Simulations of lightning-generated NOx for parameterized convection in the WRF-Chem model

Production of nitrogen oxides (NO_x) by lightning is an important precursor for ozone in the upper troposphere. In this study, the prediction of lightning flash rates and lightning-generated NO_x in the Weather and Research Forecasting model coupled with Chemistry (WRF-Chem) is evaluated using data from the Deep Convective Clouds and Chemistry (DC3) field campaign. The WRF-Chem simulations are performed at grid spacings of 15 km where convection is parameterized. Thus, lightning flash rate is predicted using the Price and Rind (1992) parameterization, which is based on cloud-top height and is predicted by the level of neutral buoyancy determined in the convective parameterization. By comparing the predicted flash rate to the Oklahoma Lightning Mapping Array data for specific DC3 case studies, we found for the May 29, 2012 severe storm in northern Oklahoma that the lightning parameterization needed to be adjusted by limiting the flash rate location to regions with high cloud condensate content, in order to better match the horizontal spatial distribution and magnitude of observed flash rates. Comparison of predicted NO_x mixing ratios in the anvil region of the storm showed that the predicted NO_x is 1 to 1.5 times greater than aircraft observations. However, comparison of predicted concentrations to aircraft data collected in the convective outflow (over western North Carolina) 20-22 hours after active convection showed that modeled NO_x had similar values to observations.