO₂ estimates were calculated by applying emissions factors using Eq. 1. Where available, SCC-specific emissions factors were applied in place of default emissions factors to improve accuracy. The resulting emissions data were then aggregated by fuel ID, state, county, and mobile emissions type (e.g., \"Railroad\" or \"Marine vessels\") to create a comprehensive and spatially resolved inventory of CO₂ emissions from these source categories. These emissions are then spatially allocated to shipping lanes, ports, railroad segments, or railyards by putting the relevant county-level emission to the target spatial source. Nonroad Equipment The Nonroad sector is defined as off-road mobile sources that use gasoline, diesel, and other fuels (including agriculture equipment). We integrate the nonroad definition of nonroad emissions, which report these emissions at the county-level, though our definition excludes airplanes, marine and rail locomotives. This module processes the NEI Nonroad dataset to estimate CO₂ emissions by applying specific filtering criteria, calculation methods, and aggregation steps. The dataset is first filtered to include only data that meets the definitions outlined in (Supplementary Table S11), ensuring the analysis focuses on relevant nonroad sources. From this filtered data, CO emissions are extracted and converted to CO₂ estimates using emissions factors specific to fuel types, as detailed in Eq. 2 . Finally, the emissions are aggregated by fuel, state, and county, resulting in a structured dataset that offers a detailed spatial breakdown of CO₂ emissions. Uncertainty EPA emission factors from AP-40 and the underlying activity data (NEI emissions and EIA fuel consumption) do not report meaningful uncertainty values. As such, we follow the IPCC recommendation of using a Monte Carlo analysis to identify uncertainty 16 . Several studies have applied this Monte Carlo approach to quantify uncertainties in national or regional emissions of CO2 and other pollutants 17 – 19 . The method has been used for emissions from various sectors, including power plants, transportation, industrial processes, and oil and gas 20 – 23 . In this model, we use the Monte Carlo approach to generate multiple random iterations of the model, where input variables (in this case, emission factors and activity data) are sampled from probability distributions that reflect their potential variability. We group the emission factors into their economic sectors and fuel types to develop a probability distribution for these factors when the EPA AP-40 does not include additional information. We do not use the distribution of other input activity, e.g., NEI inputs, self-reported emissions to the EPA, or fuel consumption estimates from the EIA because these datasets do not report confidence intervals on their outputs and developing meaningful distributions of the data are obfuscated. The uncertainty is limited to the modeling process, such that no additional uncertainty analysis is performed to estimate uncertainty due to geospatial processing. Results Crosswalk ffCO 2 Estimates by Sector and Year Nationally, direct emissions of fossil fuel CO 2 have decreased by 896 MMT (13%) from 2010 to 2024. Until 2023, the highest emitting sector contributing to national CO 2 emissions came from electric power generation (1,800–2,500 MMT CO 2 , corresponding to 32–40% of national emissions). After 2023, the transportation sector overtook power generation (1,760 MMT CO 2 vs 1,670 MMTCO 2 ), though both sectors contribute to roughly 33% of total emissions. Annual sector totals are outlined in Table 1 and shown in Fig. 2. Table 1. CO 2 estimates over the US by sector for Crosswalk (units in MMT CO2). Year Airport Commercial buildings Industrial Nonroad Other Powerplants Residential Vehicles Total 2010 51 161 1362 125 22 2550 352 1541 6163 2011 50 161 1401 127 20 2451 350 1520 6082 2012 50 163 1390 130 26 2326 350 1504 5938 2013 50 166 1393 132 23 2346 352 1534 5996 2014 50 172 1410 134 23 2347 353 1558 6048 2015 50 176 1395 135 23 2195 344 1563 5880 2016 51 179 1367 135 23 2109 334 1582 5781 2017 52 183 1374 136 23 2026 339 1585 5717 2018 53 186 1414 136 23 2049 346 1609 5815 2019 54 188 1405 135 22 1889 349 1614 5656 2020 35 189 1355 133 22 1717 346 1433 5229 2021 44 191 1374 131 22 1812 344 1541 5460 2022 49 193 1369 130 22 1795 343 1542 5442 2023 52 194 1354 127 22 1666 344 1571 5330 2024 53 188 1316 129 22 1667 326 1566 5267 Table 2. CO 2 estimates for 2024 for 5 highest states (units in MMT CO 2 ). Year Airport Commercial buildings Industrial Nonroad Other Powerplants Residential Vehicles Total Texas 5 9 328 6 2 216 12 183 762 California 5 22 74 55 2 48 24 145 376 Florida 5 5 24 6 0 101 1 93 236 Louisiana 0 1 140 1 1 54 2 34 232 Pennsylvania 1 10 42 4 1 78 18 53 205 The states with the largest absolute Scope 1 emissions are Texas, California, and Florida, Louisiana, and Pennsylvania (Table 2). Texas and Louisiana have the largest industrial CO 2 emissions while California has the largest portion of emissions from vehicles, and Florida and Pennsylvania have the largest proportion of emissions from power plants. Comparison of CO 2 Product National Emissions Inventories As shown in Fig. 3, we compare the Crosswalk data product output to published national inventories. The datasets used for comparison are the EPA Greenhouse Gas Inventory (GHGI), the EPA state implementation tool (SIT), CarbonTracker v2022 product 24 , Vulcan 5 , and ClimateTrace 25 . To maintain consistency, testing datasets were cropped to include only fossil fuel emissions over the continental US, limiting the analysis to this specific emission category and spatial extent. At the national level, the Crosswalk CO 2 product agrees best with the EPA GHGI, with the lowest differences (10% on average). Crosswalk shows similar year-over-year trends as the EPA, Vulcan, and Carbon tracker datasets (r > 0.99), with the lowest agreement with ClimateTrace, though the total amount of CO 2 converges in the most recent years (Fig. 3). This high level of agreement with established inventories demonstrates the commonality in methodology and/or shared data sources between the inventories and the Crosswalk CO 2 product, while the discrepancies with ClimateTrace highlight the differences in methodology and data sources, particularly in earlier years. These variations underscore the importance of continued cross-comparison and validation to improve consistency across emissions datasets and provide a more comprehensive understanding of national CO 2 trends. Gridded Emissions Comparison Comparison of spatial emissions patterns between the Crosswalk and CarbonTracker v2022 products reveals a strong overall correlation (r = 0.61), indicating broad agreement in the spatial distribution of emissions. Notably, both products show that the eastern United States has substantially higher fossil fuel CO 2 emissions—over 250% greater—than the western United States (Fig. 4a,b). However, some regional differences persist. In total, Crosswalk estimates tend to be higher than those from CarbonTracker, though spatially there are some areas in the Western US which are lower in Crosswalk relative to CarbonTracker (Fig. 4c). These discrepancies may stem from differences in input data, sectoral assumptions, or spatial allocation methods. The CarbonTracker product is an assimilated modeling system that integrates atmospheric CO 2 observations with surface flux estimates to produce more accurate, observation-constrained emissions. By assimilating global atmospheric measurements, CarbonTracker adjusts emission fluxes to reflect atmospheric observations, with the v2022 product differentiating ffCO 2 and biogenic CO 2 with priors as opposed to observations. For spatial allocation of emissions, CarbonTracker employs nighttime lights as a proxy for human activity, enabling a finer resolution depiction of where anthropogenic CO 2 emissions are likely to originate. This contrasts to Crosswalk emissions, which use both measured fluxes allocated at the source and bottom-up emissions estimates which are spatially allocated to specific emitting shapes (e.g., roadways, oil and gas wells, buildings). As shown in Fig. 3, CarbonTracker is significantly lower than Crosswalk and the published EPA bottom-up emissions inventories (SIT, GHGI), which highlight the discrepancies between bottom-up and top-down emissions estimates and different approaches and assumptions to back out anthropogenic emissions. Discussion This study builds off existing research into anthropogenic-based CO 2 emissions and integrates updates to better address the limitations of current CO 2 emissions models. These advancements allow for a more comprehensive and flexible modeling framework capable of addressing both historical trends and future projections of CO₂ emissions. By combining high-resolution activity data with cutting-edge computational infrastructure, this study aims to improve spatial granularity and consistent releases, enabling decision-makers to better understand localized emissions patterns and their drivers. First, we integrate public and private activity datasets to best capture historical emissions (e.g., airplane types and counts from FAA, GPS-based traffic data, and self-reported quarterly emissions for large point sources from the Clean Air Markets Database (CAMD)). Second, we integrate this modeling framework into a cloud-native environment to produce reliable and consistent outputs. Finally, we integrate energy fuel forecasts to allow for emissions to be modeled to near-real time, and out to 2050. The integrated approach presented herein not only addresses existing gaps in emissions modeling, but also establishes a robust foundation for dynamic and data-driven climate action strategies. We provide data aggregated to the neighborhood scale and aggregated for all US cities with populations greater than 5,000 people. This makes our product particularly useful for smaller jurisdictions that might not otherwise have access to their own fossil fuel CO₂ inventory, and might be required to rely on or extrapolate from data from the larger metro area or county. Another strength of our approach is that by using a cloud-native environment we are able to automatically ingest and utilize data when updates or new source data come online. Integration of energy fuel forecasts makes our model particularly conducive to forecasting future emissions under varying scenarios. A key area for future work is characterization of uncertainty. The largest