Validation of a parabolic equation method against long range blast wave measurements

This report is about the validation of a method, based on a Parabolic Equation model, for predicting acoustic noise. The method is designed for the low-frequency, long-ranging noise induced by heavy weapons and explosions. It is able to take detailed meteorological information into account, which is important for this application. Hundreds of relevant blast wave measurements are available from the Norwegian Trials campaign at Finnskogen. Fairly detailed weather data are also available from the campaign. We have compared our simulations with measurements from Finnskogen, and made the following main conclusions: 1. The PE model worked as well as can be expected given the parametric uncertainties. 2. Taking the weather into account is clearly necessary in noise mapping of heavy weapons. 3. The weather data improved the prediction accuracy of the PE model. 4. Prediction accuracy varied with weather type, season, distance and sound frequency. A total of 299 unique combinations of detonation events and propagation paths have been considered, some of which were simulated several times with different parameter settings. The selected events were from six summer days and three winter days. Propagation distances varied from 1 to 23 km. Wind and temperature profiles were obtained from measurements that used a tethered balloon. Temperature inversions, which strongly influence sound propagation, occurred during both the summer and winter experiments. In the process of selecting and analysing the shot recordings and weather data, we found that there were errors and issues in the database registration, which are specified in this report. Future users of the database should therefore especially consult our chapters on shot selection and weather data.