Application of Lagrangian particle technique to modeling atmospheric transport in different scales: from LES to RAMS to GCM

The Lagrangian particle modeling technique is very flexible and attractive, though potentially expensive, approach to simulate atmospheric transport. Particles can carry different information including time and location of their release as well as characteristics of the atmosphere along their trajectories. Therefore, this approach is well suited for inverse problems in atmospheric transport: source apportionment, emission rate estimation, and interpretation of atmospheric measurements.

A common framework based on this approach has been developed to simulate atmospheric transport in different scales from local to global. In the most general case, particles are released continuously from the entire volume of modeling domain that provides a full Lagrangian representation of atmospheric flow. In a long range or global transport, particle release is usually limited to the Earth’s surface. The kernel density estimator technique is used to calculate concentrations for selected source areas, as well as concentration and flux footprints for selected receptors. Geometry and time characteristics of both sources and receptors can be changed without a need to rerun the dispersion model. The essential part of our modeling approach is to derive subgrid scale velocity components of particle motion consistently with subgrid scale parameterizations implemented in a meteorological model. The Lagrangian particle model has been coupled with several different meteorological models and used in a range of applications: (1) two LES codes: dispersion in cloud-topped boundary layer, Lagrangian characteristics of turbulence, flux measurement footprints; (2) CSU RAMS (Regional Atmospheric Modeling System): air pollution studies in meso and regional scales, interpretation of concentration measurements, and, (3) CSU GCM (Global Circulation Model): global transport of CO2 and other tracers.