source of uncertainty in the current methodology arises from the nonpoint buildings sector. The reliance on downscaling county-level emissions from the NEI introduces potential inaccuracies due to variations in building-level energy consumption that are not fully captured in aggregate data. A more robust approach would involve integrating direct utility billing data or high-resolution building energy models to better constrain emissions estimates. Access to improved fuel use data, particularly for heating and industrial processes, would enhance the spatial and temporal resolution of emissions estimates. Additionally, incorporating uncertainty estimates from the NEI would help quantify and communicate the confidence level in the final emissions inventory, providing a clearer picture of potential variability. Another key area for future work is the integration of top-down and bottom-up emissions estimation methods to improve inventory accuracy and facilitate better atmospheric modeling. Remote sensing data, atmospheric inversion models, and satellite-based CO₂ measurements could complement ground-based inventories, helping to reconcile discrepancies between modeled and observed emissions. Developing a framework for systematically merging these approaches would lead to a more representative emissions dataset, improving the accuracy of emissions inventories used in climate modeling, air quality analysis, and policy decision-making. Further refinement of temporal emissions profiles, incorporating real-time activity data where possible, would also enhance the responsiveness and applicability of the dataset for near-term emissions mitigation efforts. Declarations Data Availability The datasets generated and analysed during the current study are available in this Zenodo repository, https://zenodo.org/records/16986970?preview=1&token=eyJhbGciOiJIUzUxMiIsImlhdCI6MTc1NjM 5ODMzMiwiZXhwIjoxNzU5Mjc2Nzk5fQ.eyJpZCI6IjIyOTc3YjgxLWU4ZWEtNGY3MC1hMGQ0LTQ0ZWIxZTMwODQ5NCIsImRh dGEiOnt9LCJyYW5kb20iO iJjNWZiMDZkMmNjZmNjNzQ0Nzc1ZmFlZDA2ODg zZDFlYyJ9.wS9S1AUcWpAaTTAYvwVmc2jWEo9K1QauOB4-yWeKgZrZlz5vXmViASMdO2yOcwZ2UEznhYTe2l-FlfsIcVAzGg which can be cited as https://doi.org/10.5281/zenodo.16986970. Funding This research was supported by Crosswalk Labs. Acknowledgements The Data Foundation provided funding to publish results from this research in an open data format. References Friedlingstein, P. et al. Global Carbon Budget 2022. Earth Syst. Sci. Data 14 , 4811–4900 (2022). Zhang, W., Huang, Y. & Wu, H. The symmetric and asymmetric effects of economic policy uncertainty and oil prices on carbon emissions in the USA and China: evidence from the ARDL and non-linear ARDL approaches. Environ. Sci. Pollut. Res. 29 , 26465–26482 (2022). Kauffman, J. B., Beschta, R. L., Lacy, P. M. & Liverman, M. Livestock Use on Public Lands in the Western USA Exacerbates Climate Change: Implications for Climate Change Mitigation and Adaptation. Environ. Manage. 69 , 1137–1152 (2022). Seto, K. C., Güneralp, B. & Hutyra, L. R. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proc. Natl. Acad. Sci. 109 , 16083–16088 (2012). Gurney, K. R. et al. The Vulcan Version 3.0 High‐Resolution Fossil Fuel CO 2 Emissions for the United States. J. Geophys. Res. Atmospheres 125 , (2020). Tao, J. & Kong, X. Spatial allocation of anthropogenic carbon dioxide emission statistics data fusing multi-source data based on Bayesian network. Sci. Rep. 11 , 18128 (2021). Oda, T., Maksyutov, S. & Andres, R. J. The Open-source Data Inventory for Anthropogenic CO 2 , version 2016 (ODIAC2016): a global monthly fossil fuel CO 2 ; gridded emissions data product for tracer transport simulations and surface flux inversions. Earth Syst. Sci. Data 10 , 87–107 (2018). Asner, G. P. et al. Targeted carbon conservation at national scales with high-resolution monitoring. Proc. Natl. Acad. Sci. 111 , (2014). Chen, P. The impact of low-carbon city pilots on carbon emission – new evidence from high-resolution carbon emission data. Kybernetes 52 , 5875–5892 (2023). Feng, S. et al. Los Angeles megacity: a high-resolution land–atmosphere modelling system for urban CO 2 emissions. Atmospheric Chem. Phys. 16 , 9019–9045 (2016). Adams, S., Adedoyin, F., Olaniran, E. & Bekun, F. V. Energy consumption, economic policy uncertainty and carbon emissions; causality evidence from resource rich economies. Econ. Anal. Policy 68 , 179–190 (2020). Sargent, M. R. et al. Majority of US urban natural gas emissions unaccounted for in inventories. Proc. Natl. Acad. Sci. 118 , e2105804118 (2021). Li, T. et al. Reconstructing Global Daily CO2 Emissions via Machine Learning. Preprint at https://doi.org/10.48550/arXiv.2407.20057 (2024). Gately, C. K. & Hutyra, L. R. Large Uncertainties in Urban‐Scale Carbon Emissions. J. Geophys. Res. Atmospheres 122 , (2017). Skinner, C. C. et al. Aggregated Oil and Natural Gas Drilling and Production History of the United States (ver. 1.1, April 2023). U.S. Geological Survey https://doi.org/10.5066/P9UIR5HE (2022). Mastrandrea, M. D. et al. The IPCC AR5 guidance note on consistent treatment of uncertainties: a common approach across the working groups. Clim. Change 108 , 675–691 (2011). Wu, Y. et al. Methods to account for uncertainties in exposure assessment in studies of environmental exposures. Environ. Health 18 , 31 (2019). Super, I., Dellaert, S. N. C., Visschedijk, A. J. H. & Denier Van Der Gon, H. A. C. Uncertainty analysis of a European high-resolution emission inventory of CO 2 and CO to support inverse modelling and network design. Atmospheric Chem. Phys. 20 , 1795–1816 (2020). Zhao, Y., Nielsen, C. P., Lei, Y., McElroy, M. B. & Hao, J. Quantifying the uncertainties of a bottom-up emission inventory of anthropogenic atmospheric pollutants in China. Atmospheric Chem. Phys. 11 , 2295–2308 (2011). Tchepel, O. et al. EMISSION MODELLING OF HAZARDOUS AIR POLLUTANTS FROM ROAD TRANSPORT AT URBAN SCALE. TRANSPORT 27 , 299–306 (2012). Roest, G. & Schade, G. Quantifying alkane emissions in the Eagle Ford Shale using boundary layer enhancement. Atmospheric Chem. Phys. 17 , 11163–11176 (2017). Gurney, K. R. et al. Quantification of Fossil Fuel CO 2 Emissions on the Building/Street Scale for a Large U.S. City. Environ. Sci. Technol. 46 , 12194–12202 (2012). Liu, Z. et al. Near-real-time monitoring of global CO2 emissions reveals the effects of the COVID-19 pandemic. Nat. Commun. 11 , 5172 (2020). Peters, W. et al. An atmospheric perspective on North American carbon dioxide exchange: CarbonTracker. Proc. Natl. Acad. Sci. 104 , 18925–18930 (2007). Gore, A. Measure emissions to manage emissions. Science 378 , 455–455 (2022). Additional Declarations No competing interests reported. Supplementary Files supplemental1.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-7456083\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Article\",\"associatedPublications\":[],\"authors\":[{\"id\":513873036,\"identity\":\"602a5a3f-e022-47e9-848e-2fea73b10722\",\"order_by\":0,\"name\":\"Anastasia Montgomery\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1ElEQVRIiWNgGAWjYNCCChijIIGBjaBqsIozMJ4BsVoY25C0EATy85sffuadZ5fPz3/8mcQPgzQGPvYG/FoMjrEZS/NuS7acOSPHTLLHIIeBjecAAS1sDGbMvNsOGBjc4GG7wWNQwcAmQcBx8m3s35h55wC1nD/+7OYfYrQwHOMB2tIA1HIgwew2D8hhhLQYHMsplpxzLNlAckaO+W8ZgzQegn6Rbz6+8cObGjsDYIg9NnxTkSwn395AwGFAwMSDxOHBqQwZMP4gStkoGAWjYBSMWAAArQI5yc/JOkUAAAAASUVORK5CYII=\",\"orcid\":\"\",\"institution\":\"Crosswalk Labs\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Anastasia\",\"middleName\":\"\",\"lastName\":\"Montgomery\",\"suffix\":\"\"},{\"id\":513873037,\"identity\":\"9812060d-59d5-4d74-9c7e-ddd2e818abf0\",\"order_by\":1,\"name\":\"Geoff Roest\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Crosswalk Labs\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Geoff\",\"middleName\":\"\",\"lastName\":\"Roest\",\"suffix\":\"\"},{\"id\":513873038,\"identity\":\"6de23fb5-0af9-42df-b138-9395eb2f7459\",\"order_by\":2,\"name\":\"Jason Zou\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Crosswalk Labs\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Jason\",\"middleName\":\"\",\"lastName\":\"Zou\",\"suffix\":\"\"},{\"id\":513873039,\"identity\":\"337f5658-137b-467c-ae7c-62235b8207b1\",\"order_by\":3,\"name\":\"Erik Badger\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Crosswalk Labs\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Erik\",\"middleName\":\"\",\"lastName\":\"Badger\",\"suffix\":\"\"},{\"id\":513873040,\"identity\":\"9f522040-d5c1-4792-aee4-ade087c1a767\",\"order_by\":4,\"name\":\"Phil DeCola\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Crosswalk Labs\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Phil\",\"middleName\":\"\",\"lastName\":\"DeCola\",\"suffix\":\"\"},{\"id\":513873041,\"identity\":\"1a3d9ab4-d336-4ccc-866e-eb4b257b7dd5\",\"order_by\":5,\"name\":\"Victoria Hunt\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Crosswalk Labs\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Victoria\",\"middleName\":\"\",\"lastName\":\"Hunt\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-08-25 17:38:16\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-7456083/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-7456083/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":91465033,\"identity\":\"13c2fb98-5861-493c-8142-3b543b5b9a43\",\"added_by\":\"auto\",\"created_at\":\"2025-09-16 18:27:32\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":389276,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eCrosswalk CO\\u003csub\\u003e2\\u003c/sub\\u003e total emissions over the contiguous US, with the entire domain including all 50 US states and excluding US territories.\\u0026nbsp;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7456083/v1/6b9978c3a695237c69288f69.png\"},{\"id\":91466617,\"identity\":\"de0b1ed6-5a75-4421-88da-20c4bc932a46\",\"added_by\":\"auto\",\"created_at\":\"2025-09-16 18:51:32\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":243960,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eStacked barplot of CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions from 2010 - 2024, with each color corresponding to an emissions sector.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7456083/v1/bf372ef543ae53bf45b269e8.png\"},{\"id\":91465373,\"identity\":\"728f9d33-ef32-4622-9b93-4269b382a774\",\"added_by\":\"auto\",\"created_at\":\"2025-09-16 18:35:32\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":195823,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eAnnual ffCO\\u003csub\\u003e2\\u003c/sub\\u003e emissions estimates over the continental United States, comparing the Crosswalk product to federal (EPA GHGI and SIT) and independent (Vulcan, Climate Trace, CarbonTracker) datasets.