The effect of vertical diffusivity on photochemical model estimates of tropospheric ozone

Vertical diffusivity (K_v) is used to parameterize the subgrid scale vertical mixing in three-dimensional photochemical models. These models are commonly used for regional-scale analyses of ozone formation, transport and fate. K_v is not directly output by the meteorological models used to generate input to the photochemical models, but must be diagnosed by differing methods. The issue of the effect of K_v on photochemical models has not been extensively investigated. This paper discusses the effect of alternative K_vs on photochemical model performance and the practical implications for interpretation of modeling results from regional and local ozone modeling studies.

The UAM-V version 1.24 model was applied to the urban areas of Nashville and Memphis, Tennessee, and to a large NO_x point source in middle Tennessee. High and low values of K_v were used in 4 km grid resolution model runs. The model output for Nashville is compared for two days where in situ airborne measurements were available from a helicopter that was flying in a grid-like pattern to characterize the concentration fields. Additional comparisons were made for the point source plume between model and measurements by the helicopter.

Variation of K_v was found to significantly affect the NO_y and O₃ concentration fields, sometimes in unexpected fashion. For example, the point source plume as actually modeled to mature faster with lower vertical diffusivity.