Impact of meteorological uncertainty on dispersion predictions in complex terrain

The Weather Research and Forecast (WRF) atmospheric model is used to study the impact of meteorological uncertainty on dispersion predictions during the Diablo Canyon Nuclear Power Plant air tracer experiment (Dopptex). An ensemble modeling approach is utilized to generate a large data set of WRF simulations with varying reanalysis input, initialization time, model physics, and four dimensional data assimilation approaches. WRF results are compared with Dopptex meteorological observations that were not assimilated to quantify the variation in wind reconstruction skill among the atmospheric ensemble members.

The Lagrangian FLEXPART dispersion model, which is offline coupled to WRF, is used to generate dispersion predictions based on meteorological data from the atmospheric ensemble. Dispersion model predicted air concentrations are compared with Dopptex one hour averaged air concentration measurements. A statistical analysis is completed to study the variation of dispersion prediction skill based on different WRF ensemble members and to assess what meteorological model factors contribute the most to spread in the dispersion results. In addition, we assess the impact of source and dispersion related parameter uncertainties on concentration spread and compare these findings to the scale of meteorological uncertainty on Dopptex dispersion results.

The project (14-ERD-006) was funded by the Laboratory Directed Research and Development Program at LLNL, and was performed under the auspices of the U.S. Department of Energy by LLNL under Contract DE-AC52-07NA2734. LLNL-ABS-675728