Much of the world's open coasts are formed by estuaries and lagoons which are highly important for ecological reasons. Since the rich organic materials are carried in by the stream water, a favorable environment gets formed for plankton, the other creatures thriving on planktons and the fish feeding on them. On the other hand, the water, the reeds and other vegetation, the fish and the organic material that has settled provide the birds with an ideal environment. In this context, the problems with pollution in lagoon systems are important risk factors for the ecological balance. The purpose of the study is to generate a high quality computational network for the risk analysis of a special lagoon in order to construct a solid foundation for the next step in our study. A reduction of the number of elements is aimed at while the underlying physical domain is adequately represented via several grid refinement criteria. Furthermore, standardizing the method to be used as the calculation network, and utilizing its highly accurate risk predictions for all lagoon systems are also aimed at.