A nonhydrostatic cloud-resolving model with explicit cloud microphysics is used to simulate a thunderstorm which occurred over the northern Colorado plains in July 1996 during the STERAO-A Deep Convection experiment. The cloud model was initialized nonhomogeneously with fields from the NCEP Eta forecast model. The simulation produces a storm with characteristics (e.g., cloud-top height, horizontal dimension, radar reflectivity, etc.) similar to that of the observed storm. NO, the most important trace gas for photochemical production of ozone in the free troposphere, is
produced by lightning in addition to being convectively transported from the boundary layer. The model uses observed cloud and ground lightning flashes and a parameterization of flash placement and NO production. Wind fields from the cloud model are used to
redistribute the lightning NOx, NOx from other sources, and other trace gases important for ozone photochemistry. The resulting NOx distribution is compared with aircraft observations in and around the anvil of the storm. This comparison suggests that the common assumption that a cloud flash produces one tenth as much NO as a ground flash does not appear valid for this storm. Photochemical ozone production rates in storm outflow are compared for model calculations assuming (1) lightning as the only NOx source and (2) a combination of lightning and boundary layer sources.
Symposium on Interdisciplinary Issues in Atmospheric Chemistry