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7456083/v1/95c8f5f5bc5f4b1a6a6d36b5.png\"},{\"id\":91465861,\"identity\":\"b01a3804-9859-4518-9a58-ce3208cc2030\",\"added_by\":\"auto\",\"created_at\":\"2025-09-16 18:43:32\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":288214,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003effCO\\u003csub\\u003e2\\u003c/sub\\u003e emissions estimates over the continental United States, comparing the (a) CarbonTracker v2022 product to the (b) Crosswalk ffCO2 output, and the (c) difference. Data in this figure corresponds to 2020 emissions, the most recent year available for CarbonTracker. The 100° W is used to define the eastern and western US divide. Note the colorbar represents a log scale.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7456083/v1/be6eb904cb7753a1eedbfbd2.png\"},{\"id\":94464537,\"identity\":\"dd0f3a7c-4146-4aed-b7a4-b299965b5334\",\"added_by\":\"auto\",\"created_at\":\"2025-10-27 15:10:56\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":1792160,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7456083/v1/10100d79-4dbd-46e1-846f-d1b8711352d4.pdf\"},{\"id\":91465032,\"identity\":\"ee1918d6-ad8a-4e85-a84e-17c1236dcf24\",\"added_by\":\"auto\",\"created_at\":\"2025-09-16 18:27:32\",\"extension\":\"docx\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":20780,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"supplemental1.docx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7456083/v1/235817f994365d0871bd2dd6.docx\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"Modernizing carbon dioxide emissions inventories for action in the United States\",\"fulltext\":[{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eHigh-resolution carbon emissions data are critical for effective climate change mitigation strategies. Numerous agencies have called for significant reductions in greenhouse gases to limit global temperature rise\\u003csup\\u003e\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e\\u003c/sup\\u003e. Accurate emissions data enable policymakers to identify specific sources of emissions, assess the effectiveness of existing regulations, and develop targeted interventions to reduce greenhouse gas emissions. Carbon emissions models are integral to this process, allowing for the investigation of improved stewardship of public lands\\u003csup\\u003e\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e\\u003c/sup\\u003e, investigation of local and regional carbon dynamics\\u003csup\\u003e\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e\\u003c/sup\\u003e, and characterization of the effects of urbanization on carbon emissions\\u003csup\\u003e\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e\\u003c/sup\\u003e. Thus, the ability to map emissions at finer spatial resolutions is not merely a technical improvement; it is a fundamental requirement for effective climate action.\\u003c/p\\u003e\\u003cp\\u003eNumerous CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions models have been developed to assess and predict emissions in the U.S., each employing different methodologies and data sources. Models integrating a bottom-up approach use activity data and emission factors to estimate CO\\u003csub\\u003e2\\u003c/sub\\u003e at various resolutions\\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR6\\\" citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e–\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u003c/sup\\u003e. Other CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions models use direct measurements or satellite observations and use atmospheric inversion to get atmospherically-derived fluxes of CO\\u003csub\\u003e2\\u003c/sub\\u003e\\u003csup\\u003e8\\u003c/sup\\u003e. The diversity of these models underscores the necessity for a multi-faceted approach to emissions assessment as different models provide complementary insights into the complexities of carbon emissions in the U.S.\\u003c/p\\u003e\\u003cp\\u003eHowever, current CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions modeling frameworks in the U.S. require updates to improve spatial resolution and representativeness, and to be made readily accessible for a wide audience of stakeholders. The integration of high-resolution data is increasingly recognized as vital for understanding localized emissions patterns and their drivers\\u003csup\\u003e\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e\\u003c/sup\\u003e. By relying on coarse spatial resolutions, critical variations in emissions can be obscured across different urban landscapes\\u003csup\\u003e\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e. Furthermore, as the scientific community calls for more robust data to inform climate policy, the need for models that can incorporate real-time data and account for uncertainties in emissions estimates has become apparent\\u003csup\\u003e\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e\\u003c/sup\\u003e. This evolution in modeling approaches is essential not only for compliance with regulatory frameworks, but also for fostering public trust in climate action initiatives.\\u003c/p\\u003e\\u003cp\\u003eDespite significant progress in CO₂ emissions modeling, critical gaps remain in harmonizing diverse data sources to develop near-real-time estimates with a full spectrum of emission sources. Many current models are constrained by limitations in data availability or consistency, particularly when integrating disparate datasets with different data publication schedules, such as federal inventories, satellite observations, and activity-based emissions estimates.\\u003c/p\\u003e\\u003cp\\u003eFurther, integrating diverse data sources requires harmonization due to inherent differences in methodologies; e.g., many atmospherically-derived CO₂ emissions datasets have shown that usage-based inventories are missing emissions from natural gas\\u003csup\\u003e\\u003cspan citationid=\\\"CR12\\\" class=\\\"CitationRef\\\"\\u003e12\\u003c/span\\u003e\\u003c/sup\\u003e. These challenges are compounded by the need to address complex sectors such as the land-use sector, which can significantly impact regional emissions profiles. While advancements in machine learning and data assimilation techniques offer promising pathways for overcoming these challenges, their application to emissions modeling is still in active development\\u003csup\\u003e\\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e\\u003c/sup\\u003e. Bridging these gaps will require a concerted effort to leverage emerging technologies and develop models that are both flexible and transparent. These enhancements are not only essential for more precise emissions assessments but also for aligning modeling frameworks with the evolving needs of policymakers, researchers, and the broader public.\\u003c/p\\u003e\\u003cp\\u003eIn this paper, we will describe the methodology that the Crosswalk emissions model implements to create annual CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions, review novel data sources and methods including the use of GPS-derived traffic data for onroad vehicle emissions, describe the output from the Crosswalk CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions model, and finally compare the emissions output with other independent CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions inventories.\\u003c/p\\u003e\\n\\n\"},{\"header\":\"Methods\",\"content\":\"\\u003cp\\u003eModel Concept\\u003c/p\\u003e\\n\\u003cp\\u003eWe present a CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions inventory for the year 2010\\u0026ndash;2024, gridded to a 1 km spatial resolution. We integrate the sectoral definitions set by the National Emissions Inventory, with activities based on the EPA\\u0026rsquo;s Source Classification Codes (SCC) and other independent datasets (e.g., GPS-based traffic data). As such, this inventory follows a similar sectoral aggregation to Vulcan\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e\\u003c/sup\\u003e and the Anthropogenic Carbon Emissions System\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e\\u003c/sup\\u003e. Our domain covers the entire United States excluding US territories (Fig. \\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e).\\u003c/p\\u003e\\n\\u003cp\\u003eThe Crosswalk emissions model is built in Python with a separate module for each activity sector. Input data from public sources is accessed via APIs, webscraping, or downloading directly from an FTP server, then copied to Amazon S3 for traceable usage within the model. In the case of onroad emissions, private data sources are used as described in that subsequent section. Files are stored as Parquet files for efficient handling of larger datasets. The data are transformed to have a standardized schema so that disparate datasets can be merged. The model routine takes one or more data sources, applies emission factors, and applies a logic hierarchy to select the final emission calculation. The spatialization routine takes the calculated emissions and geolocates the emissions output. The spatial surrogates for the model output are customized for each sector, as described in subsequent sector-specific sections. The raw model output is then available in point, line, and polygon shape formats which then can be gridded. The output is also written as GeoParquet files using the newer GeoParquet 1.1 specification, which allows for more efficient filtering of data during downstream geo-processing routines.\\u003c/p\\u003e\\n\\u003cp\\u003eUsing the development framework presented herein, the Crosswalk emissions model can easily digest multiple data sources while still maintaining a consistent output format. Further, this framework allows for quarterly updating, utilizing the latest available datasets and interpolating data where older datasets have more time lag. To generate the model output, the model is run distributed on a cluster of Amazon EC2 instances using AWS Batch which takes approximately five days real-world runtime for end-to-end national output for 14 years of data. The software is version controlled using Github, and we maintain a consistent execution environment across runs via Docker containers.