A Simplified Approach for Solution of Time Update Problem during Toxic Waste Plume Spreading in the Atmosphere

Reliable and up to date information represents principal prerequisite for effective management of intervention operations targeted on emergency situations during accidental releases of harmful substances into the atmosphere. Promising way of this trend insists in development of assimilation techniques for improvement of model prognosis reliability on basis of optimal blending of predictions with observations incoming from terrain. In this paper we are concentrating on the forecast procedure which generates the new state predictions from the initial conditions standing for previous time. Two problems are encountered here. At first, the resulting predictions from the previous time step can be assimilated with incoming data before the next time predictions are evaluated. The initial conditions for the next time are then modified in advance and trajectory information accumulated from the same beginning can become fuzzy. So, the trajectory models can have troubles how to propagate plume in the next time. At second, reliable and up to date information at medium distances from the source of pollution represents basic inevitable conditions for effective management of intervention operations targeted on consequence mitigation during emergency situations. But justifiability of Gaussian plume model application in medium range distances is questionable. A simplified solution offers segmented Gaussian plume model (SGPM) which can account stepwise for the time dynamics of the admixture release and hourly changes of meteorological conditions. The extension of modelling to medium distances is facilitated by availability of the new quality of gridded spatial short-term meteorological forecasts. The Czech meteorological service provides 3-D data in ALADIN format for medium domain 160 km × 160 km around each nuclear power plant (NPP) in the Czech Republic. In the following text we are describing the SGPM application in longer distances taking into account the new gridded meteorological predictions. The comparison with former methodology is illustrated on one real meteorological situation from June 25, 2008.