Neural Network Models as Decision-Support Tools for Road Traffic Noise Management in Port Cities: A Comparison with Standard Prediction Methods

Road traffic is the main source of exposure to environmental noise in Europe and Directive 2002/49/EC requires the competent authorities to evaluate and manage it. In port cities, a significant portion of this exposure is generated by road traffic induced by ferry embarkation and disembarkation operations: a pulsating and highly non-stationary source, which standard forecasting methods, designed to return long-term average indicators, describe only in aggregate form. This work compares two forecasting approaches applied to the same waterfront of Olbia (Sardinia, Italy) and addresses a question of an operational rather than metrological nature: which model provides the information that a specific noise management decision actually needs? A physical model based on ISO 9613 and the Common Noise Assessment Methods in Europe (CNOSSOS-EU) methodology was implemented in CadnaA and calibrated on continuous monitoring data conducted in three receptor locations during a low and a high season. A nonlinear autoregressive artificial neural network, developed and experimentally validated previously on the same waterfront, was used to predict the equivalent sound pressure level from traffic flow data together with the autoregressive feedback of the previously observed noise sequence. The Lday, Levening and Lnight indicators returned by the two models are compared, together with the required input data, the spatial and temporal resolution, the latency and the installation and update costs over the five-year cycle of the strategic acoustic mapping. Against the measured levels in the same hours, the physical model reproduces the day period with a mean absolute error of 1.1 dB and the evening period with 2.8 dB, whereas in one night-time hour it under-estimates the measurement by 10 to 12 dB simultaneously at all three receiver positions; the calculation times vary between 3 and 36 h for the physical model against an almost-real-time inference for the neural one. The two approaches are complementary and not alternative: the physical model remains necessary for planning and for scenarios not yet implemented, while the operational and intraday management of noise in a ferry port is better served by data-driven prediction. Beyond the site examined, the results give port and municipal authorities a criterion for choosing the prediction instrument according to the decision it has to support, and show how a monitoring infrastructure already installed for other purposes can be reused to extend noise management from the five-year strategic cycle to the operational scale of a single ferry call.