\\u003c/p\\u003e\\n\\u003cp\\u003eSector Modeling\\u003c/p\\u003e\\n\\u003cp\\u003eBuildings\\u003c/p\\u003e\\n\\u003cp\\u003eThe building emissions sector is defined as a subset and downscaled version of the NEI nonpoint emissions. The NEI nonpoint emissions represent the large number of diffuse, small (\\u0026lt;\\u0026thinsp;25 MMT CO\\u003csub\\u003e2\\u003c/sub\\u003ee) carbon emitters associated with industrial, commercial, and residential activity, such as commercial cooking emissions. We define the industrial, residential, and commercial buildings based on key words in the level-2 SCC definitions, as defined in Supplementary Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eTo estimate building-level CO₂ emissions, the module begins by filtering NEI Nonpoint data to include only relevant sources and applying emission factors (Supplementary Table S2) to convert CO emissions to CO₂ (Eq.\\u0026nbsp;1). EIA SEDS fuel data is mapped to internal fuel types (Supplementary Table S3) and used to estimate CO₂ emissions from heat. Data from NSI, NEI Nonpoint, and EIA SEDS is then reconciled to ensure consistency. Building-level CO₂ emissions are downscaled using county-level NEI CO₂ totals, proportional to each building\\u0026rsquo;s square footage. If county emissions are zero or significantly different from expected values (40x lower or 1.5x higher than state totals), emissions are backfilled using state-level EIA SEDS estimates, scaled by building square footage. To refine backfilled estimates, an adjustment factor based on the ratio of stationary point fuel consumption to EIA state-level fuel consumption is applied, ensuring emissions align with expected fuel use while preserving self-reported NEI data.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cimg src=\\\"data:image/png;base64,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\\\"\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eOnroad\\u003c/p\\u003e\\n\\u003cp\\u003eEmissions from the onroad sector use a novel dataset - a GPS-based traffic dataset covering every road segment in the contiguous US - to estimate emissions. This sector is defined as segment-level CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions representing non-electric vehicles driving on roadways. Emissions from other off-highway vehicles are not included in this sector (e.g. construction equipment, agriculture/forestry vehicles, golf carts, etc.). The onroad emissions module provides a comprehensive framework for calculating on-road vehicle miles traveled (VMT) and associated fuel consumption and CO₂ emissions, integrating multiple datasets across various spatial scales.\\u003c/p\\u003e\\n\\u003cp\\u003eThe onroad modeling process integrates multiple sources of vehicle miles traveled (VMT) data and calculates the associated fleet mix, fuel consumption and CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions. For CONUS, VMT is sourced from GPS-based traffic data for 2023 and is merged with fleet mix information (FHWA Highway Statistics VM1, VM4, and VM2 tables) and the fuel economies (Transportation Energy Data Book), aligning on common dimensions such as year, state, road class, urban/rural land class, and vehicle class. A separate approach is used for Alaska and Hawaii, where VMT is calculated using FHWA HPMS AADT shapefiles for 2022 and aggregated at the state level. Discrepancies between measured HPMS VMT and state-level totals are allocated across road segments based on length, ensuring accurate distribution. Vehicle-type fractions are applied to derive specific VMT estimates, with gaps in truck data filled using HPMS sources.\\u003c/p\\u003e\\n\\u003cp\\u003eFor years before and after the years corresponding to the VMT data (2023 for the contiguous US and 2022 for Alaska and Hawaii), VMT was estimated using a regression model that was trained on 2023 county-level VMT for the CONUS and 2022 VMT for Alaska and Hawaii using road length and population as predictors. This model estimates past VMT, distributing values across road segments proportionally to 2023 (2022 for Alaska and Hawaii) county-level VMT. Estimates are constrained using state-level FHWA VMT and SEDS/EIA fuel data. CO₂ emissions are then computed using fleet mix, fuel economy, and year-specific fuel consumption, following the same methodology as 2023 (2022 for Alaska and Hawaii).\\u003c/p\\u003e\\n\\u003cp\\u003eHoteling\\u003c/p\\u003e\\n\\u003cp\\u003eHoteling emissions are created by long-haul combination truck vehicles idling during mandated rest times for periods of longer than one hour duration. This definition follows that of NEI. Hoteling hours for all four NEI years were read in as a long table and combined with data from the FHWA Highway Statistics Series (VM1 and VM4 tables) and the Transportation Energy Data Book (Tables A.1, A.3, and A.6). The EPA surrogate shapefile for hoteling locations was also incorporated. NEI and FHWA vehicle miles traveled (VMT) data were merged using state FIPS codes, and annual hoteling hours were interpolated for years between or beyond NEI years using VMT data for combined trucks. The interpolated NEI data were spatially joined with EPA hoteling location shapes, and the annual number of hoteling hours at each location was calculated as:\\u003c/p\\u003e\\n\\u003cdiv id=\\\"Equ1\\\" class=\\\"Equation\\\"\\u003e\\n \\u003cdiv class=\\\"mathdisplay\\\" id=\\\"FileID_Equ1\\\" name=\\\"EquationSource\\\"\\u003e$$\\\\:{hours}_{i}\\\\:=\\\\:{hours}_{county}\\\\times\\\\:\\\\frac{{p}_{i}}{{p}_{county}}$$\\u003c/div\\u003e\\n \\u003cdiv class=\\\"EquationNumber\\\"\\u003e2\\u003c/div\\u003e\\n\\u003c/div\\u003e\\n\\u003cp\\u003ewhere hours\\u003csub\\u003ei\\u003c/sub\\u003e represents the hoteling hours at a specific location, hours\\u003csub\\u003ecounty\\u003c/sub\\u003e is the total county-wide hoteling hours, p\\u003csub\\u003ei\\u003c/sub\\u003e is the number of parking spaces at the location, and p\\u003csub\\u003ecounty\\u003c/sub\\u003e is the total parking spaces in the county.\\u003c/p\\u003e\\n\\u003cp\\u003eFinally, CO₂ emissions at each location were estimated using an emission factor and fuel consumption rate, calculated as:\\u003c/p\\u003e\\n\\u003cdiv id=\\\"Equ2\\\" class=\\\"Equation\\\"\\u003e\\n \\u003cdiv class=\\\"mathdisplay\\\" id=\\\"FileID_Equ2\\\" name=\\\"EquationSource\\\"\\u003e$$\\\\:{CO2}_{i}\\\\:=\\\\:{hours}_{i}\\\\times\\\\:\\\\frac{0.71\\\\:gallons}{hour}\\\\times\\\\:\\\\frac{10.21\\\\:kg\\\\:CO2}{gallon}$$\\u003c/div\\u003e\\n \\u003cdiv class=\\\"EquationNumber\\\"\\u003e3\\u003c/div\\u003e\\n\\u003c/div\\u003e\\n\\u003ch3\\u003eEGU\\u003c/h3\\u003e\\n\\u003cp\\u003eThe Energy Generating Unit (EGU) sector is defined as point-level emissions representing emissions from EGUs, aka power plants, associated with Scope 1 electricity generation using data from the U.S. Energy Information Administration (EIA) and Clean Air Markets (CAM). The EGU module focuses on integrating emissions and facility data from the EIA 923 and EIA 860 datasets with CAM emissions data to create a comprehensive and harmonized dataset for CO₂ emissions analysis. The methodology involves several key steps to ensure accurate geospatial tagging, data consolidation, and emissions estimation.\\u003c/p\\u003e\\n\\u003cp\\u003eFirst, the EIA 923 and EIA 860 datasets are merged to tag each facility with latitude and longitude coordinates, enabling spatialization. Any missing geolocation data is supplemented using an external reference table to ensure all facilities are mapped accurately. CAM emissions data, which includes Department of Energy fuel codes, are then aligned with Supplementary Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e fuel IDs using a standardized crosswalk, ensuring consistency in fuel classifications across datasets.\\u003c/p\\u003e\\n\\u003cp\\u003eBoth EIA and CAM datasets provide heat output data, which is used to estimate CO₂ emissions by applying emission factors specific to each fuel type, measured in metric tons of CO₂ per million BTUs (tCO₂/mmBTU). This approach ensures consistency in emissions estimation across the datasets while leveraging the detailed heat output information available in the source data. The EIA and CAM datasets are aggregated by facility, fuel type, and year to create a unified view of emissions at the facility level. These aggregated datasets are then merged using a facility/unit ID crosswalk, which resolves discrepancies between the two sources. Data from the EIA is prioritized over CAM where overlaps occur, as the EIA datasets provide a more comprehensive list of facilities and corresponding data. This ensures the resulting dataset is both complete and accurate. The final dataset provides a comprehensive and geospatially tagged inventory of CO₂ emissions at the facility level, categorized by fuel type and year.\\u003c/p\\u003e\\n\\u003cp\\u003eFor edge cases, the crosswalk process aims to reconcile duplicate facilities from CAM and EIA using facility/unit IDs. Some CAM facility/unit IDs repeat across multiple generator matches in EIA, leading to situations where EIA boiler IDs are either empty or repeated. To address this, rows with empty EIA boiler IDs are filtered out first, followed by a deduplication process to resolve one-to-many mappings. This manual adjustment ensures consistency and avoids data duplication.\\u003c/p\\u003e\\n\\u003cp\\u003eStationary Point\\u003c/p\\u003e\\n\\u003cp\\u003eThe stationary point sector is defined as the facilities that emit greater than 25,000 metric tons CO\\u003csub\\u003e2\\u003c/sub\\u003ee per year (i.e., large CO\\u003csub\\u003e2\\u003c/sub\\u003e emitting facilities), which are required to report emissions to the EPA GHGRP. These facilities encompass the direct emissions from fossil fuel combustion and process emissions from the activities defined in Envirofacts.\\u003c/p\\u003e\\n\\u003cp\\u003eFor the stationary point module, self-reported, facility-level CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions are aggregated to consistent temporal and categorical dimensions, using data from the annual GHGRP, subpart GHGRP, and Envirofacts. Since the three input datasets\\u0026mdash;GHGRP Annual, Envirofacts, and GHGRP Subparts\\u0026mdash;report information for an overlapping set of facility ids, a priority order is applied to resolve discrepancies and duplicated data and retain the most reliable data. Specifically, GHGRP Annual data is prioritized over Envirofacts data, which is further prioritized over GHGRP Subparts data. Additionally, where overlapping facilities are identified, Envirofacts fuel information is appended to enrich the dataset with detailed fuel-related attributes. The data is further cleaned by excluding subpart D (electricity production) and by excluding basin-level oil and gas production emissions in subpart W, as these emissions are covered in other sectors in this model.