Positioning of aquatic animals based on time-of-arrival and random walk models using YAPS (Yet Another Positioning Solver)

Henrik Baktoft, Karl Øystein Gjelland, Finn Økland, Uffe Høgsbro Thygesen

Research output: Contribution to journalJournal articleResearchpeer-review

316 Downloads (Pure)

Abstract

Aquatic positional telemetry offers vast opportunities to study in vivo behaviour of wild animals, but there is room for improvement in the data quality provided by current procedures for estimating positions. Here we present a novel positioning method called YAPS (Yet Another Positioning Solver), involving Maximum Likelihood analysis of a state-space model applied directly to time of arrival (TOA) data in combination with a movement model. YAPS avoids the sequential positioning-filtering-approach applied in alternative tools by using all available data in a single model, and offers better accuracy and error control. Feasibility and performance of YAPS was rigorously tested in a simulation study and by applying YAPS to data from an acoustic transmitter towed in a receiver array. Performance was compared to an alternative positioning model and proprietary software. The simulation study and field test revealed that YAPS performance was better and more consistent than alternatives. We conclude that YAPS outperformed the compared alternative methods, and that YAPS constitute a vast improvement to currently available positioning software in acoustic telemetry. Additionally, in contrast to vendor-supplied solutions, YAPS is transparent, flexible and can easily be adapted and extended for further improvements or to meet study specific requirements such as three-dimensional positioning.
Original languageEnglish
Article number14294
JournalScientific Reports
Volume7
Issue number1
ISSN2045-2322
DOIs
Publication statusPublished - 2017

Fingerprint Dive into the research topics of 'Positioning of aquatic animals based on time-of-arrival and random walk models using YAPS (Yet Another Positioning Solver)'. Together they form a unique fingerprint.

Cite this