Data Sources
Assembling Source Data
Where and how you access the source data for your telemetry study will depend on the type of tag and vendor. Argos location data is available to users from the Argos website and many third party sites and repositories (e.g. movebank, sea-turtle.org) have API connections to ArgosWeb that can facilitate data access. Each tag manufacturer often has additional data streams included within the Argos transmission that require specific software or processing to translate. Both the Sea Mammal Research Unit (SMRU) and Wildlife Computers provide online data portals that provide users access to the location and additional sensor data.
Regardless of how the data are retrieved, these files should be treated as read only and not edited. ArgosWeb and vendors have, typically, kept the data formats and structure consistent over time. This affords end users the ability to develop custom processing scripts without much fear the formats will change and break their scripts.
Data Management
The quantity and diversity of data generated from bio-logging studies can quickly become overwhelming especially in situations where a single device is deployed for several months and collects data from multiple on-board sensors. For some research projects, storing these data in a central relational database (e.g. PostgreSQL) can have important benefits. If storing bio-logger data in a relational database, explore one that provides native support or extensions for spatial data types and operations (e.g. PostGIS, SpatiaLite). For many, though, simply storing data as and organized collection of plain text files (e.g. *.csv) is sufficient. This approach can be further improved by converting the comma-separated files to parquet files and leaning on the the Apache arrow package for R. These parquet files provide modern and efficient file compression, a columnar-based structure, and are optimized for big data. Even if a project starts as small data, it’s often useful to design your data management with growth in mind.
Wildlife Computers Data Sources
We have the most first-hand experience with data originating from Wildlife Computers devices and data that have been processed through the Wildlife Computers Data Portal. This doesn’t mean the techniques, packages, and analysis presented require Wildlife Computers devices. Or, that most of what is presented is not applicable to other bio-logging data sources. When possible, we’ve tried to create examples and processes that are agnostic to the type of deployed device.