The paper describes the models and program structures. methods and dat
a in the German real-time decision support system RESY, which has been
developed with the support of the Ministry of Environment (BMU), Germ
any. It is part of RODOS, the CEC real-time on-line decision support s
ystem for nuclear emergencies, whose first prototype version was demon
strated during the workshop. RESY, the system for near-range early pha
se applications. has a three steps concept of information processing:
analysis and prognosis of the radiological situation, simulation of co
untermeasures and assessment of their consequences, and evaluation of
consequences to determine the most suitable countermeasures. To that p
urpose, the following models have been developed and implemented in RE
SY: (1) a fast gaussian puff model for atmospheric dispersion calculat
ions and the determination of the concentration and radiation fields i
n the environment of the site after releases of up to 15 nuclides; (2)
simulation models for early emergency actions like sheltering and eva
cuation, distinguishing up to five action areas with different time pa
tterns of actions defined by intervention levels; methods for the calc
ulation of doses with emergency actions; (3) models for calculating de
terministic and stochastic somatic health effects and the economic cos
ts associated with them and the underlying emergency actions; (4) a ru
le based expert system containing components of multi-attribute value
techniques for evaluating and ranking alternative courses of emergency
action.