I've recently heard a lot of people talk about kurtosis (a measure of signal 'peakiness') in the context of assessing the impacts of noise exposure. (The higher the kurtosis score, the more the signal has impulsive characteristics).
Including kurtosis-corrected SPLs in an acoustics impact assessment, for example, has been shown to better predict hearing damage in humans exposed to 'complex sounds' (see Goley et al. 2011, JASA; and Zhao et al. 2010, Ear & Hearing). However, current standards for noise exposure (in humans) doesn't take this into account; current ISO standards only consider the noise duration and overall SPL.
This approach has been applied to evaluating underwater sound (Muller et al. 2020, JASA) , and very recently, it has also been used to better predict the onset of hearing threshold shifts in harbour porpoises (von Benda-Beckmann et al. 2022, JASA). Specifically, kurtosis corrections were applied to frequency-weighted SEL (sound exposure levels), and this lead to more consistently predicted TTS (temporary threshold shift) onset thresholds for porpoises. This seems like a really useful tool!
The discussion (in von Benda-Beckmann et al. 2022, JASA) only briefly speculates on how broadly the kurtosis-correction method can be applied, and I am wondering aloud (to an audience with a few marine mammal auditory specialists, I'm sure), that given the different functional hearing groups of marine mammals, should we expect the kurtosis-correction method to be of equal usefulness for these groups? If not, why not?