Farina A., Galaverna P., Truffelli G.

Whilst currently available programs for computing noise maps become even more complex, there are cases where the usage of highly detailed algorithms and procedures is meaningless, because the input data for such processing is simply not available. So it is often seen that a very complex and theoretically very accurate simulation system is driven to provide objectionable results, because the user had to 'guess' for all the required input data which were not known. The paper describes the algorithms employed in the program Citymap 3.0. This software is very simple to use, requires just a minimal geometrical description of the site, of the sources and of the receiving areas. The algorithms employed are stripped-down versions of the computational methods currently mandatory in Italy, after the EC directivity on environmental noise has been converted into Italian Law DL194 of 19/08/2005. Namely, the propagation algorithms are fully compliant with NMPB-Routes-96(SETRACERTU-LCPC-CSTB) for the road noise, with Reken-en Meetvoorschrift Railverkeerslawaai '96, Ministerie Volkshuisvesting, Ruimtelijke Ordening en Milieubeheer, 20 November 1996 for the railway noise, and with ISO 9613/2 for industrial sound sources. The program imports easily geometrical data from most CAD and GIS packages, and asks only for minimum information about sound sources and receivers. Thank to a new, ultrafast computational engine, the program can map with high spatial resolution large areas, even in presence of a lot of obstacles and on arbitrarily-shaped terrains. As the variance due to introduction of guessed input data is reduceed, the results are often more reliable than those of theoretically more accurate models, which are usually very sensitive to errors in the input data. Some comparative results are presented, verified by measurements in the field, for demonstrating the above assertion.