Multicopter bat surveillance : noise emission versus the hearing abilities of bats

Abstract

Covertly monitoring bats to obtain information on their population size and species in hard-to-reach areas is challenging. By attaching an acoustic sensor to a multicopter (MC), a large area can be covered without the limitations experienced by sensors mounted at fixed locations. It has previously been shown that the emitted MC noise may either attract or deter the bats. Although the cause of this effect is still unclear, we reveal that some bat species can perceive the emitted noises even far away from the MC, whilst other bat species do not hear the MC. We measured the sound pressure level of multiple drone drivetrains up to a frequency of 100 kHz in the laboratory and applied a two-stage environmental damping model of sound. A cylindrical model for free field sound propagation was used to simulate the worst-case geometrical attenuation of sound propagation in air. This model depends on the radiation characteristic and the distance from the point source. In addition, the atmospheric model simulates the effect of molecular absorption in the air and is dependent on the distance from the object, relative humidity, atmospheric pressure, and temperature. We further investigated the effect of temperature and humidity on noise attenuation for the atmospheric model and the best possible scenario is recommended. The auditory brainstem response (ABR) data of 38 bat species were overlaid with our simulated attenuation results to draw conclusions in which frequency ranges and to what extent the emitted sound pressure level could possibly be heard by the bats. We can conclude that there are many bat species that will not hear the MC, but there are also some bat species that will hear the MC. The analysis of several drivetrain combinations indicates that the propeller is the most noise-emitting part of MCs. The noise emission of MCs is clearly dependent on the thrust level, size, and shape of the propeller.