Page 57 - Summer 2021

were recommended for impulsive noise, regardless of propagation distance.
This oversimplification led to practical challenges and unintended consequences for both those establish- ing and enforcing regulations and for those seeking to comply with them. For instance, it proved difficult to effectively categorize certain rapid-onset sonar signals.
Additionally, when onset criteria levels were applied to relatively high-intensity impulsive sources (e.g., pile driv- ing), TTS onset was predicted in some instances at ranges of tens of kilometers from the sources. In reality, acous- tic propagation over such ranges transforms impulsive characteristics in time and frequency (see Hastie et al., 2019; Amaral et al., 2020; Martin et al., 2020). Changes to received signals include less rapid signal onset, longer total duration, reduced crest factor, reduced kurtosis, and narrower bandwidth (reduced high-frequency content).
A better means of accounting for these changes can avoid overly precautionary conclusions, although how to do so is proving vexing. Several recent studies address this issue, which was taken up by a noise sources subgroup of the reassembled criteria panel.
Hastie et al. (2019) evaluated field recordings of seismic airguns and pile-driving at different ranges and compared received parameters (rise time, peak pressure and duration interactions, duration, and crest factor). They found range- dependent changes and that signals generally began to lose impulsive characteristics at ranges of several kilometers, especially beyond 10 km. However, they noted that the lack of direct hearing data precluded determining the most appropriate metric(s) for defining impulsiveness.
Martin et al. (2020) suggested a novel effective quiet calcu- lation to compare accumulated exposures to group-specific TTS onset estimates. After calculating crest factor, kurto- sis, and the Harris impulse factor (applied by Southall et al., 2007) for noise exposures measured or estimated at variable ranges, Martin et al. (2020) compared values with those associated with residual auditory effects in humans and some terrestrial mammals. They concluded that kurto- sis (see Qiu et al., 2020) is the most appropriate indicator of impulsivenessfortheirscenariosbutsimilarlynotethelack of direct supporting empirical data in marine mammals. Martin et al. (2020) further proposed a kurtosis threshold for distinguishing impulsiveness using auditory studies in humans (Hamernik et al., 2010). They suggested that this
threshold be used for marine mammals, noting the general consistency from a study in one species (harbor porpoise; see bit.ly/31oncy2) using recorded pile-driving noise at a single kurtosis level (Kastelein et al., 2016) that yielded TTS results for harbor porpoises similar to those predicted by the impulse criteria from Southall et al. (2019a). Finally, based on these assumptions and assessments, Martin et al. (2020) postulated that propagation-related changes for signals are effectively not relevant considerations in a real- world application.
Although providing new data on propagation effects, developing novel applications of existing concepts, and proposing several testable hypotheses, Martin et al. (2020) is arguably overreaching and consequently potentially misleading. At present, there are no properly designed, comparative studies evaluating TTS for any marine mammal species with various noise types, using a range of impulsive metrics to determine either the best metric or to define an explicit threshold with which to delineate impulsiveness. More substantial research similar to that in human studies (Hamernik, 2010) for multiple species in controlled conditions is needed to support such explicit conclusions.
Such studies should be conducted under testing condi- tions involving more realistic presentation of stimuli than can be achieved in traditional laboratory settings. The most needed relevant studies would avoid artificial near- field, highly reverberant testing contexts. These are often associated with the use of a scaled source projector very close to the animals, intending to simulate high-intensity sources by matching received levels at some presumed range. This does not account for range-induced changes in spectral and temporal features. The most appropriate environments for such studies are arguably less reflec- tive open ocean conditions and with full-scale projectors located at several to tens of kilometer ranges over which these potential transitions may be occurring. However, such studies are challenging to conduct, especially for multiple individuals and species, and the desired data may thus be unavailable for some time.
Although the predictions of Martin et al. (2020) may ultimately be validated for marine mammals, a sim- pler interim approach is possible. Namely, members of the noise sources subgroup noted that the presence of high-frequency noise energy could be used as a proxy
Summer 2021 • Acoustics Today 57