Quantification of the effect of modeled lightning NO₂ on UV–visible air mass factors

Abstract. Space-borne measurements of tropospheric nitrogen dioxide (NO₂) columns are up to 10x more sensitive to upper tropospheric (UT) NO₂ than near-surface NO₂ over low-reflectivity surfaces. Here, we quantify the effect of adding simulated lightning NO₂ to the a priori profiles for NO₂ observations from the Ozone Monitoring Instrument (OMI) using modeled NO₂ profiles from the Weather Research and Forecasting–Chemistry (WRF-Chem) model. With observed NO₂ profiles from the Deep Convective Clouds and Chemistry (DC3) aircraft campaign as observational truth, we quantify the bias in the NO₂ column that occurs when lightning NO₂ is not accounted for in the a priori profiles. Focusing on late spring and early summer in the central and eastern United States, we find that a simulation without lightning NO₂ underestimates the air mass factor (AMF) by 25 % on average for common summer OMI viewing geometry and 35 % for viewing geometries that will be encountered by geostationary satellites. Using a simulation with 500 to 665 mol NO flash⁻¹ produces good agreement with observed NO₂ profiles and reduces the bias in the AMF to  <  ±4 % for OMI viewing geometries. The bias is regionally dependent, with the strongest effects in the southeast United States (up to 80 %) and negligible effects in the central US. We also find that constraining WRF meteorology to a reanalysis dataset reduces lightning flash counts by a factor of 2 compared to an unconstrained run, most likely due to changes in the simulated water vapor profile.