Precision and bias in estimating detection distances for beaked whale echolocation clicks using a two-element vertical hydrophone array

TitlePrecision and bias in estimating detection distances for beaked whale echolocation clicks using a two-element vertical hydrophone array
Publication TypeJournal Article
Year of Publication2017
AuthorsBarlow J, Griffiths E.T
JournalJournal of the Acoustical Society of America
Volume141
Pagination4388-4397
Date Published2017/06
Type of ArticleArticle
ISBN Number0001-4966
Accession NumberWOS:000403467300042
Keywordscuviers; passive acoustics; sensors; ziphius-cavirostris
Abstract

Detection distances are critical for cetacean density and abundance estimation using distance sampling methods. Data from a drifting buoy system consisting of an autonomous recorder and a two-element vertical hydrophone array at similar to 100-m depth are used to evaluate three methods for estimating the horizontal distance (range) to beaked whales making echolocation clicks. The precision in estimating time-differences-of-arrival (TDOA) for direct-and surface-reflected-path clicks is estimated empirically using repeated measures over short time periods. A Teager-Kaiser energy detector is used to improve estimates of TDOA for surface-reflected signals. Simulations show that array tilt in the direction of the source cannot be reliably estimated given this array geometry and these measurements of TDOA error, which means that range cannot be reliably estimated. If array tilt can be reduced to less than 0.5 degrees, range can be reliably estimated up to similar to 3000 m. If array depth is increased to 200m and array tilt is less than 1 degrees, range can be reliably estimated up to similar to 5000 m. Prior information on the depth of vocalizing beaked whales and estimates of declination angle can be used to precisely estimate range, but different analytical methods are required to avoid bias and to treat distributions of depth probabilistically.

DOI10.1121/1.4985109
Student Publication: 
No
sharknado