ROBUST STATE-SPACE MODELING OF ANIMAL MOVEMENT DATA

摘要: Remotely sensed tracking data collected on animal movement is vastly un- derutilized due to a lack of statistical tools for appropriate analysis. Features such that make analysis particularly challenging include the presence estimation errors are non-Gaussian and vary in time, observations occur irregularly com- plexity underlying behavioral processes. We develop state-space framework simultaneously deals with these features demonstrate our method by analyzing three seal pathway sets. show how known information regarding error distributions can be used improve inference process(es) frame- work provides powerful flexible fitting different models data.