When it comes to wildlife conservation, land managers have relied mostly on expert opinion to determine whether a particular species might be endangered. There hadn't been a systematic way to know what is happening with wildlife using real data -- until now.
In 2013, Conservation International (CI) and Hewlett-Packard launched HP Earth Insights, a partnership that uses HP big data technology to improve the speed and accuracy of analysis of data on the biodiversity of tropical regions. The findings provide early warnings about threats to species and tropical forests.
Data is collected by the Tropical Ecology, Assessment and Monitoring (TEAM) Network, a coalition that includes CI, the Smithsonian Institution and the Wildlife Conservation Society.
Strategically placed sensor cameras take photos when an animal walks by -- producing more than 500,000 images per year. A data management tool allows users to take those images and collect information from them. An analytics system is then used for modeling and calculating statistics, taking into account variants such as current environment, human presence and land use changes.
"The correlation data can show on a per-species basis whether or not there's a significant impact" in a particular area, says Eric Fegraus, director of information systems for the TEAM Network. For instance, "if a species' downward trend correlates with human presence, then maybe we need to rethink where our trails in our park systems go," he says.
The project has amassed more than 6 million climate measurements, 2.3 million camera trap photos and 4TB of critical biodiversity data. Manu National Park in Peru is one of 16 sites in 15 countries that use the data.
"It is the only study that is monitoring animals for five years straight in a methodical way in the same place," says Patricia Alvarez-Loayza, the TEAM Network's site manager for Manu National Park. She notes that most studies end after one to three years and are not conclusive.
This story, "Conservation International" was originally published by Computerworld.