\\u003c/p\\u003e\\n\\u003cp\\u003eOil and Gas\\u003c/p\\u003e\\n\\u003cp\\u003eThe Oil and Gas emissions sector is defined as a subset of the NEI nonpoint emissions that relate to oil and gas production. SCCs for this sector are matched using the SCC level two descriptions matching \\u0026ldquo;Oil and Gas production\\u0026rdquo;.\\u003c/p\\u003e\\n\\u003cp\\u003eTo estimate CO₂ emissions from oil and gas production, the NEI Nonpoint datasets are processed through a series of filtering and calculation steps. First, the dataset was filtered to include only records that meet the definition, ensuring that only relevant nonpoint sources were included in the analysis. From the filtered data, CO emissions are isolated, and CO₂ emissions were calculated using the emissions factors specified in Supplementary Table S2. Finally, the emissions are aggregated by fuel ID, county, and state to create a spatially resolved inventory. Emissions are spatialized by identifying oil and gas production wells outlined in the USGS oil and gas production dataset\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e\\u003c/sup\\u003e, placing emissions based on production at the well site.\\u003c/p\\u003e\\n\\u003cp\\u003eAirport\\u003c/p\\u003e\\n\\u003cp\\u003eThe airport sector contains emissions for all aircraft (planes and helicopters) during the landing/take-off (LTO) cycle. This sector models CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions by using data on the number of aircraft and aircraft/engine specific emission factors.\\u003c/p\\u003e\\n\\u003cp\\u003eTo estimate CO₂ emissions using the airport module, there are several steps of data filtering, aggregation, and emissions calculations using both FAA 5010 and TFMSC datasets. First, the FAA 5010 dataset is filtered to include only data for years when airports were operational. The resulting data is aggregated by airport, user class, and year, with user class grouping following the rules outlined in Supplementary Table S4, which describes how native FAA 5010 LTO columns are pooled into general user classes. Similarly, the TFMSC dataset is aggregated by airport, user class, engine type, and aircraft type, with user class and engine type mappings defined in Supplementary Table S5 and Supplementary Table S6 respectively. As both FAA 5010 and TFMSC datasets report data for overlapping airports and user classes, FAA 5010 data is prioritized for these cases. The assumption is made that all FAA 5010 reported LTOs are for propeller-type aircraft, as the dataset does not include commercial operations. For non-overlapping TFMSC data, FAA 5010 location data is used to gapfill missing location information, and any remaining TFMSC data lacking location details is dropped, as it primarily represents foreign airports. Emission factors are then applied to calculate CO₂ emissions from LTO operations, with IPCC aircraft emission factors (Supplementary Table S7) being prioritized over EPA engine type emission factors (Supplementary Table S8). In cases where engine type or aircraft type is not reported, default user-class emission factors from the EPA (Supplementary Table S9) are used for CO₂ calculations.\\u003c/p\\u003e\\n\\u003cp\\u003eMobile\\u003c/p\\u003e\\n\\u003cp\\u003eThe mobile emissions sector is defined as a subset of the NEI nonpoint, point, and nonroad emissions that relate to Railroad and Shipping activity. SCCs for this sector are matched using the SCC level two descriptions shown in Supplementary Table S10.\\u003c/p\\u003e\\n\\u003cp\\u003eThis module processes the NEI nonpoint, point, and nonroad datasets to estimate CO₂ emissions by applying specific filtering criteria, calculation methods, and aggregation steps. First, the NEI datasets are filtered to include only data meeting the definitions outlined in Supplementary Table S10. From these filtered datasets, CO emissions are isolated, and CO₂ estimates were calculated by applying emissions factors using Eq.\\u0026nbsp;1. Where available, SCC-specific emissions factors were applied in place of default emissions factors to improve accuracy. The resulting emissions data were then aggregated by fuel ID, state, county, and mobile emissions type (e.g., \\u0026quot;Railroad\\u0026quot; or \\u0026quot;Marine vessels\\u0026quot;) to create a comprehensive and spatially resolved inventory of CO₂ emissions from these source categories. These emissions are then spatially allocated to shipping lanes, ports, railroad segments, or railyards by putting the relevant county-level emission to the target spatial source.\\u003c/p\\u003e\\n\\u003cp\\u003eNonroad Equipment\\u003c/p\\u003e\\n\\u003cp\\u003eThe Nonroad sector is defined as off-road mobile sources that use gasoline, diesel, and other fuels (including agriculture equipment). We integrate the nonroad definition of nonroad emissions, which report these emissions at the county-level, though our definition excludes airplanes, marine and rail locomotives.\\u003c/p\\u003e\\n\\u003cp\\u003eThis module processes the NEI Nonroad dataset to estimate CO₂ emissions by applying specific filtering criteria, calculation methods, and aggregation steps. The dataset is first filtered to include only data that meets the definitions outlined in (Supplementary Table S11), ensuring the analysis focuses on relevant nonroad sources. From this filtered data, CO emissions are extracted and converted to CO₂ estimates using emissions factors specific to fuel types, as detailed in Eq. \\u003cspan class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003e. Finally, the emissions are aggregated by fuel, state, and county, resulting in a structured dataset that offers a detailed spatial breakdown of CO₂ emissions.\\u003c/p\\u003e\\n\\u003cp\\u003eUncertainty\\u003c/p\\u003e\\n\\u003cp\\u003eEPA emission factors from AP-40 and the underlying activity data (NEI emissions and EIA fuel consumption) do not report meaningful uncertainty values. As such, we follow the IPCC recommendation of using a Monte Carlo analysis to identify uncertainty\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e\\u003c/sup\\u003e. Several studies have applied this Monte Carlo approach to quantify uncertainties in national or regional emissions of CO2 and other pollutants\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e\\u0026ndash;\\u003cspan class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e\\u003c/sup\\u003e. The method has been used for emissions from various sectors, including power plants, transportation, industrial processes, and oil and gas\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e\\u0026ndash;\\u003cspan class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e\\u003c/sup\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eIn this model, we use the Monte Carlo approach to generate multiple random iterations of the model, where input variables (in this case, emission factors and activity data) are sampled from probability distributions that reflect their potential variability. We group the emission factors into their economic sectors and fuel types to develop a probability distribution for these factors when the EPA AP-40 does not include additional information. We do not use the distribution of other input activity, e.g., NEI inputs, self-reported emissions to the EPA, or fuel consumption estimates from the EIA because these datasets do not report confidence intervals on their outputs and developing meaningful distributions of the data are obfuscated. The uncertainty is limited to the modeling process, such that no additional uncertainty analysis is performed to estimate uncertainty due to geospatial processing.\\u003c/p\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cp\\u003eCrosswalk ffCO\\u003csub\\u003e2\\u003c/sub\\u003e Estimates by Sector and Year\\u003c/p\\u003e\\n\\u003cp\\u003eNationally, direct emissions of fossil fuel CO\\u003csub\\u003e2\\u003c/sub\\u003e have decreased by 896 MMT (13%) from 2010 to 2024. Until 2023, the highest emitting sector contributing to national CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions came from electric power generation (1,800–2,500 MMT CO\\u003csub\\u003e2\\u003c/sub\\u003e, corresponding to 32–40% of national emissions). After 2023, the transportation sector overtook power generation (1,760 MMT CO\\u003csub\\u003e2\\u003c/sub\\u003e vs 1,670 MMTCO\\u003csub\\u003e2\\u003c/sub\\u003e), though both sectors contribute to roughly 33% of total emissions. Annual sector totals are outlined in Table 1 and shown in Fig. 2.\\u003c/p\\u003e\\n\\u003cp\\u003eTable 1. CO\\u003csub\\u003e2\\u003c/sub\\u003e estimates over the US by sector for Crosswalk (units in MMT CO2).\\u003c/p\\u003e\\n\\u003ctable border=\\\"1\\\" cellspacing=\\\"0\\\" cellpadding=\\\"0\\\" width=\\\"624\\\"\\u003e\\n \\u003ctbody\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eYear\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eAirport\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eCommercial