Three Dimensional Flow-field Reconstruction using LIDAR Measurements for Near-Range Atmospheric Dispersion

Near-range atmospheric dispersion modelling is an essential part in licensing new nuclear installations as well as in nuclear and radiological emergency preparedness and response. In general, the impact of radioactive releases to the atmosphere is highest close to the release point and measurement stations surrounding nuclear installations are the first off-site tools delivering information on on-going releases. Many dispersion models have been developed for the near-range from simple Gaussian models towards highly complex techniques such as Computational Fluid Dynamics[1]. However, the estimation of the local flow-field is crucial in near-range atmospheric dispersion. This is very challenging, since to date, measurement techniques that provide three-dimensional time-resolved flow-field data in the atmospheric boundary layer do not exist. Although Doppler laser LIDAR profilers can be used with different scanningpatterns to provide information on the flow field, it is still a technique one-dimensional in nature, but can provideessential input, if combined with Large Eddy Simulations (LES) to perform a 3D flow-field reconstruction. Bayesianinference and four-dimensional variational data assimilation (4DVar) methods in combination with a LES model and LIDAR data will be explored to reconstruct the 3 dimensional flow field. This flow field will be further input to calculate radionuclide concentration fields and ambient dose rates at specific positions for well characterized releases during normal operations at certain facility allowing to test the methods developed. In a further stage the concentration and/or dose rate measurements can be added in the data assimilation methodologies developed leading to a combined flow-field and concentration field reconstruction, including improved source identification and source term estimation.