Measured and modelled ozone photochemical production in the Baltimore-Washington airshed. (April 2019)

• Record Type: Journal Article
• Authors: Hembeck, Linda; He, Hao; Vinciguerra, Timothy P.; Canty, Timothy P.; Dickerson, Russell R.; Salawitch, Ross J.; Loughner, Christopher
• Is Part Of: Atmospheric environment. Volume 2(2019)
• Publication Date: 2019-04
• Subjects: Air quality -- Surface ozone -- CMAQ -- Ozone production efficiency
• DOI: 10.1016/j.aeaoa.2019.100017
• Languages: English
• ISSNs: 2590-1621

Abstract

Abstract: Ozone production efficiency (OPE) represents the number of ozone (O3 ) molecules produced per molecule of NOx (NO2 +NO), before NOx is converted to a sink or reservoir species. Observations from NASA's July 2011 DISCOVER-AQ (D-AQ) campaign in the Baltimore-Washington region are used to quantify OPE, which is compared to output from the Community Multiscale Air Quality (CMAQ) model. The baseline model run of CMAQ (CB05-TUCL mechanism; mobile NOx emissions from the National Emissions Inventory; biogenic emissions from BEIS model v3.61) underestimates observed OPE by 34 ± 20%. The payload of D-AQ lacked direct observations of HO2 and RO2, so we infer the rate at which these peroxy radicals react with NO (termed inROx ) from observations of NO, NO2, O3, and j(NO2 ). The baseline run of CMAQ underestimates inROx by 35 ± 19%. A modified run of CMAQ with a factor of 10 reduction in the lifetime of organic nitrates, a factor of 2 reduction in mobile NOx emissions, an update to the thermal dissociation rate of PAN and PANX to the IUPAC 2014 value, and a correction for the rate constant for OH + PANX results in better overall model performance based on a number of comparisons to data that include column NO2 measured by the Ozone Monitoring Instrument, a value of OPE that is 21 ± 13% less than observed, and inROx that is low by 30 ± 19%. Sensitivity analysis indicates that the peroxy radical discrepancy involves either emission of VOCs other than isoprene or HCHO, or a problem with the model production of HO2 and RO2 . Our analysis suggests surface O3 may exhibit stronger declines to further reductions of NOx than is indicated by baseline runs of CMAQ. Index terms: Troposphere: constituent transport and chemistry. Highlights: CMAQ using CB05-TUCL underestimates observed O3 production efficiency (OPE) by ∼34%. CMAQ also underestimates the peroxy radical (HO2 +RO2 ) reactivity with NO by ∼35%. CMAQ overestimates NOy and peroxy nitrates (PNs) by 40% and 100%, respectively. CMAQ was also run with reduced emission of NOx and faster decomposition of PNs. This modified CMAQ simulation provides better representation of OPE, NOy and PNs.