Importance of Using Observations of Mixing Depths in order to Avoid Large Prediction Errors by a Transport and Dispersion Model

Abstract

The mixing depth of the boundary layer is an input to most atmospheric transport and dispersion (ATD) models, which obtain mixing depths in one of four ways: 1) observations by radiosondes, sodars, or other devices; 2) simulations by regional or mesoscale meteorological models; 3) parameterizations based on boundary layer similarity theory; or 4) climatological averages. This paper describes a situation during a field experiment when exceptionally low mixing depths persisted in the morning and led to relatively high observed tracer concentrations. The low mixing depths were caused by synoptic effects associated with a nearby stationary front and the outflow from a mesoscale thunderstorm complex located 20?50 km away. For the same time period, the ATD model-parameterized mixing depth was a factor of 5?10 higher, leading to predicted concentrations that were less than the observations by a factor of 5?10. The synoptic situation is described and local radiosonde and radar observations of mixing depth are presented, including comparisons with other more typical days. Time series of local observations of near-surface sensible heat fluxes are also plotted to demonstrate the suppression of turbulence by negative sensible heat fluxes during the period in question.