The Movebank data model facilitates the integration of animal movement data across a range of species, study designs and tracking methods. The main concepts included in the model are the animal, tag, deployment and event. All data in Movebank are stored and managed within studies, which are created and managed by user accounts.
An animal refers to an individual of a known species or other taxon for which tracking or other sensor data have been collected. Because animals are referenced within studies, if multiple studies refer to the same animal, there will be multiple records of it in Movebank. In these cases it is best practice for users to assign the same identifier to the animal in each study. Users can store reference data with additional information about the animal, such as birthdate, nickname or sex. Locations and other sensor measurements are only linked to animals if the animals are related to a tag using a deployment. Within a study, the user-provided animal identifier is stored as individual-local-identifier in event data and as animal-ID in reference data. Within the reference data, animals may also be uniquely defined using animal-ring-id and animal-nickname. In addition, each animal has an internal Movebank database identifier that is sometimes shown.
The taxon (typically species) for each animal is defined using the Integrated Taxonomic Information System (ITIS, itis.gov). The user-provided taxon is stored as animal-taxon in reference data and as individual-taxon-canonical-name in event data. Additional information, such as a subspecies or names not supported by ITIS, can be stored in the reference data as animal-taxon-detail.
A tag represents a specific tag. A tag can have multiple sensors and can only deployed on one animal at a time. A unique identifier (Tag ID) must be linked to every event in Movebank. Because tags are referenced within studies, if multiple studies refer to the same tag, there will be multiple records of it in Movebank. In these cases it is best practice for users to assign the same identifiers to the tag across all studies. Users can store reference data with additional information about the tag, such as manufacturer, mass or serial number. Within a study, the user-provided tag identifier is stored as tag-id in reference data and as tag-local-identifier in event data. Each tag also has an internal Movebank database identifier that is sometimes shown.
A sensor represents a specific sensor, such as a GPS or accelerometer, on a tag. A sensor type, chosen from a controlled list, must be linked to every event in a study, and multiple sensors can be linked to the same tag. If an event contains measurements that were in reality made by more than one sensor (for example a GPS tag that records locations and temperature), the user assigns the most relevant sensor type, using a location sensor type if location estimates are included. Movebank distinguishes between location and non-location sensors. The sensor is assigned to each event with the data attribute sensor-type.
An event is a measurement or set of measurements recorded at or associated with a known time. An event is equivalent to one record or row from an imported data file. Events can include location estimates, measurements recorded by or derived from the tag (such as temperature or acceleration) and information provided by the data owner (such as a behavioural classification). Every event must contain an associated tag, sensor and timestamp. Because events are referenced within studies, if the same data are imported to multiple studies, there will be multiple events for it in Movebank. In these cases it is best practice for users to assign the same animal, tag and deployment identifiers to the event across all studies. Within a study, events are uniquely described using the data attribute event-id that is automatically assigned when data are imported.
A deployment represents the period during which a specific tag is attached to a particular animal. Deployments allow users to define when tags are deployed on and removed from animals in order to exclude pre- or post-deployment events that may be present and to properly manage tag redeployments. Because deployments are referenced within studies, if multiple studies refer to the same deployment, there will be multiple records of it in Movebank. In these cases it is best practice for users to assign the same identifiers to the deployment across all studies. Users can store reference data with additional information about the deployment, such as deployment end type, maniulation type and study site. Within a study, deployments are uniquely described using the data attribute deployment-id. Each deployment also has an internal Movebank database identifier that is sometimes shown.
A study is created by a user to store data. The scope of a study can vary depending on the user's needs. Users might create separate studies for specific field sites, field seasons, species, collaborations or collection methods, or for the data used in a published analysis. Data uploads, data downloads, assignment of sharing and permissions settings and definition of individual animals, tags and deployments are all done within a study. The Study Details provide summary information about a study. Additional considerations for creating and organizing studies can be found in Archiving Best Practices.
A file format is a defined mapping between attributes and values in an original data file and those stored in Movebank. File formats for live data feeds and other supported standard formats can be used in all Movebank studies, and file formats used to import custom tabular data can be saved within individual studies. Each file format has a name and an internal ID. The format name used to import a file can be shown in the "File Details" on the Studies page.
Reference data are information describing individual animals, tags and deployments. Reference data are used to manage deployments and store details that are critical to understanding and using tracking data, such as an animal's age at the time a tag was attached. Information that does not change over time (such as sex or ring ID) are linked to the animal, while information that changes over time (such as reproductive condition and age) are linked to the deployment. Technical specifications of the tag and sensors are linked to the tag. A limited amount of reference data—the animal and tag IDs and taxon—is included with downloads of event data tables from Movebank. The complete reference data are stored in a separate table.
For a current list of data terms used in Movebank, see the Movebank Attribute Dictionary. Movebank's vocabulary is published at the NERC Vocabulary Server hosted by the British Oceanographic Data Centre: http://vocab.nerc.ac.uk/collection/MVB/current
For more about Movebank's data model, see Kranstauber et al. (2011).