buildings\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eIndustrial\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eNonroad\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eOther\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003ePowerplants\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eResidential\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eVehicles\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eTotal\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2010\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e51\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e161\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1362\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e125\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2550\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e352\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1541\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e6163\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2011\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e50\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e161\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1401\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e127\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e20\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2451\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e350\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1520\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e6082\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2012\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e50\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e163\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1390\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e130\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e26\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2326\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e350\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1504\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5938\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2013\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e50\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e166\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1393\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e132\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e23\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2346\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e352\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1534\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5996\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2014\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e50\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e172\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1410\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e134\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e23\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2347\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e353\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1558\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e6048\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2015\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e50\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e176\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1395\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e135\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e23\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2195\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e344\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1563\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5880\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2016\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e51\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e179\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1367\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e135\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e23\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2109\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e334\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1582\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5781\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2017\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e52\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e183\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1374\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e136\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e23\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2026\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e339\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1585\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5717\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2018\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e53\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e186\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1414\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e136\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e23\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2049\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e346\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1609\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5815\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2019\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e54\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e188\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1405\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e135\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1889\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e349\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1614\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5656\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2020\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e35\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e189\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1355\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e133\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1717\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e346\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1433\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5229\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2021\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e44\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e191\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1374\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e131\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1812\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e344\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1541\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5460\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2022\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e49\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e193\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1369\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e130\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1795\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e343\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1542\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5442\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2023\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e52\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e194\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1354\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e127\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1666\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e344\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1571\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5330\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e2024\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e53\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e188\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1316\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e129\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1667\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e326\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e1566\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e5267\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003c/tbody\\u003e\\n\\u003c/table\\u003e\\n\\u003cp\\u003eTable 2. CO\\u003csub\\u003e2\\u003c/sub\\u003e estimates for 2024 for 5 highest states (units in MMT CO\\u003csub\\u003e2\\u003c/sub\\u003e).\\u003c/p\\u003e\\n\\u003ctable border=\\\"1\\\" cellspacing=\\\"0\\\" cellpadding=\\\"0\\\" width=\\\"624\\\"\\u003e\\n \\u003ctbody\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eYear\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eAirport\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eCommercial buildings\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eIndustrial\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eNonroad\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eOther\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003ePowerplants\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eResidential\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eVehicles\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd\\u003e\\n \\u003cp\\u003e\\u003cstrong\\u003eTotal\\u003c/strong\\u003e\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003eTexas\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e5\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e9\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e328\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e6\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e2\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e216\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e12\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e183\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e762\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003eCalifornia\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e5\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e22\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e74\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e55\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e2\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e48\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e24\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e145\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e376\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003eFlorida\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e5\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e5\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e24\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e6\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e0\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e101\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e93\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e236\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003eLouisiana\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e0\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e140\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e54\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e2\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e34\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e232\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003ePennsylvania\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e10\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e42\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e4\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e1\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e78\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e18\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e53\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\"\\u003e\\n \\u003cp\\u003e205\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003c/tbody\\u003e\\n\\u003c/table\\u003e\\n\\u003cp\\u003eThe states with the largest absolute Scope 1 emissions are Texas, California, and Florida, Louisiana, and Pennsylvania (Table 2). Texas and Louisiana have the largest industrial CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions while California has the largest portion of emissions from vehicles, and Florida and Pennsylvania have the largest proportion of emissions from power plants.