As shown by the recently completed NKS-B projects MUD and FAUNA, the influence of meteorological uncertainties on long-range atmospheric dispersion calculations can be large, e.g. up to an order of magnitude depending on the weather situation, with significant implications for nuclear emergency preparedness and decision making. The question that the MESO project is going to answer is to what extent this also applies to short-range (mesoscale) dispersion models employed for nuclear emergency preparedness up to about two hundred kilometres from the source.
The assessment of such uncertainties is facilitated by recent developments in numerical weather prediction (NWP) modelling through the use of ensemble methodology. The computer resource demanding procedures are now being, or are planned to be, implemented at a number of national weather services. This development enables operational quantitative calculation of uncertainties of the concentration and deposition patterns from accidental releases of radionuclides to be used by nuclear decision-support systems (DSSs), which thereby provide a more comprehensive basis for the decision making.
Short-range atmospheric dispersion models differ from long-range models not only by the use of finer resolution terrain and land-use data, but also by the fact that short-range models may utilize weather radar data for simulation of wet deposition of radionuclides. Obviously, observational data, e.g. from radars, can be used only for hindcasting, but these data, which are expected to represent the precipitation intensity more accurately than NWP model data, are useful for nuclear emergency preparedness in the period of time until radiological monitoring data have become available. However, there are a number of uncertainties associated with such use of weather radar data. These include the use of a parameterization of the precipitation rate depending on the attenuation of the reflected radar signal, filtering of false radar echoes arising from e.g. clutter or flocks of birds, precipitation from low clouds not being registered by the radar beam, and precipitation evaporating before reaching ground. A new possibility is facilitated by the next-generation dual-pole Doppler radars, namely observed distinction between rain and snow which can have important consequences for short-range modelling since deposition of radionuclides differs substantially between rain and snow.
Thus, the MESO project has two work packages: one devoted to the study of uncertainties of short-range atmospheric dispersion forecasting involving the use of NWP model data only, the other focusing on hindcasting including the combined use of NWP model data and weather radar data.