Recent advances in global positioning system (GPS) and satellite tracking technologies have enabled unprecedented collection of locational information for mobile objects. Though it is possible to capture highly accurate spatial information describing the movements of people, animals, vehicles, and other types of moving objects, the theory and methods available to support analysis of such data remain limited. Within geographic information science, the area known as "time geography" is the foundational body of knowledge for movement pattern analysis. However, as it is currently theorized and understood, traditional time geography suffers from one critical shortcoming: it is based on a discrete representation of space that does not permit the mapping and quantification of an object's movement distribution with uncertainty. While traditional time geography is adept at describing the spatial extent of an object's possible movement over a period of time, it does not reveal the likelihood of where the object was located within this boundary. This is an important question in many substantive domains, particularly in wildlife ecology, where quantifying and mapping the space-use characteristics, or home ranges, of animals is a key management task. Building on recent work in spatial statistics and time geography, this project develops probabilistic measures of time geography that can be used to analyze tracking data. In particular, several time-geographic density estimators will be developed in order to generate continuous probability density functions from tracking data. Formulation of these methods will proceed along several tracks that include examining appropriate sampling schemes, tracking intervals, distance-weighting functions, velocity parameter specifications, moving object inactivity, and positional uncertainty. To support the development of these new methods, the investigators will collect a rich set of high-frequency baseline data from tagged Muscovy ducks (Cairina moschata) in the State of Florida. This data will also be used as a basis for evaluating new and existing methods of animal home range estimation. Lastly, the project will adapt the developed probabilistic time geographic methods for use in network space.
The project will advance basic knowledge in geographic information science regarding how the movement patterns of mobile objects can best be analyzed. Several new spatial statistical techniques will be produced. A baseline dataset describing animal movements will be collected and shared with the other researchers. Although the project applications primarily are ecological in nature, developed methods will be relevant to a number of fields, as spatially-referenced moving object data are increasingly used in other areas such as crime, transportation, and homeland security. In addition to the research outcomes being incorporated into the PI's course offerings, the project will directly support at least two graduate students.
