Probabilistic models of pollution of the surface waters of large marine areas

Abstract

Version of Monte-Carlo model for calculation of propagation of contamination in large marine areas with taking into account a coastline is presented. Based on long standing database of sea currents, bathimetry, and mixed layer depth, the probabilistic model of zoning marine areas by concentration of pollutant is offered. In accordance with modern IAEA recommendations general principles of estimation of radiation situation on radiation hazardous object in normal expluataion mode, conclusion of normal expluatation mode, or in accident mode exists. Practical methodology of safety case and tactics of monitoring of state of radiation hazardous object should rely not only on measuring, but also on model calculations and forecasts. Optimal tactics of monitoring includes definition of probabilistic picture of possible pollution different marine areas assuming hypothetical leak of pollutant with violation of defence barriers. One way of building such probabilistic pictures is execution of multi-variant calculations using long standing databases. Different calculations in multi-variant mode executes assuming different accident start time, in advance models multi-variant dynamic of marine area polluton. Not only maximum values of pollution to be interested, but and statistical estimations with different confidence levels (for filtration of rarery big values) for hypothetical leak of pollutant. In the base of the numerical algorithm of proposed technique lies Monte-Carlo method using lagrangian particles with non-zero sizes.