a ray-tracing method is not appropriate for the description of narrow waveguides (Küsel and Siderius 2019). Modeling was performed in two-dimensional planes (distance x depth), considering a single sound source described as a monopole. The data used as input for the numerical modelling was the data available in regional noise registries, hosted at ICES. To consider important regional differences, the North Sea, the Baltic Sea, and the Ligurian Sea in the Mediterranean Sea were selected as areas for case studies. For each of the case study regions, a sound- intensive activity was selected that represents a realistic impulsive noise event reported in the region. In addition to the regional perspective of each case study, a speci?c technical focus was given to challenges of quantitative assessments for each of the case studies. Test case study 1 addressed seismic airgun arrays in the Mediterranean Sea. Airgun blasting is an activity commonly used by the oil and gas industries to locate reservoirs deep beneath the sea?oor, involving high levels of impulsive sound. The context of the Mediterranean Sea requires modeling in very deep waters. An objective of this test case was to assess the sensitivity of acoustic propagation with respect to regional variability in bathymetry and bottom type, as well as seasonal variability in sound propagation conditions in water as detailed in Table 2. The seasonal variation of oceanographic conditions had a strong impact on sound propagation with up to 20 dB difference in broadband level in the surface layer between the summer and winter season based on results from the modeling of scenarios 01A and 01B. The effect of sound speed variability between summer and winter concerned mainly frequencies above 1000 Hz in the upper water layer (up to 30 dB difference at 1000 Hz). This is related to the proximity of the source to the surface (with a source depth of 6 m), as well as by the fact that only the surface layer is affected by seasonal variations in sound speed. The depth-dependent gradi- ent of sound speed in water depends on the temperature, pressure, and salinity of the water.It causes an oriantation of the wavefront by refracting the sound rays toward the surface or toward the seabed. Hence, sound propagation becomes affected by the state of the sea at the surface, or by the geo-acoustic properties of the seabed and its varying re?ectivity. This phenomenon adds uncertainty to the estimation of sound levels at the receiving point, and requires that the environment is accurately described. Figure 3 shows an example of seasonal differences in the sound propa- gation at 500 Hz in the deep water context of the case study region in the Mediter- ranean Sea. Based on the modelling scenarios 01C and 01D, representing two different propagation directions from the same sound source, the bathymetry and variability of the sea?oor properties have a strong impact on sound propagation. One one hand, the propagation of sound may be limited or interrupted in shallow water, e.g., when the sound ?eld reaches the coast. On the other hand, the interaction of sound with the seabed increases for decreasing water depth and the sound propaga- tion becomes affected by the geo-acoustic properties of the bottom. This caused an increase in sound levels of up to 5 dB based on the scenario results. In order to study the in?uence of frequency discretization on the modeling results, the scenario 01C for a distance of up to 30 km was repeated by multiplying the frequencies in the calculation of the frequency range between 32 and 1000 Hz. To EU Marine Strategy Framework Directive-Compatible Approaches for. . . 11