A field test of the accuracy of a passive acoustic location system

In this study, we tested the ability of a computer-based passive acoustic location system (ALS) to determine the two-dimensional locations of vocalizing animals. The ALS uses multi-track tape recordings to estimate locations based on arrival time delays between widely-spaced microphones. We tested the accuracy of ALS location estimates using tape recordings of wild free-ranging birds made with 4 microphones placed at the corners of a 40m square. We compared ALS location estimates for these birds with locations determined by surveying the locations of the perches birds vocalized from. ALS location estimates were typically less than 1m away from surveyed locations when birds vocalized within the square microphone array, rising to just over 2m for birds vocalizing from within 25m beyond the array boundary. Beyond 25m, accuracy diminished rapidly with increasing distance. ALS accuracy did not depend on the bird species located, but location estimates based on frequency-modulated tonal notes were more than twice as accurate than those made using relatively unmodulated notes from the same songs. Two examples are given to demonstrate the usefulness of the ALS in field studies.