MB FUTURE SCIENCE BRIEF 
Although many international conventions have a clause that 
exempt military activities, there is a requirement that Defence 
organisations should ‘endeavour to ensure that such activities are 
zonducted in a manner that is compatible, so far as reasonable 
and practicable, with the objectives of that convention or 
regulation' (see e.g. MSFD chapter 1, art. 2-2). Most national military 
organizations have mitigation measures in place to minimise risk 
to marine mammals when deploying military mid-frequency sonar, 
Several institutions and nations such as NATO (NATO, 2018), the UK, 
Norway, the Netherlands, Spain, Australia, Germany and Italy have 
also adopted other mitigation measures (Dekeling et al., 2016). The 
main elements are risk assessment, avoidance of sensitive areas, 
considering the source level needed for a specific activity and soft- 
start / ramp-up (UN, 2018b). 
4.4 Emerging technologies and methods 
Since the last EMB Position Paper in 2008 there have been 
several technologies or approaches that have advanced 
understanding of the effects of noise on marine organisms 
and have helped with field studies, risk assessment and 
noise management. Some of these are listed here according 
to their relevance to the risk framework (see Figure 1) 
4.4.1 Exposure assessment 
Just a decade ago, impact assessments commonly applied rather 
simplistic calculations for sound transmission to estimate impact 
“anges for different sources. Now numerical modelling which uses 
nore advanced mathematical models and include environmental 
data such as bottom topography and sound speed vertical profiles, 
is common practice. Its application has much improved noise 
assessments but more is needed to apply the appropriate models 
to the specific circumstances (see review by Farcas et al., 2016). 
The emergent uptake in the use of autonomous underwater 
vehicles offers a potential cost-effective way to make extensive 
sound recordings over large areas, and to further validate noise 
propagation models. In terms of quantifying the number of 
animals exposed, statistical modelling has increasingly been used 
in Identification of animal hot-spots for the identification of risk 
areas. For example, using dynamic habitat modelling, animal 
distribution can be correlated to environmental variables (e.g. 
current speed) to identify those habitat variables that drive the 
distribution of marine fauna in space and time (Heinänen et al, 
2018). Furthermore, agent-based models (ABM) simulate the 
movement of individual animals in response to habitat drivers 
and pressures such as noise, providing more realistic assessments 
of exposure to sound than those based on stationary animals 
Double Bia Bubble Curtair in getion. used to mitigation sound propagation fram pıle-Anmı