Contrasting air pollution responses to hourly varying anthropogenic NOx emissions in the contiguous United States

Short Summary: Global models often rely on highly simplified emissions patterns that lack real-world hourly variations. Our study finds that representing their changes throughout the day substantially changes predicted air pollution levels. The impact can be large, comparable to a uniform 30% reduction in total emissions, and varies significantly by region, especially between the eastern and western US and between urban and rural areas.

Abstract: Some global atmospheric chemistry modeling applications assume that intra-month variability in anthropogenic emissions averages out at monthly timescales. To systematically quantify the impacts of resolving daily and hourly emissions, we use a global model with a refined ∼ 14 km resolution over the contiguous United States (CONUS; MUSICAv0) and a regional CONUS inventory for July 2018. 

Switching from daily to hourly nitric oxide (NO) emissions (typically higher during the day and lower at night) yields contrasting spatial responses in nitrogen oxides (NO_x ≡  NO+ nitrogen dioxide (NO₂)) and ozone (O₃) concentrations in the western versus eastern CONUS and in urban versus rural areas. Neglecting hourly variations in CONUS NO emissions leads to grid-cell level discrepancies in monthly mean surface O₃ concentrations of −22 % to +11 % (−7 to +5 ppb) and surface NO₂ of −49 % to +86 % (−1 to +8 ppb), with tropospheric NO₂ columns showing similar spatial patterns (−12 % to +56 %). While comparable in magnitude to a uniform 30 % NO emission reduction (grid-cell level surface O₃ differences of −12 % to +9 %, −7 to +3 ppb), the spatial response patterns differ with location-specific timing of emissions and meteorology. For example, Los Angeles shows higher morning NO_x concentrations and stronger NO_x-saturated O₃ suppression relative to New York City. 

A simple scaling analysis suggests that neglecting hourly emissions variability can bias NO_x emissions inferred from monthly mean tropospheric NO₂ columns, with absolute relative differences ranging from ∼ 1 % to ∼ 56 % within individual model grid cells.