Evaluation of turbulent mixing in HYSPLIT using measurements from controlled tracer experiments

HYSPLIT, a dispersion model developed by NOAA's Air Resources Lab, has different options to estimate the turbulent mixing depending on the availability of stability and turbulent parameters in the meteorological data used to drive the dispersion simulation. The estimation can be based on the stability function, diagnosed by meteorological variables such as friction velocity and mixing height. However, if the meteorological model does not include these variables, temperature and wind soundings can be used to diagnose the boundary layer stability parameters. Other ways to estimate the mixing include the use of the total turbulent kinetic energy or mixing diffusivity. This study aims to understand the mixing characteristics generated by different estimations of the standard deviation of the turbulent velocity. Dispersion simulations using WRF meteorological data were conducted to simulate two tracer experiments - the Project Sagebrush phase 1 (PSB1) and the Cross Appalachian Tracer Experiment (CAPTEX). PSB1 consisted of tracer releases aiming for the sub-kilometer scale transport with near neutral or unstable stability conditions. The sampling network for measuring tracer concentrations was set within a 3-km range from the release location and samples were taken in 10-minute averages. Flux measurements including turbulent velocity variance are available for comparisons with model estimations corresponding to different mixing options. On the other hand, CAPTEX, consisted of six 3-h releases aimed to simulate the long-range transport and diffusion of pollutants. The comparisons of dispersion results with measured concentrations provide a unique insight to assess the performance of the dispersion model using different turbulent mixing methods.