\\u003c/p\\u003e\\n\\u003cp\\u003eComparison of CO\\u003csub\\u003e2\\u003c/sub\\u003e Product\\u003c/p\\u003e\\n\\u003ch3\\u003eNational Emissions Inventories\\u003c/h3\\u003e\\n\\u003cp\\u003eAs shown in Fig. 3, we compare the Crosswalk data product output to published national inventories. The datasets used for comparison are the EPA Greenhouse Gas Inventory (GHGI), the EPA state implementation tool (SIT), CarbonTracker v2022 product\\u003csup\\u003e24\\u003c/sup\\u003e, Vulcan\\u003csup\\u003e5\\u003c/sup\\u003e, and ClimateTrace\\u003csup\\u003e25\\u003c/sup\\u003e. To maintain consistency, testing datasets were cropped to include only fossil fuel emissions over the continental US, limiting the analysis to this specific emission category and spatial extent.\\u003c/p\\u003e\\n\\u003cp\\u003eAt the national level, the Crosswalk CO\\u003csub\\u003e2\\u003c/sub\\u003e product agrees best with the EPA GHGI, with the lowest differences (10% on average). Crosswalk shows similar year-over-year trends as the EPA, Vulcan, and Carbon tracker datasets (r \\u0026gt; 0.99), with the lowest agreement with ClimateTrace, though the total amount of CO\\u003csub\\u003e2\\u003c/sub\\u003e converges in the most recent years (Fig. 3). This high level of agreement with established inventories demonstrates the commonality in methodology and/or shared data sources between the inventories and the Crosswalk CO\\u003csub\\u003e2\\u003c/sub\\u003e product, while the discrepancies with ClimateTrace highlight the differences in methodology and data sources, particularly in earlier years. These variations underscore the importance of continued cross-comparison and validation to improve consistency across emissions datasets and provide a more comprehensive understanding of national CO\\u003csub\\u003e2\\u003c/sub\\u003e trends.\\u003c/p\\u003e\\n\\u003ch3\\u003eGridded Emissions Comparison\\u003c/h3\\u003e\\n\\u003cp\\u003eComparison of spatial emissions patterns between the Crosswalk and CarbonTracker v2022 products reveals a strong overall correlation (r = 0.61), indicating broad agreement in the spatial distribution of emissions. Notably, both products show that the eastern United States has substantially higher fossil fuel CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions—over 250% greater—than the western United States (Fig. 4a,b). However, some regional differences persist. In total, Crosswalk estimates tend to be higher than those from CarbonTracker, though spatially there are some areas in the Western US which are lower in Crosswalk relative to CarbonTracker (Fig. 4c). These discrepancies may stem from differences in input data, sectoral assumptions, or spatial allocation methods. The CarbonTracker product is an assimilated modeling system that integrates atmospheric CO\\u003csub\\u003e2\\u003c/sub\\u003e observations with surface flux estimates to produce more accurate, observation-constrained emissions. By assimilating global atmospheric measurements, CarbonTracker adjusts emission fluxes to reflect atmospheric observations, with the v2022 product differentiating ffCO\\u003csub\\u003e2\\u003c/sub\\u003e and biogenic CO\\u003csub\\u003e2\\u003c/sub\\u003e with priors as opposed to observations. For spatial allocation of emissions, CarbonTracker employs nighttime lights as a proxy for human activity, enabling a finer resolution depiction of where anthropogenic CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions are likely to originate. This contrasts to Crosswalk emissions, which use both measured fluxes allocated at the source and bottom-up emissions estimates which are spatially allocated to specific emitting shapes (e.g., roadways, oil and gas wells, buildings). As shown in Fig. 3, CarbonTracker is significantly lower than Crosswalk and the published EPA bottom-up emissions inventories (SIT, GHGI), which highlight the discrepancies between bottom-up and top-down emissions estimates and different approaches and assumptions to back out anthropogenic emissions.\\u003c/p\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eThis study builds off existing research into anthropogenic-based CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions and integrates updates to better address the limitations of current CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions models. These advancements allow for a more comprehensive and flexible modeling framework capable of addressing both historical trends and future projections of CO₂ emissions. By combining high-resolution activity data with cutting-edge computational infrastructure, this study aims to improve spatial granularity and consistent releases, enabling decision-makers to better understand localized emissions patterns and their drivers. First, we integrate public and private activity datasets to best capture historical emissions (e.g., airplane types and counts from FAA, GPS-based traffic data, and self-reported quarterly emissions for large point sources from the Clean Air Markets Database (CAMD)). Second, we integrate this modeling framework into a cloud-native environment to produce reliable and consistent outputs. Finally, we integrate energy fuel forecasts to allow for emissions to be modeled to near-real time, and out to 2050. The integrated approach presented herein not only addresses existing gaps in emissions modeling, but also establishes a robust foundation for dynamic and data-driven climate action strategies.\\u003c/p\\u003e\\u003cp\\u003eWe provide data aggregated to the neighborhood scale and aggregated for all US cities with populations greater than 5,000 people. This makes our product particularly useful for smaller jurisdictions that might not otherwise have access to their own fossil fuel CO₂ inventory, and might be required to rely on or extrapolate from data from the larger metro area or county. Another strength of our approach is that by using a cloud-native environment we are able to automatically ingest and utilize data when updates or new source data come online. Integration of energy fuel forecasts makes our model particularly conducive to forecasting future emissions under varying scenarios.\\u003c/p\\u003e\\u003cp\\u003eA key area for future work is characterization of uncertainty. The largest source of uncertainty in the current methodology arises from the nonpoint buildings sector. The reliance on downscaling county-level emissions from the NEI introduces potential inaccuracies due to variations in building-level energy consumption that are not fully captured in aggregate data. A more robust approach would involve integrating direct utility billing data or high-resolution building energy models to better constrain emissions estimates. Access to improved fuel use data, particularly for heating and industrial processes, would enhance the spatial and temporal resolution of emissions estimates. Additionally, incorporating uncertainty estimates from the NEI would help quantify and communicate the confidence level in the final emissions inventory, providing a clearer picture of potential variability.\\u003c/p\\u003e\\u003cp\\u003eAnother key area for future work is the integration of top-down and bottom-up emissions estimation methods to improve inventory accuracy and facilitate better atmospheric modeling. Remote sensing data, atmospheric inversion models, and satellite-based CO₂ measurements could complement ground-based inventories, helping to reconcile discrepancies between modeled and observed emissions. Developing a framework for systematically merging these approaches would lead to a more representative emissions dataset, improving the accuracy of emissions inventories used in climate modeling, air quality analysis, and policy decision-making. Further refinement of temporal emissions profiles, incorporating real-time activity data where possible, would also enhance the responsiveness and applicability of the dataset for near-term emissions mitigation efforts.\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003eData Availability\\u003c/p\\u003e\\n\\u003cp\\u003eThe datasets generated and analysed during the current study are available in this Zenodo repository, https://zenodo.org/records/16986970?preview=1\\u0026amp;token=eyJhbGciOiJIUzUxMiIsImlhdCI6MTc1NjM\\n5ODMzMiwiZXhwIjoxNzU5Mjc2Nzk5fQ.eyJpZCI6IjIyOTc3YjgxLWU4ZWEtNGY3MC1hMGQ0LTQ0ZWIxZTMwODQ5NCIsImRh\\ndGEiOnt9LCJyYW5kb20iO\\niJjNWZiMDZkMmNjZmNjNzQ0Nzc1ZmFlZDA2ODg\\nzZDFlYyJ9.wS9S1AUcWpAaTTAYvwVmc2jWEo9K1QauOB4-yWeKgZrZlz5vXmViASMdO2yOcwZ2UEznhYTe2l-FlfsIcVAzGg which can be cited as https://doi.org/10.5281/zenodo.16986970.\\u003c/p\\u003e\\n\\n\\u003cp\\u003eFunding\\u003c/p\\u003e\\n\\u003cp\\u003eThis research was supported by Crosswalk Labs. \\u003c/p\\u003e\\n\\u003cp\\u003eAcknowledgements \\u003c/p\\u003e\\n\\u003cp\\u003eThe Data Foundation provided funding to publish results from this research in an open data format.\\u003c/p\\u003e\\n\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n\\u003cli\\u003eFriedlingstein, P. \\u003cem\\u003eet al.\\u003c/em\\u003e Global Carbon Budget 2022. \\u003cem\\u003eEarth Syst. Sci. Data\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e, 4811\\u0026ndash;4900 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eZhang, W., Huang, Y. \\u0026amp; Wu, H. The symmetric and asymmetric effects of economic policy uncertainty and oil prices on carbon emissions in the USA and China: evidence from the ARDL and non-linear ARDL approaches. \\u003cem\\u003eEnviron. Sci. Pollut. Res.