Biological explanation of Tau velocity

[Vilis Nams]

I understand a biological explanation of Tau location (~time to cross home range), but what would be a biological explanation of Tau velocity? The "timescale over which autocorrelation decays" is an intuitive, but statistical one. What would this mean from an animal's point of view? I understand that  Tau velocity affects path tortuousity, but  since tortuousity is a scale-dependent measure, Tau velocity can't be a measure of it. 

Also,  how would estimates of Tau velociy be affected by inactivity? I have datasets for leopards with locations taken 15 min apart. During a certain proportion of those  locations the leopards are not travelling, and thus locations only reflect error. Do periods of inactivity bias estimates of Tau velocity? If so, can I remove the bias by removing the periods of inactivity (and, of course, adjusting the times appropriately)?

Thanks, Vilis

[Christen Fleming]

Hi Vilis,

tau-velocity can be thought of as an average "step" time, or timescale under which movement is ballistic. sqrt(sigma*tau[velocity]/tau[position]) is a more familiar average "step" length, where sigma is the location variance.

I would say that tau-velocity is a measure of tortuosity, as for sampling intervals dt >> tau[velocity] the paths appear fractal and for sampling intervals dt << tau[velocity] the paths are continuous with continuous derivative (smooth). I think there can be other kinds/measures of tortuosity, though.

The models currently in ctmm are stationary, so fitting one stationary model to two behavioral states will return an average behavioral state model. This can bias the parameter estimates and segmenting can improve things.

Best,

Chris