CAMx ozone source attribution in the eastern United States using guidance from observations during DISCOVER‐AQ Maryland

Abstract

A Comprehensive Air‐Quality Model with Extensions (CAMx) version 6.10 simulation was assessed through comparison with data acquired during NASA's 2011 Deriving Information on Surface Conditions from Column and Vertically Resolved Observations Relevant to Air Quality (DISCOVER‐AQ) Maryland field campaign. Comparisons for the baseline simulation (Carbon Bond 2005 (CB05) chemistry, Environmental Protection Agency 2011 National Emissions Inventory) show a model overestimate of NOy by +86.2% and an underestimate of HCHO by −28.3%. We present a new model framework (Carbon Bond 6 Revision 2 chemistry (CB6r2), Model of Emissions of Gases and Aerosols from Nature (MEGAN) version 2.1 biogenic emissions, 50% reduction in mobile NOx, enhanced representation of isoprene nitrates) that better matches observations. The new model framework attributes 31.4% more surface ozone in Maryland to electric generating units (EGUs) and 34.6% less ozone to on‐road mobile sources. Surface ozone becomes more NOx limited throughout the eastern United States compared to the baseline simulation. The baseline model therefore likely underestimates the effectiveness of anthropogenic NOx reductions as well as the current contribution of EGUs to surface ozone.

Top‐Down Estimates of NOx and CO Emissions From Washington, D.C.‐Baltimore During the WINTER Campaign

“…The SMOKE analysis predicted slightly higher afternoon CO (1,500 ± 500 mol s −1 ) and NO x (170 ± 20 mol s −1 ) emission rates compared to the interpolated, forward‐projected, temporally allocated NEI CO (~1,100 mol s −1 ) and NO x (~115 mol s −1 ) emission estimates. A step‐by‐step explanation of the scaling/temporal allocations conducted on the D.C.‐Balt emission inventories, as well as discussion of the SMOKE analysis (EPA, , ; Goldberg et al, , ; Houyoux & Vukovich, ; Kota et al, ; Vinciguerra et al, ), is provided in Text S4 and Figures S6 and S7.…”

“…WINTER 2015 CO and NO x emissions were also calculated from Sparse Matrix Operator Kernel Emissions (SMOKE) outputs based on the NEI-11, as described by Anderson et al (2014). The SMOKE analysis predicted slightly higher afternoon CO (1,500 ± 500 mol s À1 ) and NO x (170 ± 20 mol s À1 ) emission rates compared to the interpolated, forward-projected, temporally allocated NEI CO (~1,100 mol s À1 ) and NO x (~115 mol s À1 ) (EPA, 2014a(EPA, , 2014bGoldberg et al, 2015Goldberg et al, , 2016Houyoux & Vukovich, 1999;Kota et al, 2012;Vinciguerra et al, 2017), is provided in Text S4 and Figures S6 and S7.…”

A high-resolution and observationally constrained OMI NO<sub>2</sub> satellite retrieval

“…Many literature sources, including others using different model setups (all are based on the NEI), also show a high bias in simulating summertime column NO 2 (Canty et al, 2015;Souri et al, 2016), NO x (Travis et al, 2016), and NO y (Goldberg et al, 2016). In Fig.…”

Ozone production and its sensitivity to NO<sub><i>x</i></sub> and VOCs: results from the DISCOVER-AQ field experiment, Houston 2013

“…Depending on physical and chemical conditions, the production of ozone can be either NO x -sensitive or VOC-sensitive due to the complexity of these photochemical processes. Therefore, effective ozone control strategies rely heavily on the accurate understanding of how ozone responds to reduction of NO x and VOC emissions, usually simulated by photochemical air quality models (e.g., Sillman et al, 2003;Lei et al, 2004;Mallet and Sportisse, 2005;Li et al, 2007;Chen et al, 2010;Tang et al, 2010;Xue et al, 2013;Goldberg et al, 2016). However, those model-based studies have inputs or parameters subject to large uncertainties that can affect not only the simulated levels of ozone but also the ozone dependence on its precursors.…”