\\u003c/em\\u003e \\u003cstrong\\u003e29\\u003c/strong\\u003e, 26465\\u0026ndash;26482 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eKauffman, J. B., Beschta, R. L., Lacy, P. M. \\u0026amp; Liverman, M. Livestock Use on Public Lands in the Western USA Exacerbates Climate Change: Implications for Climate Change Mitigation and Adaptation. \\u003cem\\u003eEnviron. Manage.\\u003c/em\\u003e \\u003cstrong\\u003e69\\u003c/strong\\u003e, 1137\\u0026ndash;1152 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eSeto, K. C., G\\u0026uuml;neralp, B. \\u0026amp; Hutyra, L. R. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. \\u003cem\\u003eProc. Natl. Acad. Sci.\\u003c/em\\u003e \\u003cstrong\\u003e109\\u003c/strong\\u003e, 16083\\u0026ndash;16088 (2012).\\u003c/li\\u003e\\n\\u003cli\\u003eGurney, K. R. \\u003cem\\u003eet al.\\u003c/em\\u003e The Vulcan Version 3.0 High‐Resolution Fossil Fuel CO\\u003csub\\u003e2\\u003c/sub\\u003e Emissions for the United States. \\u003cem\\u003eJ. Geophys. Res. Atmospheres\\u003c/em\\u003e \\u003cstrong\\u003e125\\u003c/strong\\u003e, (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eTao, J. \\u0026amp; Kong, X. Spatial allocation of anthropogenic carbon dioxide emission statistics data fusing multi-source data based on Bayesian network. \\u003cem\\u003eSci. Rep.\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, 18128 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eOda, T., Maksyutov, S. \\u0026amp; Andres, R. J. The Open-source Data Inventory for Anthropogenic CO\\u003csub\\u003e2\\u003c/sub\\u003e, version 2016 (ODIAC2016): a global monthly fossil fuel CO\\u003csub\\u003e2\\u003c/sub\\u003e; gridded emissions data product for tracer transport simulations and surface flux inversions. \\u003cem\\u003eEarth Syst. Sci. Data\\u003c/em\\u003e \\u003cstrong\\u003e10\\u003c/strong\\u003e, 87\\u0026ndash;107 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eAsner, G. P. \\u003cem\\u003eet al.\\u003c/em\\u003e Targeted carbon conservation at national scales with high-resolution monitoring. \\u003cem\\u003eProc. Natl. Acad. Sci.\\u003c/em\\u003e \\u003cstrong\\u003e111\\u003c/strong\\u003e, (2014).\\u003c/li\\u003e\\n\\u003cli\\u003eChen, P. The impact of low-carbon city pilots on carbon emission \\u0026ndash; new evidence from high-resolution carbon emission data. \\u003cem\\u003eKybernetes\\u003c/em\\u003e \\u003cstrong\\u003e52\\u003c/strong\\u003e, 5875\\u0026ndash;5892 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003eFeng, S. \\u003cem\\u003eet al.\\u003c/em\\u003e Los Angeles megacity: a high-resolution land\\u0026ndash;atmosphere modelling system for urban CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions. \\u003cem\\u003eAtmospheric Chem. Phys.\\u003c/em\\u003e \\u003cstrong\\u003e16\\u003c/strong\\u003e, 9019\\u0026ndash;9045 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eAdams, S., Adedoyin, F., Olaniran, E. \\u0026amp; Bekun, F. V. Energy consumption, economic policy uncertainty and carbon emissions; causality evidence from resource rich economies. \\u003cem\\u003eEcon. Anal. Policy\\u003c/em\\u003e \\u003cstrong\\u003e68\\u003c/strong\\u003e, 179\\u0026ndash;190 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eSargent, M. R. \\u003cem\\u003eet al.\\u003c/em\\u003e Majority of US urban natural gas emissions unaccounted for in inventories. \\u003cem\\u003eProc. Natl. Acad. Sci.\\u003c/em\\u003e \\u003cstrong\\u003e118\\u003c/strong\\u003e, e2105804118 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eLi, T. \\u003cem\\u003eet al.\\u003c/em\\u003e Reconstructing Global Daily CO2 Emissions via Machine Learning. Preprint at https://doi.org/10.48550/arXiv.2407.20057 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eGately, C. K. \\u0026amp; Hutyra, L. R. Large Uncertainties in Urban‐Scale Carbon Emissions. \\u003cem\\u003eJ. Geophys. Res. Atmospheres\\u003c/em\\u003e \\u003cstrong\\u003e122\\u003c/strong\\u003e, (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eSkinner, C. C. \\u003cem\\u003eet al.\\u003c/em\\u003e Aggregated Oil and Natural Gas Drilling and Production History of the United States (ver. 1.1, April 2023). U.S. Geological Survey https://doi.org/10.5066/P9UIR5HE (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eMastrandrea, M. D. \\u003cem\\u003eet al.\\u003c/em\\u003e The IPCC AR5 guidance note on consistent treatment of uncertainties: a common approach across the working groups. \\u003cem\\u003eClim. Change\\u003c/em\\u003e \\u003cstrong\\u003e108\\u003c/strong\\u003e, 675\\u0026ndash;691 (2011).\\u003c/li\\u003e\\n\\u003cli\\u003eWu, Y. \\u003cem\\u003eet al.\\u003c/em\\u003e Methods to account for uncertainties in exposure assessment in studies of environmental exposures. \\u003cem\\u003eEnviron. Health\\u003c/em\\u003e \\u003cstrong\\u003e18\\u003c/strong\\u003e, 31 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eSuper, I., Dellaert, S. N. C., Visschedijk, A. J. H. \\u0026amp; Denier Van Der Gon, H. A. C. Uncertainty analysis of a European high-resolution emission inventory of CO\\u003csub\\u003e2\\u003c/sub\\u003e and CO to support inverse modelling and network design. \\u003cem\\u003eAtmospheric Chem. Phys.\\u003c/em\\u003e \\u003cstrong\\u003e20\\u003c/strong\\u003e, 1795\\u0026ndash;1816 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eZhao, Y., Nielsen, C. P., Lei, Y., McElroy, M. B. \\u0026amp; Hao, J. Quantifying the uncertainties of a bottom-up emission inventory of anthropogenic atmospheric pollutants in China. \\u003cem\\u003eAtmospheric Chem. Phys.\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, 2295\\u0026ndash;2308 (2011).\\u003c/li\\u003e\\n\\u003cli\\u003eTchepel, O. \\u003cem\\u003eet al.\\u003c/em\\u003e EMISSION MODELLING OF HAZARDOUS AIR POLLUTANTS FROM ROAD TRANSPORT AT URBAN SCALE. \\u003cem\\u003eTRANSPORT\\u003c/em\\u003e \\u003cstrong\\u003e27\\u003c/strong\\u003e, 299\\u0026ndash;306 (2012).\\u003c/li\\u003e\\n\\u003cli\\u003eRoest, G. \\u0026amp; Schade, G. Quantifying alkane emissions in the Eagle Ford Shale using boundary layer enhancement. \\u003cem\\u003eAtmospheric Chem. Phys.\\u003c/em\\u003e \\u003cstrong\\u003e17\\u003c/strong\\u003e, 11163\\u0026ndash;11176 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eGurney, K. R. \\u003cem\\u003eet al.\\u003c/em\\u003e Quantification of Fossil Fuel CO\\u003csub\\u003e2\\u003c/sub\\u003e Emissions on the Building/Street Scale for a Large U.S. City. \\u003cem\\u003eEnviron. Sci. Technol.\\u003c/em\\u003e \\u003cstrong\\u003e46\\u003c/strong\\u003e, 12194\\u0026ndash;12202 (2012).\\u003c/li\\u003e\\n\\u003cli\\u003eLiu, Z. \\u003cem\\u003eet al.\\u003c/em\\u003e Near-real-time monitoring of global CO2 emissions reveals the effects of the COVID-19 pandemic. \\u003cem\\u003eNat. Commun.\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, 5172 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003ePeters, W. \\u003cem\\u003eet al.\\u003c/em\\u003e An atmospheric perspective on North American carbon dioxide exchange: CarbonTracker. \\u003cem\\u003eProc. Natl. Acad. Sci.\\u003c/em\\u003e \\u003cstrong\\u003e104\\u003c/strong\\u003e, 18925\\u0026ndash;18930 (2007).\\u003c/li\\u003e\\n\\u003cli\\u003eGore, A. Measure emissions to manage emissions. \\u003cem\\u003eScience\\u003c/em\\u003e \\u003cstrong\\u003e378\\u003c/strong\\u003e, 455\\u0026ndash;455 (2022).\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"carbon dioxide emissions, greenhouse gas emissions, fossil fuel, emissions inventory\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-7456083/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-7456083/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eWe present 2010\\u0026ndash;2024 annual model output of a new carbon dioxide emissions model, with a raw resolution at points, lines, and polygons corresponding to emission sources, and a gridded resolution of 1 km\\u0026sup2; spatial resolution. The underlying model modernizes emissions modeling by incorporating web-scraping and data-fusion methods to update input emissions values as they are ingested, allowing for reliable and consistent updates of the emissions model. From the output, direct CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions from fuel combustion for the entire US show that emissions were 5,267 MMTCO\\u003csub\\u003e2\\u003c/sub\\u003e in 2024. The largest activity sectors contributing to the national emissions was emissions from electricity production (1,667 MMTCO\\u003csub\\u003e2\\u003c/sub\\u003e, 31.6% of national total) and the onroad sector (1,566 MMTCO\\u003csub\\u003e2\\u003c/sub\\u003e, 29.7% of national total). By integrating energy forecasts, this model also highlights the potential for enhanced decarbonization policy applications given macroeconomic trends. Further comparison against existing federal datasets such as those of the Environmental Protection Agency\\u0026rsquo;s Greenhouse Gas Emissions Inventory and State Inventory Tool, and independent datasets such as Vulcan, Open-Data Inventory for Anthropogenic Carbon dioxide (ODIAC), and CarbonTracker, demonstrate robust agreement, though variation exists in spatial patterns and the presented dataset comprises relatively higher CO\\u003csub\\u003e2\\u003c/sub\\u003e emissions estimates. The flexibility and scalability of the model make it a valuable tool for monitoring CO₂ emissions trends and informing mitigation strategies.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Modernizing carbon dioxide emissions inventories for action in the United States\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-09-16 18:27:27\",\"doi\":\"10.21203/rs.3.rs-7456083/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"e38126a8-7588-4056-8e49-98616d12297a\",\"owner\":[],\"postedDate\":\"September 16th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[{\"id\":54595148,\"name\":\"Earth and environmental sciences/Climate sciences\"},{\"id\":54595149,\"name\":\"Physical sciences/Energy science and technology\"},{\"id\":54595150,\"name\":\"Earth and environmental sciences/Environmental sciences\"},{\"id\":54595151,\"name\":\"Earth and environmental sciences/Environmental social sciences\"}],\"tags\":[],\"updatedAt\":\"2025-10-27T12:57:30+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2025-09-16 18:27:27\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-7456083\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-7456083\",\"identity\":\"rs-7456083\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}