Developing the United States GReenhouse gas And Air Pollutants Emissions System (GRA2PES)

Published

October 11, 2025

Author(s)

Congmeng Lyu, Colin Harkins, Meng Li, Kimberly Mueller, Jack Prothero, Bert Verreyken, John Miller, Scott Lehman, Jeff Peischl, Jessica Gilman, CARSTEN WARNEKE, MATTHEW COGGON, Chelsea Stockwell, Steve Brown, Kristen Zuraski, Aaron Lamplugh, kevin gurney, Lech Gawuc, Pawlok Dass, Rachel Hoesly, Steve Smith, Tomohiro Oda, Lucy Hutyra, Conor Gately, claire granier, James Whetstone, BRIAN MCDONALD

Abstract

In the U.S., emissions of greenhouse gas (GHG) and air quality (AQ) are often developed independent of one another. Here, we develop the GReenhouse gas And Air Pollutant Emissions System (GRA2PES), providing gridded emissions of fossil-fuel carbon dioxide (ffCO2) and 93 air pollutant species across 17 sectors, at 4 km x 4 km spatial resolution across the contiguous US (CONUS). We find the AQ species most strongly correlated spatially with ffCO2 are nitrogen oxides (NOx, ρ = 0.80), followed by sulfur dioxide (SO2, ρ = 0.64), fine particulate matter (PM2.5, ρ = 0.49), and carbon monoxide (CO, ρ = 0.39). We evaluate GRA2PES ffCO2 emissions with publicly available regional and global inventories. At the national, state, and urban-scale, GRA2PES is within +1.5%, +1.5% normalized mean bias (NMB, R2 = 0.98), and +12.0% NMB (R2 = 0.97) of an ensemble average, respectively. Consistency across inventories diverge at state (NMB = -9.8% to +5.6%, R2 = 0.93 to 0.99) and urban scales (NMB = -27.5% to +16.3%, R2 = 0.92 to 0.98). Utilizing measurements from the Southwest Urban NOx and VOC Experiment (SUNVEx) in 2021, we evaluate GRA2PES ffCO2 and AQ emissions in the Los Angeles basin. The convolution of GRA2PES ffCO2 emissions against radiocarbon (14C)-derived ffCO2 observations is -23%, which is within the range of convolutions across publicly available inventories (-44% to +94%). GRA2PES AQ/ffCO2 ratios generally agree with field observations (NMB = 3%, log R2 = 0.92). Taken together, we demonstrate a method by which to utilize GRA2PES to derive AQ emission fluxes from ffCO2 emissions. Lastly, we show that at the urban-scale, the AQ species most strongly influenced by fossil fuel combustion are CO, ethyne, ethene, pentanes, total reactive nitrogen (NOy), benzene, and propene.

Citation

Journal of Geophysical Research-Atmospheres

Volume

130

Issue

20

Pub Type

Journals

Download Paper

https://doi.org/10.1029/2025JD043597

Local Download

Citation

Lyu, C. , Harkins, C. , Li, M. , Mueller, K. , Prothero, J. , Verreyken, B. , Miller, J. , Lehman, S. , Peischl, J. , Gilman, J. , WARNEKE, C. , Coggon, M. , Stockwell, C. , Brown, S. , Zuraski, K. , Lamplugh, A. , Gurney, K. , Gawuc, L. , Dass, P. , Hoesly, R. , Smith, S. , Oda, T. , Hutyra, L. , Gately, C. , granier, C. , Whetstone, J. and McDonald, B. (2025), Developing the United States GReenhouse gas And Air Pollutants Emissions System (GRA2PES), Journal of Geophysical Research-Atmospheres, [online], https://doi.org/10.1029/2025JD043597, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=959404 (Accessed December 12, 2025)