GPS measurement error influences on movement model parameterization

Global positioning system (GPS) technology has increased the accuracy and efficiency in recording animal locations and has provided data used to parameterize movement models. Although numerous studies have investigated the quality and accuracy of location data associated with different brands of GPS collars, none of these studies has investigated the influence of measurement error on the parameters used to create movement models. We used Monte Carlo simulation to quantify the measurement error for estimates of turning angle and step length as a function of distance between consecutive locations. We show that estimates of turning angle and step length are accurate only when the distance between two locations is large relative to the measurement error. Estimates of turning angle are particularly susceptible to error for short step lengths. The consequences of choosing poor data-collecting schedules are discussed, and suggestions for designing appropriate data-collecting schedules are provided.