Field experiments involving the release of reactive chemical species can be difficult and costly endeavors. As a way to plan for these exercises or to test operational readiness in case of an accidental release, exercises can be conducted that use virtual weather, dispersion, and chemistry scenarios. In this presentation, we describe a virtual weather scenario that created realistic meteorological data for input into transport, dispersion and atmospheric chemistry models. To obtain the required meteorological data, we began by running an Observing System Simulation Experiment (OSSE). OSSEs are designed to examine the impact of new and/or current observations on meteorological analyses and forecasts under differing and realistic virtual weather regimes.

An OSSE includes a nature run to provide the assumed truth and extracted simulated observations. These simulated observations are then incorporated into another model using a data assimilation cycle to generate the subsequent analyses and forecasts. The impact of assimilating temperature, winds, etc. from the nature run into the other model simulation can be assessed using subjective and objective verification. Additionally, the three-dimensional gridded nature fields provide the plume truth from the transport and dispersion models which can then be compared with the results from the assimilation runs. These results determine what impact, if any, the simulated observations have on subsequent plume analyses and forecasts. Wide ranging atmospheric chemistry release scenarios can be investigated that include different release rates and schedules, species, and locations with varying land use and topography. The atmospheric chemistry module is based on a puff model using LSODE as a solver where chemistry compounds of interest are easily appended to different choices of background mechanisms such as Carbon Bond or SAPRC. In this presentation, we will describe a virtual chemical release run during September 2006 over an urban area of the western U.S. Two different mesoscale models along with a dispersion model coupled with an atmospheric chemistry model provided the realistic scenarios.